← The Saudi aviation reference, in one place.
NTSB Safety Study - Weather-Related General Aviation Accidents (SS0501)
Reproduced for study. Always verify against the official GACAR at gaca.gov.sa.
Risk Factors Associated with Weather-Related
General Aviation Accidents
Safety Study
NTSB/SS-05/01
PB2005-917004
Notation 7565A
National
TransportationSafety Board
Washington, D.C.
National TransportationSafety Board
Washington, D.C.National Transportation
Safety Board
Washington, D.C. 20594
OFFICIAL BUSINESS
Penalty for Private Use, $300PRSRT STD
Postage & Fees Paid
NTSB
Permit No. G-200
Safety Study
Risk Factors Associated with Weather-Related
General Aviation Accidents
NTSB/SS-05/01
PB2005-917004 National Transp ortation Safety Board
Notation 7565A 490 LEnfant Plaza, S.W.
Adopted September 7, 2005 Washington, D.C. 20594
National Transportation Safety Board. 2005. Risk Factors Associated w ith Weather-Related General
Aviation Accidents. Safety Study NTSB/SS-05/01. Washington, DC.
Abstract: The goal of this National Tran sportation Safety Board study w as to better understand the risk
factors associated with ac cidents that occur in weat her conditions characterized by IMC or poor visibility
(weather-related accidents ). Safety Board air safety investigator s collected data from 72 general aviation
accidents that occurred between August 2003 and Ap ril 2004. When accidents occurred, study managers
also contacted pilots of flights that were operatin g in the vicinity at the time of those accidents for
information about their flight activity . A total of 135 nonaccident flights were included in the study. All
nonaccident pilots voluntarily consen ted to interviews and provided info rmation about their flights, their
aircraft, and details about their trai ning, experience, and demographics. Additionally, the Federal Aviation
Administration provided informatio n about pilots practical and writte n test results and their previous
accident/incident involvement. Statis tical analyses were used to determine the rela tionships between study
variables and accident/nonaccident stat us and to identify variables that co uld be linked to an increased risk
of weather-related general aviation accident involvemen t. The analysis revealed several pilot- and flight-
related factors associated with increased risk of accid ent involvement. The safety issues discussed in this
report include: 1) pilot age and training-related differen ces, 2) pilot testing, accident, and incident history,
and 3) pilot weather briefing sources and methods. Sa fety recommendations concerning these issues were
made to the Federal Aviation Administration.
The National Transportation Safety Board is an independent Federa l agency dedicated to promotin g aviation, railroad, highway, m arine,
pipeline, and hazardous ma terials safety. Estab lished in 1967, the agency is mandated by Congress through the Independent Safet y Board
Act of 1974 to investigate transportation acci dents, determine the probable causes of the accidents, issue safety recommendatio ns, study
transportation safety issues, and evaluate the safety effectiven ess of government agencies invol ved in transportation. The Safe ty Board
makes public its actions a nd decisions through accide nt reports, safety studie s, special investigation reports, safety recommen dations, and
statistical reviews.
Recent publications are av ailable in their entirety on the Web at <http:// www.ntsb.gov>. Other inform ation about available pub lications also
may be obtained from the Web site or by contacting:
National Transportation Safety Board
Records Management Division
490 LEnfant Plaza, S.W.
Washington, D.C. 20594
(800) 877-6799 or (202) 314-6551
Safety Board publications may be purchased , by individual copy or by subscription, fro m the National Techni cal Information Serv ice. To
purchase this publication, order report number PB2005-917004 from:
National Technical Information Service5285 Port Royal Road
Springfield, Virginia 22161
(800) 553-6847 or (703) 605-6000
The Independent Safety Board Act, as codified at 49 U. S.C. Section 1154(b), precludes the admi ssion into evidence or use of Boa rd reports
related to an incident or accident in a civil action for damages resulting from a matter mentioned in the report.
iii Safety Study
Contents
Acronyms and Abbrev iations ......................................................................................... v
Executive Summary ....................................................................................................... vii
Chapter 1: Background ....................................................................................................1
Historical Trends in IMC Accidents .........................................................................................1
Previous Safety Board Studies of Weathe r-Related GA Accidents ..........................................1
Previous Safety Board Reco mmendations Pertaining to Weather ............................................3
Other Research on Flight in IMC ..............................................................................................4
Chapter 2: Federal Aviation Regulations Pertaining to Instrument Flight ................ 7
Visual Flight Rules (VFR) Minimums ......................................................................................7
Training Requirements ......................................................................................................... .....7
Recency Requirements .......................................................................................................... ....8
Chapter 3: Aviation Weather .........................................................................................10
Preflight Requirements ........................................................................................................ ....11
Preflight Weather Information ................................................................................................. 11
In-Flight Weather Information ................................................................................................. 13
Chapter 4: Study Design and Methodology ..................................................................14
Case Control Methodology ...................................................................................................... 14
Study Procedures .............................................................................................................. .......15
Study Inclusion Criteria ...................................................................................................15
Matching Accident and Nonaccident Fl ights ...................................................................16
Identifying Pilots of Matching Nonaccid ent Flights.........................................................16
Interviewing Nonaccident Pilots ......................................................................................17
Accident Data ................................................................................................................. .........17
Additional Data Sources ....................................................................................................... ...18
Weather Data .................................................................................................................. ...18
Pilot History ................................................................................................................. .....18
Statistical Analyses ...... ....................... ...................... ....................... .................. .............. ........18
Chi-Square Analysis ..........................................................................................................1 9
Binary Logistic Regression Analysis ................................................................................20
Chapter 5: Results ..........................................................................................................21
Description of Study Accidents ...............................................................................................2 1
Pilot Information ............................................................................................................. .........21
Accident Pilots ............................................................................................................... ....21
Nonaccident Pilots ............................................................................................................ .24
Pilot Flight Hours ............................................................................................................ ..26
Pilot Testing History ......................................................................................................... 27
Contents iv Safety Study
Accident, Incident, and Violation History ........................................................................28
Flight Information ............................................................................................................ ........28
Planned Length of Flights .................................................................................................28
Purpose of Flight ............................................................................................................. ..30
Aircraft Information .......................................................................................................... ......30
Aircraft Ownership ............................................................................................................ 30
Aircraft Equipment ............................................................................................................ 31
Summary of Chi-Square Analysis ...........................................................................................31
Binary Logistic Regression Model...........................................................................................31
Flight Leg Length ............................................................................................................. .32
Age at Private Certificate ..................................................................................................34Prior Accident/Incident Involvement .... ............................................................................34
Instrument Rating ............................................................................................................. .34
Chapter 6: Discu ssion .....................................................................................................35
Age-Related Differences ....................................................................................................... ..35
Flight Training Differences ...............................................................................................37
Testing, Accident, and Incident History ..................................................................................39
Airman Knowledge and Prac tical Test Requirements ......................................................39
Test Performance and Accident Risk ................................................................................40
Accidents and Incidents ....................................................................................................41Pilot History of Accidents and Incidents ..........................................................................42
Tracking a Pilots Testing and Accident/Incident History ................................................42
Pilot Weather Briefing Sources and Meth ods ..........................................................................43
Summary ....................................................................................................................... ....46
Findings ...................................................................................................................... ......47
Recommendations ...........................................................................................................48
Appendixes
A: Previous National Transporta tion Safety Board Safety Re commendations .......................49
B: Basic VFR Weather Minimums ..........................................................................................59
C: Evaluation of Matching Procedure .....................................................................................60
D: Guided Interview ............................................................................................................. ...62
E: Supplemental Accident Form ............ .................................................................................65
F: A Comparison of Forecast and Actual Visib ility Conditions for Study Accidents ............67
G: Study Accidents .............................................................................................................. ...69
v Safety Study
Acronyms and Abbreviations
ADMS Accident Data Management System
AFSS automated flight service station
AIRMET airmens meteorol ogical information
ASI air safety investigator
ASOS automated surface observing system
ATC air traffic control
ATIS automated terminal information system
AWOS automated weather observation system
BFR biennial flight review
CFR Code of Federal Regulations
DUATS direct user access terminal system
FA area forecast
FAA Federal Aviation Administration
FBO fixed base operator
Flight Watch en route flight advisory service
FSS flight service station
GA general aviation
HIWAS hazardous inflight weather advisory service
IFR instrument flight rules
IMC instrument meteorological conditions
METAR aviation routine weather repor t/meteorological aerodrome
report
NEXRAD next generation weather radar
nm nautical miles
NOAA National Oceanic and Atmo spheric Administration
NWS National Weather Service
PATWAS pilots automatic telephone weather answering system
Acronyms and Abbreviations vi Safety Study
PIREP pilot report
PRIA Pilot Records Improvement Act of 1996
QICP Qualified Internet Communication Provider
SIGMET significant meteorol ogical information
TAF terminal aerodrome forecast
TIBS telephone information briefing system
TWEB transcribed weather en route broadcast
VFR visual flight rules
VMC visual meteorological conditions
VOR very high frequency om nidirectional range
WINGS FAA Pilot Proficiency Award Program
vii Safety Study
Executive Summary
Historically, about two-thirds of all general aviation (G A) accidents that occur in
instrument meteorological conditions (IMC) ar e fatala rate much higher than the overall
fatality rate for GA accident s. The goal of this Nationa l Transportation Safety Board
study was to better understand th e risk factors associated with accidents that occur in
weather conditions characterized by IMC or poor visibility ( weather-related accidents).
The study accomplished this goal usi ng the case control methodology, which
compared a group of accident fl ights to a matching group of nonaccident flights to identify
patterns of variables that distinguished th e two groups from each ot her. This methodology
expands on previous Safety Board efforts that have typically conc entrated on summaries
of accident cases. The advantage of the ca se control methodology is that, instead of
focusing on the factors that accidents have in common, and possibly being misled by
characteristics common to most pi lots/flights, it identifies ch aracteristics that set accidents
apart and contribute to their occurrence.
For this study, Safety Board air safety i nvestigators (ASI) collected data from 72
GA accidents that occurre d between August 2003 and Ap ril 2004. When accidents
occurred, study managers also c ontacted pilots of flights that were operating in the vicinity
at the time of those accidents for informati on about their flight ac tivity. A total of 135
nonaccident flights were included in the study. All nonaccident pilots voluntarily
consented to interviews and provided informa tion about their flights, their aircraft, and
details about their training, experience, and demographics . That information was
compared with data that regional ASIs collec ted about the accident fli ghts as part of their
normal investigations. Additionally, th e Federal Aviation Administration provided
information about pilots practic al and written test results and their previous accident/
incident involvement.
Statistical analyses were used to dete rmine the relationships between study vari-
ables and accident/nonaccident status and to id entify variables that could be linked to an
increased risk of weather-rela ted GA accident involvement. Th e analysis revealed several
pilot- and flight-related fact ors associated with increased risk of accide nt involvement,
and the recommended approaches to mitigating those risk factors are discussed in the con-
text of three issue areas:
1. Ensuring a minimum level of proficiency for all pilots to recognize and safely
respond to hazardous weather situations.
2. Identifying and providing additional support for pilots whose performance
history indicates an increased ri sk of weather-related accidents.
3. Providing GA pilots with additional guidance regarding s ources of preflight
weather information.
viii Safety Study
The Safety Board emphasizes that the c onclusions reached in this study are not
based on a summary of accident cases, alt hough the merits of such Board studies have
proven valuable in the past. Rather, the re sults are based on a st atistical comparison of
accident and nonaccident flight s that allows for the genera lization of findings from this
study to the wider population of GA pilots and flights that ma y be at risk for a weather-
related accident.
The Safety Board wishes to acknowledge th e significant contribution to this study
by the GA pilot community. All of the p ilots contacted voluntarily provided needed
information concerning their flight and aviation experience , providing the control group
necessary for this study.
As a result of this study, six recommendations were issued to the Federal Aviation
Administration.Executive Summary
1 Safety Study
Chapter 1
Background
Historical Trends in IMC Accidents
The total number of general aviation (GA) a ccidents per year has declined over the
past two decades. However, as shown in fi gure 1, the relative pr oportion of GA accidents
that occur during instrument meteorological conditions (IMC) has rema ined fairly stable,
ranging from 5 to 9 percent of annual GA accident totals. The National Transportation
Safety Board has long been
concerned about GA accidents that
occur in poor weather or in IMC
(referred to in this study as
weather-related accidents), espe-
cially because they are far more
likely to be fatal than accidents that
take place in visu al meteorological
conditions (VMC). Over the past
20 years, about two-thirds of all
IMC accidents have resulted in at
least one fatality, a rate that is threetimes higher than the fatality rate
of all GA accidents. In 2004, 103
accidents, or 6 percent of all GA
accidents, occurred in IMC. Of
these accidents, 70 percent were
fatal, and the 147 fatalities thatresulted from these accidents
accounted for more than 25
percent all deaths from GA acci-dents in 2004.
Previous Safety Board Studies of Weather-Related GA Accidents
The Safety Board published reports on weat her- or visibility-related GA accidents
in 1968,1 1974,2 1976,3 and 1989.4 The 1968 study considered in detail all GA accidents
1 National Transportation Safety Board, An Analysis of U.S. General Aviation Accidents Involving
Weather as a Cause/Related Factor 1966 , Notation 155 (Washington, DC: 1968).
2 National Transportation Safety Board, Special Study of Fatal, Weather-Involved General Aviation
Accidents, NTSB-AAS-74-2 (Washi ngton, DC: 1974).
3 National Transportation Safety Board, Nonfatal, Weather-Involved General Aviation Accidents ,
NTSB AAS-76-3 (Washington, DC: 1976).
4 National Transportation Safety Board, General Aviation Accidents In volving VFR Flight Into IMC,
NTSB/SR-89-01 (Washington, DC: 1989).On November 2, 2003, at 1847 central standard time, a
Commander Aircraft 114TC, N6 107Z, piloted by a private
pilot, was destroyed during an in-flight collision with terrain at
the Hutchinson Municipal Airpor t, Hutchinson, Kansas. The
pilot was fatally injured. Night instrument meteorological
conditions prevailed at the time of the accident
(CHI04FA025).
Chapter 1 2 Safety Study
that occurred in 1966 and included an analysis of the weather- and pilot-related causes and
factors cited in the Boar ds accident findings. The 1974, 1976, and 1989 studies each
examined a large number of accidents sa mpled over several years based on common
causal or contributing factors:
The 1974 study focused on fatal GA accidents involving weather.
The 1976 follow-up study focused on nonfatal weather-related GA accidents.
The 1989 study considered accidents in which VFR-into-IMC was cited as a
probable cause or contributing factor.
In each of these studies, descriptive sta tistics were used to characterize the
accident group in terms of fact ors relating to the pilot, airc raft, environment, and accident
circumstances. A dditionally, the 1989 study included some comparisons between the
VFR-into-IMC group and two other groupsone representing pilots in all GA accidents
and one representing al l active GA pilots.
In all of the studies, the selected gr oup of accidents typically occurred during
flights for which no fl ight plan had been filed. Th ey usually involved relatively
inexperienced5 pilots with private pilot certificates and no instrument ratings. Pilots in
fatal weather-related accidents were generally older than those in non fatal accidents, and
pilots involved in VFR-into-IMC accidents we re older than the comparison group of all
active pilots. The most co mmon weather phenomena for fa tal accidents were low cloudFigure 1
5 For example, in the 1974 study, 53 percent of pilots had less than 600 total flight hours. In the 1976 study,
84 percent of pilots had less than 100 total flight hours, and in the 1989 study, 52 percent of pilots involved in
VFR-into-IMC accidents had less than 500 total flight hours compared to 41 pe rcent of pilots in all GA accidents.The Proportion of GA Accidents that Occurred
During IMC Between 1983 and 2004
0%20%40%60%80%100%
Chapter 1 3 Safety Study
ceilings, fog, rain, and snow ; nonfatal accidents, on the other hand, were dominated by
accidents attributed to unfavorable wind.
Recommendations from the 1968, 1974, and 1976 st udies were directed to the Federal
Aviation Administration (FAA), the Environm ental Science Services Administration,6 and the
National Oceanic and Atmosphe ric Administration (NOAA). Recomme ndations from the
1968 and 1974 studies focused on increasing pilo ts knowledge and awareness of weather
through training and through expa nsion of weather forecasting and reporting products tailored
to pilots. Conversely, recommendations as sociated with the 1976 study emphasized hazards
associated with winds during th e landing phase of flight and di d not pertain to conditions of
reduced visibility. The 1989 report did not generate any new recommendations.
Previous Safety Board Recommendations Pertaining to Weather
In addition to the recommendations as sociated with safe ty studies, numerous
recommendations concer ning weather and visibi lity issues have resulted from accident
investigations. The Safety Boards Recommendations Databa se shows that the Board has
issued 82 recommendations rela ting to GA flight in IMC or visibility-related weather
conditions since 1968.7 These recommendations have addr essed a variety of topics, which
may be grouped into three broad areas: th e collection and disse mination of weatherFigure 2
6 The Environmental Science Services Administra tion was the predecessor agency to the National
Oceanic and Atmospheric Administration.
7 Recommendations were selected fo r this group if they affected GA flight in weathe r conditions, even
if they were issued in the context of an accident that did not involve GA operations.The Proportion of all GA Accidents and GA Accidents in
IMC that Resulted in a Fatality Between 1983 and 2004
20040%20%40%60%80%100%
IMC All GA
Chapter 1 4 Safety Study
information, pilot training and operations, and air traffic cont rol (ATC). Most of these
recommendations were addressed to FAA, NOAA, or the National Weather Service
(NWS),8 which is part of NOAA. Appendix A summarizes these recommendations and
lists each recommendation and its status.
Other Research on Flight in IMC
Weather-related accidents in general, a nd VFR-into-IMC accide nts in particular,
have generated considerable interest from other governmental agencies and the aviation
research community.9 Studies have focused on pilots evaluations and assessments of
deteriorating visibility10 or on their flight-related decisi ons in the presence of weather.11
Additionally, some researchers have suggested that a lack of good weather information
during flight contributes to the inci dence of weather-related accidents.12
Studies of accidents involving continued VF R flight into IMC often focus on pilot
judgment and the factor s influencing aeronautical decisi on-making. For instance, Jensen
and Benel13 concluded that approximate ly 50 percent of aviation fatalities were related to
poor pilot judgment. Examples of so-calle d poor judgment accidents in IMC included
pilots continuing VFR flight into IMC or pilots descending be low a published minimum
altitude while attempting an instrument appr oach. In some case s, these actions were
intentional violations of rules or safety procedures; in other cases, they were unintentional
or resulted from a misinterpre tation of available information.
Other researchers have at tempted to characterize th e types of decision-making
errors that lead pilots to make unsafe decisions. One class of decision-making error
attributed to pilots in we ather accidents is known as a plan continuation error . A plan
continuation error is defined as failure to revise a flight plan despite emerging evidence
that suggests it is no longer safe.14 For example, rather than revising the intended route
8 NWS responsibilities include providing weather fo recasts in support of av iation and the mission of
the FAA.
9 For a recent example, see General Aviation Pilot Behaviours in the Face of Adverse Weather
(Australian Transport Safety Board: June 2005).
10 D.A. Wiegmann, J. Goh, and D. OHare, Pilots Decisions to Continue Visual Flight Rules (VFR)
Flight into Adverse Weather: Effects of Distance Traveled and Flight Experience, FAA Technical Report,
ARL-01-11/FAA-01-3 (2001).
11 B. Burian, J. Orasanu, and J. Hitt, Weather-Related Decision Errors: Differences Across Flight
Types, Proceedings of the 14th Triennial Congress of the International Ergonomics Association/44th Annual
Meeting of the Human Factors and Ergonomics Society (San Diego, CA: 2000), 1, 22-25.
12 K. Latorella, S. Lane, and D. Garland, General Aviation Pilots Perceived Usage and Valuation of
Aviation Weather Sources , NASA Technical Memorandum 211443 (2002).
13 R.S. Jensen and R.A. Benel, Judgment Evaluation and Instruction in Civil Pilot Training, Final
Report FAA-RD-78-24 (Springfield, V A: National Technical Information Service, 1977).
14 J. Orasanu, L. Martin and J. Davison, Cognitiv e and Contextual Factors in Aviation Accidents, in
E. Salas and G.A. Klein (Eds.), Linking Expertise and Naturalistic Decision Making (Mahwah, NJ:
Lawrence Erlbaum Associates: 2001), pp. 209-225.
Chapter 1 5 Safety Study
of flight by changing course or altitude, deviating to an alte rnate airport, or returning to
the departure airport, pilots may opt to press on into deteri orating weather. Another type
of decision-making error can occur when pilots continue visual fl ight into instrument
conditions because they incorrectly assess the risks of the situation.15 In these cases, pilots
who appear to be intentionall y engaging in risky behavior may actually be making choices
that they mistakenly believe to be safe. Such diagnostic errors have been found to result in
more serious accidents than have errors in aircraft control.16 Even if pilots are able to
correctly assess current weather conditi ons, they may still underestimate the risk
associated with continued fl ight under those condi tions, or they may overestimate their
ability to handle that risk.17
Errors in decision-making, such as plan continuati on errors or incorrect
assessments of weather-relate d risk, may be made by pilots who are unfamiliar with the
climate of the local area, who lack total and/or recent experien ce identifying marginal
weather conditions, or who la ck experience accessing or re ading weather reports. For
example, operational experience with weather has been found to affect weather-related
decision-making and informati on acquisition. Targeted weathe r-related training programs
have demonstrated some success in t eaching pilots to recognize and respond to
deteriorating weather conditions.18
It is also possible that decision-making errors and/or accide nt involvement are
associated with more general pilot-related factors, such as tota l flight experience,
certification-level, or risk-t aking behavior. However, comp arisons of pilots experience,
self-assessment of ability, and actual accident involveme nt have yielded conflicting
results.19 For example, low-time pilots may l ack the experience they need for making
decisions, whereas high-time pilots may misj udge the risk associated with weather
conditions because they have successfully ope rated in similar weather conditions in the
past. A comparison of pilot su rvey data with past and fu ture accident involvement found
that individual differences in attitude toward risk-t aking behaviors made certain
individuals more likely to be involved in an accident , regardless of experience.20 The
15 D. Wiegmann and J. Goh, Visual Flight Rules (VFR) Flight Into Adverse Weather: An Empirical
Investigation of Factors Affecting Pilot Decision Making , FAA Technical Report ARL-00-15/FAA-00-8
(Washington, DC: FAA, 2000).
16 D. Wiegmann and S.A. Shappell, Human Factors An alysis of Postaccident Data: Applying Theoretical
Taxonomies of Human Error, The International Journal of Aviation Psychology, 7 (1997): 67-81.
17 D. OHare, Pilots Pe rception of Risks and Hazards in General Aviation, Aviation, Space, and
Environmental Medicine, 61 (1990): 599-603.
18 (a) M. Wiggins and D. OHare, Expertise in Aero nautical Weather-Related Decision Making: A Cross-
Sectional Analysis of General Aviation Pilots, Journal of Experimental Psychology: Applied, 1(4) (1995):
305-320; (b) M. Wiggins and D. OHare, Weatherwise: Evaluation of a Cue-Based Training Approach for the
Recognition of Deteriorating Weat her Conditions During Flight, Human Factors 45(2) (2003): 337-345.
19 (a) D.R. Hunter, Airman Research Questionnaire: Methodology and Overall Results, Report No.
DOT/FAA/AM-95/27 (Washington, DC: FAA, 1995); (b ) D.R. Hunter, Retrospective and Prospective
Validity of Aircraft Accident Risk Indicators, Human Factors, 43(4) (2001): 509-518; (c) M. Lubner, A
Risk Profile for Aviation A ccidents, Incidents, and Violations Among U.S. Pilots, Proceedings of the 9th
International Symposium on Aviation Psychology (Columbus, OH: 1997).
20 Hunter, 2001.
Chapter 1 6 Safety Study
same study found that previous accident involvement was associ ated with future accident
risk. Some authors have also suggested that pilots willi ngness to accept weather-related
risks depends upon their tendency to focus on ei ther the gains or losses associated with
each option.21 In addition to differences in individual decision-making style, the
cost/benefit determination may be further in fluenced by the amount of time, money, and
effort a pilot has already inve sted in a particular flight.
A substantial amount of ex isting research has used questionnaires or flight
simulators in laboratory settings. Laboratory st udies allow for scientif ic control but fail to
replicate the complex demands placed on pilo ts conducting real-world flight operations.
For example, flight simulator studies have provided a better understanding of the types of
evaluation and decision-making errors that ca n lead to weather-related accidents. Far
fewer studies have linked specific pilot, aircra ft, or flight-related fa ctors to the occurrence
of actual weather-related accident s. Because of its role as th e primary investigator of U.S.
civil aviation accidents, the Sa fety Board is uniquely suited to gather information related
to weather-related accidents and to identify the factors that distinguish them from
successful flights.
21 D. OHare and T. Smitheram, Pressing On into Deteriorating Conditions: An Application of
Behavioral Decision Theory to Pilot Decision Making, The International Journal of Aviation Psychology,
5(4) (1995): 351-370.
7 Safety Study
Chapter 2
Federal Aviation Regulations Pertaining to Instrument
Flight
Visual Flight Rules (VFR) Minimums
According to the FAA Pilot/Controller Glossary , instrument meteorological condi-
tions, or IMC, are defined as meteorologica l conditions expressed in terms of visibility,
distance from clouds, and ceiling less than th e minimums specified for visual meteorolog-
ical conditions (VMC).22 This legal definiti on is used to prescrib e the minimum weather
conditions in which a pilot may operate an aircraft without meet ing additional require-
ments of pilot qualification,
aircraft equipment, and communi-cation. These minimums account
for altitude in relation to terrain
and obstructions, time of day, and
location in relation to areas of high
air traffic density, such as airports.
An example of criteria defining
VMC are 3 statute miles flight vis-
ibility and a clearance of 1,000
feet above, 500 feet below, and
2,000 feet horizont al distance
from clouds. Me teorological con-
ditions are also commonlyreferred to as either VFR (visual
flight rules) or IFR (instrument
flight rules) in reference to theregulations that apply to pilots
who are operating under those
conditions.
Training Requirements
Because of the unique demands related to safely managing weather, the FAA has
established knowledge and traini ng requirements for pilots to help them recognize and
respond to weather hazards. A ll pilot applicantsregardle ss of certificate levelmust
receive training in how to a void potential weather hazard s when possible, and how to
safely respond to weather hazards if they are encountered. These training requirements
include the recognition and avoi dance of hazardous weather, preflight actions related to
22 Basic VFR weather minimums as defined in 14 Code of Federal Regulations (CFR) 91.155 are
presented in appendix B.On March 21, 2004, about 2050 eastern standard time, a
Piper PA-32R-301, N8173U, was destroyed when it
impacted mountainous terrain while in cruise flight near
Harlan, Kentucky. The cerf icated private pilot and five
passengers were fatally injured. Instrument meteorological
conditions prevailed near the accident site (NTC04FA092).
Chapter 2 8 Safety Study
the procurement and use of ae ronautical weather reports and forecasts, and aeronautical
decision-making and risk manage ment. For instrument-rated pi lots, this training is meant
to provide the additiona l knowledge and skills ne eded for safe flight in IMC. Conversely,
the goal of instrument flight training for VFR-only pilots is to enable them to maintain
control of an aircraft while making a course reversal or di version if they inadvertently
enter clouds. Specifically, privat e pilot applicants who are not instrument rated must have
logged a minimum of 3 hours of flight training on the cont rol and maneuvering of an
airplane solely by reference to instruments, including strai ght and level flight, constant
airspeed climbs and descents, turns to a heading, recovery from unusual flig ht attitudes,
radio communications, and the use of navigation systems/fa cilities and radar services
appropriate to instrument fli ght. Commercial pilot applicants who are not instrument rated
must have logged a minimum of 10 hour s of similar instrument training.23
Because aircraft used in recreationa l and sport operations are typically not
equipped for instrument flight, requirements for the recreationa l and sport pilot certificates
do not include flight training or proficiency standards for fl ight by reference to aircraft
instruments. Consequently, pilots holding recrea tional pilot or sport pi lot certificates are
prohibited from operating aircraft unless they maintain at l east 3 statute miles visibility
and visual contact with the surface.24 Additionally, recreational pilot certificate holders
may not operate an aircraft at night with less than 5 statute mil es visibility, and sport pilot
certificate holders may not operate an aircraft at night.
In order to pilot an aircra ft on an IFR flight plan in IMC, private pilot and
commercial pilot certificate holde rs must add an instrument ra ting to their pilot certificates
by completing additional training and pa ssing both knowledge a nd practical tests.
Instrument rating app licants are required by 14 Code of Federal Regulations (CFR) 61.65
to also receive additional ground training in how to obtain and use aviation weather
reports and forecasts, the forecasting of w eather trends based on that information and on
their personal observation of weather conditions , the safe and ef ficient operation of
aircraft under instrument fl ight conditions, the recogniti on and avoidance of critical
weather situations, and ae ronautical decision-making and judgment. The flight
requirements of the instrument rating include 40 hour s of actual or simulated instrument
flight, including at l east 15 hours of instrume nt flight training.25
Recency Requirements
Like any complex skill, flying an aircra ft solely by reference to aircraft flight
instruments requires periodic pract ice to maintain proficiency. Pilots with an instrument
23 Title 14 CFR 141, Appendix D, requires that co mmercial pilot training co nducted by a certificated
flight school include 5 hours of training in the same category and class of aircraft. Commercial pilot
applicants trained under this part must either hold an instrument rating or be concurrently enrolled in an
instrument pilot course.
24 Title 14 CFR 61.101 and Title 14 CFR 61.315, respectively.
25 Title 14 CFR 141, Appendix C.4, requires pilots trained at a certificated pilot school to have a
minimum of 35 hours of instrument training if the course is for an initial instrument rating or 15 hours ofinstrument training if the course is for an additional instrument rating.
Chapter 2 9 Safety Study
rating on their private pilot or commercial pilot certific ates must therefore meet
instrument flight recency re quirements in order to opera te aircraft in instrument
conditions. Title 14 CFR 61.57 states that no person may act as pilot-in-command under
IFR or in weather c onditions less than the minimums pr escribed for VFR unless, within
the preceding 6 calendar months, that pers on has conducted at least 6 instrument
approaches, holding procedures, and intercep ting and tracking courses through the use of
navigation systems in actual or simulated instrument conditi ons. Pilots who do not meet
this requirement have an additional 6 months after the prescribed time to meet the
requirement, but may not act as pilot-in-com mand under IFR. Pilots who do not meet the
requirement after that must pass an instrume nt proficiency check given by an authorized
instructor, examiner, or check pilot.
Even though they are subject to an initial certification requirement, non-
instrument-rated pilots are not currently re quired to receive recurr ent instrument flight
training, nor are they required to periodical ly demonstrate proficiency in flight by
reference to instru ments. All pilots are subject to 14 CFR 61.56, which specifies that, to
act as pilot-in-command of an aircraft, they must have satisfactorily completed a flight
review during the pr evious 24 months.26 That reviewcommonly referred to as a
biennial flight review or BFRmust include a minimum 1 hour each of ground and flight
instruction covering general knowledge, operating rules, and procedures. The instructor
giving the flight review is free to determine the content; therefore, the BFR may or may
not include a demonstration of the weather knowledge and instrument flight skills
required for initial certification.
26 Title 14 CFR 61.56 specifies several additional ways for pilots to satisfy the flight review
requirement, such as a practical te st for an additional certificate or rating, a pilot proficiency check, or
recurrent training, such as the FAA voluntary Pilot Proficiency Award Program (WINGS).
10 Safety Study
Chapter 3
Aviation Weather
Over the years, the measuring and repo rting of aviation weather conditions has
shifted from human observations to the use of automated systems. For example, upper-air
weather data collection that once relied on sounding balloons and pilot reports (PIREP)
has been supplemented by information gathered from wind profilers and aircraft equipped
with instruments that automatically downli nk weather observations. In addition, manual
surface observations have now been augmen ted or replaced by th e automated surface
observing system (ASOS) and the automa ted weather observing system (AWOS).
Established in the 1990s, ASOS and AWOS sy stems are now installed at more than 1,500
airports in the United States.
Radar and satellites have
also become major sources of
weather information. In the 1980s
and 1990s, using Doppler technol-
ogy, the NWS developed next-
generation radar (NEXRAD),
which provides advanced anddetailed information about precipi-
tation and winds. Weather
satellites provide additional infor-
mation in the form of visible,
infrared, and other images that are
made available on a near-real-time
basis to NWS and FAA facilities.
Pilots use of weather
information for flight-planning anddecision-making has also changed
greatly over the years. Up until the
late 1960s, most pilots planned
their flights with the aid of face-to-
face weather briefings from trained
FAA personnel working at flight service st ations (FSS). In the 1970s, telephone briefings
from FSS and prerecorded weather informat ion like the telephone information briefing
service (TIBS) and the pilots automatic telephone weather answering system (PATWAS)
became commonplace. In the cockpit, pilots ga ined access to radio br oadcasts of airport
information, including current weather conditions , using the automatic terminal information
service (ATIS). The 1970s also saw the advent of in-flight we ather information like the en
route flight advisory service (Flight Watch), transcribed weather broadcasts (TWEB), and
the hazardous in-flight weathe r advisory service (HIWAS).On February 11, 2004, about 2110 eastern standard time, a
Piper PA-28-180, N6473J, collided with trees and the ground
during approach to Bacon County Airport, Alma, Georgia.
Instrument meteorological cond itions prevailed. The private
pilot, the pilot-rated passenger, and the rear-seated
passenger received fatal injuries, and the airplane sustained
substantial damage (ATL04FA075).
Chapter 3 11 Safety Study
Today, many pilots receive preflight weat her briefings via computer using the FAA
direct user access terminal syst em (DUATS). Additionally, In ternet websites and computer
terminals displaying live radar and graphical weather data ar e available at airports and
fixed base operators (FBOs) to support flight planning with supplemental weather
information. For local and en route applic ations, real-t ime weather information from
ASOS has joined many of the existing in -flight weather info rmation sources.
Most preflight and in-flight weather briefing services employ text-based or audio-
based products like aviation routine weather re ports (METAR), area fo recasts (FA), terminal
aerodrome forecasts (TAF), ai rmens meteorological informat ion reports (AIRMETs), and
significant meteorological information reports (SIGMETs). Additionally, numerous graphical
products are now availabl e to pilots. For example, surfa ce analysis and weather depiction
charts portray recent atmospheric pressure patterns and surface we ather observations from
across the United States. Ra dar summary charts provide info rmation about the location and
intensity of thunderstorms and other forms of precipitation, and low-level prognostic charts
provide forecast information pertaining to IM C, turbulence, and icing for 12- and 24-hour
periods. Graphic products like th ese, as well as radar and sa tellite images, are available to
pilots on the ground, and new servi ces are making them available in the cockpit with the aid of
datalink or satellite communications a nd advanced multifunctional displays.
Preflight Requirements
Title 14 CFR 91.103 requires th at all pilots, before beginning a flight, become
familiar with all available in formation concerning that flight . Specific examples of such
information are airport elev ation and runway lengths, aircraft takeoff and landing
distances, and aircraft gross we ight. Pilots of IFR flights a nd non-local flights must also be
familiar with weather reports and forecasts, fu el requirements, and alternatives available if
the planned flight cannot be completed. When considered with 14 CFR 91.3, which
stipulates that pilots-in-comm and are directly responsible fo r, and the final authority as
to, the operation of their aircraft, pilots are solely responsible for knowing everything
there is to know about the weather along their planned route of flight.
Preflight Weather Information
Preflight weather informati on enables pilots to prepar e for conditions they may
encounter during flight, or to decide that they are not qualified to fly in the reported
conditions. Unlike regulations governing Parts 121 and 135, whic h limit operators to the
use of approved weather sources,27 Part 91 regulations do not specify a particular source
27 According to the FAA Air Transport Operations Inspectors Handbook (8400.10), for all operations
conducted under Parts 121 and 135, weather reports e ither must be prepared by the National Weather
Service (NWS) or by sources approved by the NWS or Federal Aviation Administration (FAA). These
sources include NWS offices, FAA Flight Service St ations, Automated Surface Observations, and certain
Qualified Internet Communication Providers (QICP).
Chapter 3 12 Safety Study
of weather information for GA pilots. The FAA and NWS provide numerous sources of
weather information to pilots, including FSS telephone briefings, DUATS, and Internet
website services such as <http://www.aviationweather.gov >. Commercial providers also
offer data and services, such as satellite weat her or flight planning software packages that
include weathe r information.
In its Instrument Flying Handbook (FAA-H-8083-15), the FAA suggests that, in
preparing for an IFR flight, p ilots should call an automated fl ight service st ation (AFSS) to
obtain a weather briefing. A s tandard preflight briefing is meant to provide pilots with
sufficient weather info rmation to prepare for their flight s, and is defined by FAA Advisory
Circular AC00-45E as comprising the following components:
1.Adverse conditions. Meteorological or aeronaut ical conditions reported or
forecast that may influence a pilot to alter the proposed flight.
2.VFR flight not recommended. VFR flight proposed and sky conditions or
visibilities present or for ecast, surface or aloft, th at, in the judgment of the
AFSS/FSS briefer, would ma ke flight under VFR doubtful.
3.Synopsis. A brief statement de scribing the type, locat ion, and movement of
weather systems and/or air masses that might affect the proposed flight.
4.Current conditions. A summary from all available sources reporting weather
conditions applicable to the flight.
5.En route forecast. A summary from appropriate data forecast conditions
applicable to the proposed flight.
6.Destination forecast. Destination forecast that includes significant changes
expected within 1 hour before and after the expected time of arrival.
7.Winds aloft. Forecast winds aloft for the proposed route; temperature
information on request.
8.Notices to airmen. A notice containing information concerning the
establishment, condition, or change in any component of, or hazard in, the
National Airspace System.
9.ATC delays . Any known ATC delays and/or flow control advisories that may
affect the proposed flight.
10.Request for PIREP. A request made if a repor t of actual in-flight conditions
would be beneficial or wh en conditions meet the cr iteria for so licitation of
PIREPs.
11.Flight Watch. The availability of in-flight weather informati on (for example,
Flight Watch) fo r weather updates.
12. Any other information the pilot requests (for example, military training activity
along the route of flight).
Chapter 3 13 Safety Study
In-Flight Weather Information
Each time they fly, pilots must continuou sly evaluate in-flight weather to identify
conditions that could potentially affect the safety of their fli ghts. In so doing, pilots must
integrate what they see outside the cockpit wi th data presented by aircraft instruments and
any additional information they receive from ATC, weather reporting facilities, and other
pilots. Pilots must then compare that info rmation against their e xpectations based on the
forecast weather.
Pilots without a current instrument rating, or those flying aircraft certified for VFR
flight only, must base their decisions on the need to maintain minimum VFR cloud
clearances at all times. A pilot may be able to change altitude or deviate from the planned
route of flight in order to maintain VFR requirements. If not, the pilot must either turn
back or find a suitable alte rnate route or destination.
Because they are permitted to operate in degraded weather conditions, instrument-
rated pilots operating on IFR flight plans ar e more concerned with avoiding hazards like
icing, turbulence, and embedded thunderstorms that they ma y encounter when flying in
the clouds. Detection and avoida nce of in-flight we ather hazards require pilots either to
use weather avionics like rada r or lightning det ection equipment or to communicate with
ATC or Flight Watch to obtain this inform ation. Pilots can comm unicate directly with
most local FSSs through Flight Watch or HIWAS, or if controller workload allows, request
weather information through ATC. Recentl y, aircraft-based equipment and radio
communications have begun to merge with avionics capable of displaying weather data,
such as METARs, TAFs, FAs, AIRMETS, SIGMETS, and PIREPs transmitted through
either a ground-based array or a satellite network.28
28 For an example of a system that combines datalinked traffic, w eather, and terrain information with
advanced aircraft displays, see the FAAs Alaska Capstone project, <http://www.alaska.faa.gov/capstone/ >.
14 Safety Study
Chapter 4
Study Design and Methodology
The goal of this study was to identify factors that are predictive of accident risk for
GA pilots flying in weather-re lated conditions characterized by poor visibility, like IMC
or marginal VMC.29 This goal
was accomplished by comparing
a set of weather- related GA acci-
dents to a corresponding set ofnonaccident flights that tookplace under similar weather con-
ditions. This approach, referred
to as the case control methodol-ogy, allows for a determination of
the odds of involvement in acci-
dents based on identified risk
factors. The advantage of this
approach is that, instead of focus-
ing on what accidents have in
common, and possi bly being mis-
led by characteristics common tomost pilots/flights, it identifiescharacteristics that set accidents
apart and contribute to their
occurrence.
Case Control Methodology
The case control methodology is commonly used in epid emiological research to
compare a group of interest, such as people with a certain dis ease (that is, cases) with a
group of individuals from the same population who do not exhibit the disease (that is,
controls). Control groups may be random ly selected from wi thin the population of
interest or may be selected to match case s on certain variables, such as age, sex, or
exposure to potential risk factors.
In the aviation domain, the case control methodology has been used in only a handful
of studies. For example, one study compared characteristics of accide nt pilots to a larger pilot
sample and found that accident pilots were more likely to have medical problems than the
29 Marginal VMC is used to refer to conditions in which visibility is between 3 and 5 miles or when
ceiling height is between 1,000 and 3,000 feet.On January 10, 2004, about 1840 central standard time, a
Cessna 182P, N5787J, was destroyed during an in-flight
collision with trees and terrain 7 1/2 miles southwest of the
Baudette International Airport, Baudette, Minnesota. Night
instrument meteorological cond itions prevailed at the time
of the accident. The non-instrument-rated pilot and
passenger sustained fatal injuries (CHI04FA055).
Chapter 4 15 Safety Study
overall pilot population.30 A similar approach was used to a lesser extent in the Safety
Boards 1989 safety study, which found that pilo ts involved in VFR-into-IMC accidents were
older and less likely to have an instrument rating than a sample of all active GA pilots.31
Other case control studies have gathered preexisting survey or accident data from
groups of pilots and then cate gorized the data by a variable of interest, like the pilots
involvement in an accident or an accidents fatal or nonfat al outcome. For example, one
study, which compared fatal-to-the-pilot GA cr ashes to those in which the pilot survived,
found that aircraft fires, of f-airport locations, nighttime fl ight, and IMC were linked to
pilot fatality.32 Another study examin ing predictors of pilo t fatality among weather-
related GA accidents result ed in similar findings.33
One challenge associated with using histor ical accident records or surveys is that
the data they contain may be incomplete or in adequate to address the research goals. The
present study addressed that ch allenge by identifying variables of interest a priori, which
facilitated more consistent and complete data collection from cases and controls.
Study Procedures
Study managers from the Safety Board s Office of Research and Engineering
worked closely with air safety investigators (ASI), air traffic specialists, and Safety Board
meteorologists from the Office of Aviation Safety. When GA accidents occurred during
the study time frame (August 2003 through April 2004), ASIs used preliminary
information to determine if the accident met the study inclus ion criteria and notified the
study managers accordingly. Study managers al so monitored FAA incident/accident daily
reports to identify additional qualifying accidents.
Study Inclusion Criteria
Before data collection began, study managers established a set of criteria to guide
the selection of accident cases. Study inclus ion criteria were base d on initial observations
of the weather at the time of the accident and characteristics of the accident flight.
Accidents were selected if they involved a GA operation and happene d in either IMC or
marginal VMC. Also included were accident s that appeared to have involved spatial
disorientation, loss of control, or collision with te rrain or object due to a lack of visual
references or encounter with weather.
30 C.R. Harper, Physical Defects of Civilian Pilots Related to Aircraft Accidents, Aerospace
Medicine, 35 (1964): 462-464.
31 NTSB/SR-89-01.
32 G. Li and S.P. Baker, Correlates of Pilot Fatality in General Aviation Crashes, Aviation, Space, and
Environmental Medicine, 70 (1999): 305-309.
33 J. M. Price and L.S. Groff, R isk Factors for Fatal General Aviation Accidents in Degraded Visual
Conditions, Proceedings of the 13th International Symposium on Aviation Psychology (Oklahoma City, OK,
2005), 469-474.
Chapter 4 16 Safety Study
Matching Accident and Nonaccident Flights
Once a candidate accident wa s identified, staff began to identify and locate
matching nonaccident flights based on the factors listed below.
Weather conditions
Location (within 30 miles of accident)
Time (within 30 minutes of accident)
Rules of flight
Number of engines
Engine type
For example, if an accident flight was operating under VFR, corresponding
nonaccident flights were select ed from other VFR flights operating in similar conditions
and in similar aircraft near the accident site. A subsequent review of data from accident
and nonaccident flights, presented in appendi x C, confirmed the su ccess of the matching
procedure.
Identifying Pilots of Matc hing Nonaccident Flights
If the accident flight was operating on an IFR flight plan, nonaccident aircraft were
identified with the use of commercia lly available flight tracking software.34 Information
presented by the software is based on composite radar data used by the FAA for system-
wide traffic monitoring and fl ow control. As shown in fi gure 3, the software provided
graphical displays of aircraft radar data, as well as su ch weather products as NEXRAD
radar. After an accident occurred, Safety Bo ard staff used the software to review flight
traffic and weather information around the accident site at the time of the accident and to
obtain registration numbers of matching nonaccident flights.
Because the software depicted only f lights receiving ATC radar services, an
alternative procedure was used for accident flights that were operating under VFR. For
those flights, study managers te lephoned airports and FBOs with in a 30-mile ra dius of the
accident and along the accident r oute of flight to identify matching aircraft. Airport and
FBO employees were informed about the st udy and its purpose and were asked if they
were aware of any aircraft that had arrived, departed, or passed through the area around
the time of the accident. An ef fort was also made to identify pilots who intended to fly but
chose not to because of the weather. For VF R accidents, multiple i nquiries were typically
necessary to identify appropr iate nonaccident pilots. In addition to telephone inquires,
primary radar was used when n ecessary to determine whether any other aircraft had
passed through the area.
34 Flight Explorer Professional (Version 4.6) [Com puter software]. Alexandria, V A: Flight Explorer,
Inc.
Chapter 4 17 Safety Study
Interviewing Nonaccident Pilots
Once candidate nonaccident aircraft ma tches were identified, a study manager
contacted the registered owners via telephone. Aircraft owners were informed about the
purpose of the study and allowed either to accept or decline pa rticipation. If the owner was
not piloting the aircraft during th e flight of interest, or if the aircraft was being operated
for rent, the flying pilot was identified and c ontacted. In most cases , study managers were
able to identify and interview pilots within 72 hours of the accident flight.
Of the aircraft pilots who were contact ed, 100 percent volunteered to participate in
a structured interview designed to elicit details of their flight s, available equipment in the
aircraft they were flying, and their previous flight experience. A copy of the interview
form is included in appendix D. Much of the information collected from nonaccident
pilots corresponded to informat ion typically collected as pa rt of Safety Board accident
investigations. No personally id entifiable information pertai ning to nonaccident pilots was
retained after the collec tion of study-related data.
Accident Data
Every effort was made to assemble matching data concerning both the accident
and nonaccident flights. Safety Board accident investigators routinely collect a core set of
factual data for aviation accident investigat ions and for populating a census of all U.S.
civil aviation accidents. This census, known as the Aviation Accident/Incident Database,
includes details about accident events (for exampl e, time, location, and weatherFigure 3. Screenshot of flight tracking software depicting weather,
an accident aircraft, and potential nonaccident matches.
accident
aircraft
potential
nonaccident
matches
Chapter 4 18 Safety Study
conditions), aircraft (for example, mode l type, inspection records, and engine
specifications), and pilots (for example, certificates, ratings, a nd flight hours flown).
Most of the data needed for this study were available using the Aviation
Accident/Incident Database . For data not typicall y collected during accident
investigations, the investigat or completed a supplemental fo rm and sent it to the study
managers. The supplemental fo rm, shown in appendix E, re quested such information as
the number of instrument instruction hours r eceived and whether the aircraft had various
equipment used to display navi gation and weather information.
Additional Data Sources
Weather Data
To determine the forecast weather conditi ons for each accide nt, Safety Board
meteorologists requested weathe r products that were released immediately before the
accident airplanes departur e time for the region surrounding the accident location. The
actual conditions at th e time and location of the accident were determined from the Safety
Board factual report a ssociated with the accident. (A s part of all factual reports,
investigators identify the b asic weather conditions as either IMC or VMC.) A
subsequent analysis, described in appendix F, found that in many cases, the weather at the
time of the accident was similar to forecast conditions.
Pilot History
FAA records were used to obtain a variety of historical data for both accident and
nonaccident pilots, including prev ious accidents, incidents, a nd violations, as well as the
results of all knowledge and pract ical tests. Testing records included the type (for example,
private, instrument, or commerci al), date, and outcome (pass or fail) of each practical and
knowledge test taken by a pilot. Practical test records in cluded each test attempted, the
outcome of each test, a nd the flight hours reporte d by the applicant at the time of the test.
Knowledge test reports were copies of the resu lts presented to test applicants, which typically
include the number of tests attempted and the percentage score for the most recent attempt.
Additional pilot data, such as certificates, ratings, and total flight hours reported
when pilots renew their medical certificate, were also obtained from the FAA Airmans
Registry. This information was used to substi tute for missing data (f or example, when log
books were destroyed in an accident.)
Statistical Analyses
Study analyses involved initial comparis ons of the accident and nonaccident
groups with regard to indivi dual variables of interest, fo llowed by development of a
multivariate prediction model. Using statisti cal methods typical of case control studies,
Chapter 4 19 Safety Study
initial comparisons were c onducted using chi-square anal yses, and the multivariate
prediction model wa s developed using binary logistic regression. All stat istical analyses
were conducted using the SPSS statistical package.35
Chi-Square Analysis
Data obtained from accident flights (cases) and from matching nonaccide nt flights
(controls) were used to compare accident risk ba sed on factors specific to the pilot, flight,
and aircraft.36 Chi-square tests were used to dete rmine the extent to which accidents and
nonaccidents differed on individual variables. Chi-square is a statistical test that can be
used to determine whether two or more samp les differ significantly with respect to the
proportional distribution of a given characteristic or qua lity. Tests of statistical
significance provide a measur e of the probability that a particular finding was due to
chance. If that probability is found to be very low (for ex ample, 5 percent or less), the
finding is considered stat istically significan t and thereby unlikely due to random
differences. Statistical signi ficance tests are sensitive to the number of observations
included in the study sample, as well as to the size of any observed difference(s).
In the context of this study, chi-square st atistics were used to determine whether
the study accident and nonaccident groups were different enough with regard to variables
of interest to suggest that study results represent GA weather accidents in general.
The following variables were selected for the chi-square analysis:
Pilot Information
Pilot age at the time of the accident
Pilot age at the time of initial private certification
Years as pilot
Pilot highest certification level and instrument rating
Pilot flight hours
FAA knowledge test performance
FAA practical test performance
Accident/incident history Violation history
Flight Information
Planned length of flight
Purpose of flight
35 SPSS for Windows, Rel. 13.0. 2004. Chicago: SPSS Inc.
36 A table containing data used in this study is available in the NTSB Docket Management System.
Chapter 4 20 Safety Study
Aircraft Information
Aircraft ownership
Aircraft equipment
Binary Logistic Regression Analysis
Binary logistic regression was used to identify the variables that were most
predictive of acci dent involvement and to develop an overall accident prediction model.
Binary logistic regression evaluates the combined effect of a set of predictor variables on a
dichotomous outcome variable (for example, accident or nonaccide nt status). One
strength of binary logistic re gression is that the results not only indicate which variables
are associated with accident i nvolvement, but also provide an estimate of the relative risk
associated with different leve ls of a given predictor variablefor example, the accident
risk associated with fl ights of various lengths.37 For this study, predictor variables were
selected for inclusion in th e regression model based on the initial findings from the chi-
square tests, as well as on hypothesized relationships be tween specific variables and
weather-related accidents.
37 For a more detailed discussion of binary logistic re gression, refer to the following: (a) D. Hosmer and
S. Lemeshow, Applied Logistic Regression, 2nd ed. (New York: Wiley and Sons, 2000) or (b) B.G.
Tabachnick and L.S. Fidell, Using Multivariate Statistics , 3rd ed. (New York: Harper Collins, 1996).
21 Safety Study
Chapter 5
Results
Description of Study Accidents
Study data were collected for GA accide nts that occurred between August 1, 2003,
and April 30, 2004,38 that appeared, at the time of in itial notification, to involve hazardous
weather or visibility-related factors. A tota l of 72 accidents met the predefined selection
criteria and were selected for the
study.39 Of these, 78 percent
were fatal accidents that resulted
in a total of 108 fatalities.
Figure 4 shows the
distribution of study accidents bymonth and figure 5 by time of
day. About 41 percent of
accidents occurred at night and
59 percent during daylight.
Additionally, about 56 percent
were operating on an instrument
flight plan and 44 percent
operated under VFR. A review
of the narrative da ta available for
the 72 study accidents showedthat 7 occurred during takeoff orclimb, 23 during cr uise flight, 17
during descent or maneuveringflight, and 25 during approach,landing, or go-around (figure 6).
Pilot Information
Accident Pilots
Accident pilots ranged in age from 23 to 81, with a m ean age of 53, and all pilots
but one were male. Accident pilot training re cords showed that the mean number of years
38 Between August 1, 2003, and April 30, 2004, th e Safety Board investigated a total 1,129 GA
accidents involving 1,144 aircraft. Of these, 226 were fa tal accidents that resulted in a total of 381 deaths.
39 Refer to appendix G for a list of the accidents included in this study.On December 7, 2003, about 1708 Pacific standard time, a
Piper PA-28R-200, N16264, collided with hilly terrain about 6
miles southwest of Chino Hills, California. The private pilot
and passenger were fatally injured and the airplane was
destroyed. Instrument meteorol ogical conditions prevailed in
the accident area, and no flight plan had been filed
(LAX04FA061).
Chapter 5 22 Safety Study
as pilot at the time of the accident was appr oximately 18 years and the mean age at which
accident pilots received their first pilot certificate was 35 year s. Of the 72 accident pilots,
4 percent held airline transport pilot certif icates, 32 percent commercial pilot certificates,
61 percent private pilot certif icates, and 3 percent student pilot certificates. As shown in
figure 7, the distribution of study accident pilots by highest ce rtificate was similar to theFigure 4
Figure 5Distribution of Study Accidents by Month
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
Distribution of Study Accidents
by the Time of Occurrence
Chapter 5 23 Safety Study
distribution of pilots invol ved in weather-related acci dents from 1983 through 2001.40 In
comparison, 2003 FAA airman registry data41 indicated that 23 percen t of active pilots at
that time held airline transport pilot certif icates, 20 percent commercial pilot certificates,
39 percent private pilot ce rtificates, and 14 percent student pilot certificates.42
Approximately 68 percent of accid ent pilots were rated for in strument flight, compared to
the 51 percent of the active pilot populat ion who held an instrument rating. Figure 6
40 Price and Groff (2005).
41 U.S. Department of Transportatio n, Federal Aviation Administration, U.S. Civil Airmen Statistics , 2003,
available online at <http://www.faa.gov/data_statistics/aviation_da ta_statistics/civil_airmen_statistics/>.
42 The remainder includes recreati onal pilot, rotorcraft (only), and glider (only) certificates.Figure 7Distribution of Study Accidents by Phase of Flight
Cruise Takeoff/Climb Descent/Manuever Approach/Landing/Go-around
Comparison of Study Pilots to Population Totals
and Previous Accident Research
0%20%40%60%80%100%
Student Private Commercial Airline Transport2003 Active
Pilots
(FAA Registry)Current Study
Nonacc PilotsWeather Acc
Pilots
(Price & Groff, 2005)Current Study
Acc Pilots
Chapter 5 24 Safety Study
Nonaccident Pilots
Interviews were conducted with 135 pi lots of matching nonaccident flights.
Nonaccident study participants included 131 male and 4 female pilots, ranging in age from
19 to 74 years, with a mean age of 46 years. The mean num ber of years as a pilot was
approximately 18 years for the nonaccident gr oup, and the mean age at which nonaccident
pilots received their first pilo t certificate was 28 years.
Data pertaining to pilot age at the time of the flight, year s as pilot, and age at first
certificate were cate gorized in four groups43 for analysis (figures 8, 9, and 10). The chi-
square test indicated a significant differen ce between the accident and nonaccident groups
with regard to age at the time of accident, χ2 (3, N = 207) = 12.33, p < .01.44 Specifically,
the nonaccident group included a higher perc entage of pilots in the youngest ( ≤40) group,
while the accident group included a higher percen tage of pilots in the oldest (>60) group.
However, there was no corresponding differen ce in the years of piloting experience
between the accident and nonaccident groups, χ2 (3, N = 207) = 1.52, p = .679. The
accident and nonaccident pilots did differ significantly by the age at which they first
obtained their pilot certificates, χ2 (3, N = 207) = 22.62, p < .001. Figure 10 illustrates that
a higher percentage of nonaccident pilots received their first pi lot certificates before age 25,
while the percentages represented in all other age groups were higher fo r accident pilots. .
43 Continuous variables were divided into discrete variables for analysis to simplify interpretation of
results. Discrete variable categories were determined by dividing continuous values into quartiles. The
resulting category cut-points were adjusted as necessary to create meaningful groups.
44 In this study, results were considered statistically significant if the probability of the result being due
to chance was less than 5 percent ( p < .05).Figure 8Distribution of Study Pilots by Age at Accident
0%20%40%60%80%100%
≤40 40.1-50 50.1-60 >60
Accident Nonaccident
Chapter 5 25 Safety Study
As shown in figure 7, 27 percent of the nona ccident pilots held airline transport
pilot certificates, 37 percent commercial pilot certificat es, 35 percent private pilot
certificates, and 1 percent stude nt pilot certificates. The chi-square analysis compared
pilots with student or private pilot certificates to those with commercial or transport pilotFigure 9
Figure 10≤5 5.1-15 15.1-25 >25Distribution of Study Pilots by
Years of Pilot Experience
0%20%40%60%80%100%
Accident Nonaccident
Distribution of Study Pilots
by Age at First Certificate
0%20%40%60%80%100%
Accident Nonaccident≤25 25.1-35 35.1-45 >45
Chapter 5 26 Safety Study
certificates and found a signi ficant difference between the accident and nonaccident
groups, χ2 (1, N = 207) = 15.21, p < .001, with a higher proportion of student and private
pilots in the accident group. Additionally, about 90 percent of pilots in the nonaccident
group held an instrument rating, compared to 68 percent of pilots in the accident group.
The proportion of pilots with in strument ratings differed signi ficantly for the accident and
nonaccident groups, χ2 (1, N = 207) = 16.28, p < .001.
Pilot Flight Hours
For most flight hour variable s (for example, total time in aircraft make and model,
actual instrument time, and ti me in the last 90 days), sta tistical comparis ons could not be
calculated because fli ght hour data for a large number of accident pilots were unavailable
due to incomplete or missing pilot logs. The most commonl y available in formation was
that for total flight hours in all aircraft, partly because p ilots are asked to report this
information when renewing a medical certifi cate. Using this information, study managers
were able to estimate total flight hours fo r accident pilots w hose logbooks could not be
recovered. The median total number of fl ight hours for the acci dent group (1,300 hours)
was lower than the median total flight hours for the nonaccident group (2,270 hours).
However, when pilots were di vided into groups fo r analysis by total flight hours (shown in
figure 11), the distributions were not found to be significan tly different for the accident
and nonaccident pilot groups, χ2(3, N = 207) = 3.89, p = .274.
Figure 11Total Flight Hours of Study Pilots
0%20%40%60%80%100%
≤500hrs 500.1-1500hrs 1500.1-3000hrs >3000hrs
Accident Nonaccident
Chapter 5 27 Safety Study
Pilot Testing History
Using test information fo r all accident and nonaccident pilots involved in the
study, study managers calculated separate knowledge a nd practical test pass rates for each
pilot using the results of his or her private, commercial, and in strument test(s). To allow
them to compare the testing hi story of all study pilots rega rdless of the number of tests
taken, study managers determined a cumulat ive pass rate by dividing the total number
of written or practical te sts passed by the total number of tests a pilot had taken.
Knowledge and practical test r ecords indicated that accident pilots had a mean cumulative
pass rate of 86 percent for knowledge tests (max = 100 percen t, min = 30 percent) and, for
practical tests, a mean cumulative pass ra te of 84 percent (max = 100 percent, min = 43
percent). Pilots in the nonacci dent control group had a mean cumulative pass rate of 95
percent (max = 100 percent, mi n = 59 percent) for knowledge te sts and, for practical tests,
a mean cumulative pass rate of 95 percen t (max = 100 percent, min = 50 percent).
Pilots were divided into tw o groups for the statistical analysis: a high pass rate
group for all pilots with a cumulative pass ra te of 70 percent or higher and a low pass
rate group for those with a cumu lative pass rate of less th an 70 percent. As shown in
figures 12 and 13, larger percentage s of accident pilots were in the low pass rate group for
both knowledge and practical test s. For the practical test, th is difference be tween accident
and nonaccident pilots was statistically significant, χ2 (1, N = 207) = 4.42, p = .036, with
28 percent of accident pilots and 16 percent of nonaccident pi lots falling into the low pass
rate group. The difference in knowledge test pass rates between the groups was
marginally significant, χ2 (1, N = 207) = 3.75, p = .053.
Figure 12Practical Test Pass Rate of Study Pilots
(private, instrument, and commercial pilot tests)
0%20%40%60%80%100%
Accident Nonaccident≥70% <70%
Chapter 5 28 Safety Study
Accident, Incident, a nd Violation History
A review of the accident a nd incident history of study pilots indicated that 19
percent of accident pilots had an accident or incident before the study accident. In four of
these cases, the event included in the study was the pilots third recorded mishap. In
contrast, figure 14 shows that a smaller proportion of pilots in the nonaccident control
group were ever involved in an accident or incident. The differences in accident and
incident history betw een groups was statis tically significant, χ2 (1, N = 207) = 3.99, p =
.046, with a larger percentage of accident pilots having been involved in a prior
occurrence.
In addition to accidents and incidents, FA A records indicated that 6 of the 72 study
accident pilots had civil aviati on violations on their records, in contrast to only 3 of the
135 nonaccident pilots. The desc riptive comparison suggests th at a higher percentage of
accident pilots had vi olation records, but because of th e small numbers of all study pilots
with violations, statistical analyses were not conducted for these values.
Flight Information
Planned Length of Flights
The mean planned length of a ccident flights was 232 nauti cal miles, compared to a
mean of 162 nautical miles for nonaccident fli ghts. Study flights were separated into four
groups according to the length of the planned flight (figure 15). Resu lts of the chi-square
analysis indicated a significant difference be tween the accident and nonaccident flightsFigure 13Knowledge Test Pass Rate of Study Pilots
(private, instrument, and commercial pilot tests)
0%20%40%60%80%100%
Accident Nonaccident≥70% <70%
Chapter 5 29 Safety Study
with regard to the inte nded length of flights, χ2 (3, N = 207) = 8.25, p = .041. As figure 15
shows, the accident group included a higher pe rcentage of planned flights in the longest
group (>300nm), and the nonaccide nt group included a higher per centage of flights in the
shortest planned length group ( ≤50nm).Figure 14
Figure 15Accident/Incident History of Study Pilots
0%20%40%60%80%100%
Accident NonaccidentNo Prior Acc/Inc Prior Acc/Inc
Distribution of Study Flights by Planned Length
0%20%40%60%80%100%
Accident Nonaccident≤50nm 50.01-150nm 150.01-300nm >300nm
Chapter 5 30 Safety Study
Purpose of Flight
As for the purpose of flight , the available information indicated that 17 percent of
accident flights were being operated for pay, and the re maining 83 percent were unpaid
operations. In contrast, 33 perc ent of the matching nonaccident flights were conducted for
pay, and the remaining 67 percent were unpai d. Analysis results indicated a significant
difference betw een the groups, χ2 (1, N = 207) = 6.54, p = .011. The differences in
proportion of paid flights ar e illustrated in figure 16.
Aircraft Information
Aircraft Ownership
As shown in figure 17, 76 percent of the aircraft involved in study accidents were
owned by the pilot either directly, through a limited liability co rporation, or through
shared ownership. About 10 percent of accident aircraft were rented, and the remaining 14
percent were owned by the pilo ts employer or another entity . In comparison, 56 percent
of the nonaccident aircraft were owned by the pilot, 13 percent were rented, and
approximately 30 percent were owned by the pilots employer or another entity.45
Differences in aircraft ow nership between the accident and nonaccident groups were
statistically significant, χ2 (2, N = 207) = 8.68, p = .013.Figure 16
45 Values do not sum to 100 percent due to rounding.Distribution of Study Flights by Purpose of Flight
0%20%40%60%80%100%
Accident NonaccidentNon-paid Paid
Chapter 5 31 Safety Study
Aircraft Equipment
Pilot interviews, aircraft records, and ex aminations of accident aircraft wreckage
were used in combination to collect info rmation about the equi pment on board accident
and nonaccident aircraft. The equipment inform ation available for ac cident aircraft was
markedly less than for nonaccident aircraft. Fift een of the accident aircraft in this study, or
about 21 percent, sustained impa ct and/or post-crash fire dama ge so extensive that little or
no information was available a bout aircraft avionics. For ma ny of the accident aircraft,
only partial equipment informat ion was available from airc raft maintenance logs and
repair and alteration records.46 The extent of missing data fo r accident cases, and the lack
of information about equipment use at the time of the accident, prevented study managers
from making meaningful comparisons of acci dent and nonaccident aircraft equipment.
Summary of Chi-Square Analysis
Overall, the chi-square analysis identifi ed several variables indicating significant
differences between the accident and nonaccident groups: 1) age at accident, 2) age at first
certificate, 3) highest certif ication, 4) instrument rating, 5) practical test cumulative pass
rate, 6) accident/incident hist ory, 7) planned length of fli ght, 8) purpose of flight, and 9)
aircraft ownership.
Binary Logistic Regression Model
Following the individual comparisons, a binary logistic re gression model was
developed based on the results of the chi-square analyses and relationships hypothesized byFigure 17
46 As reported on FAA form 337, Major Repair and Alteration.Distribution of Study Aircraft by Ownership
0%20%40%60%80%100%
Accident NonaccidentOther Own Rent
Chapter 5 32 Safety Study
other researchers. The decision to include an individual variab le in the logistic regression
model was ultimately made ba sed on the resulting effect on the ability of the model to
predict accident/nonaccident status. In some ca ses, variables were included (for example,
total flight time) because, although not statis tically significant on th eir own, they added to
the overall predictive ca pability of the regression model. In other cases, the strongest
predictor was selected from two or more rela ted variables (for example, age at private
certificate instead of age at accid ent). Table 1 lists each of th e variables included in the final
binary logistic regression mode l, along with the results of th e logistic regression analysis.
An overall test of the logi stic regression model signifi cance was calculated using a
chi-square test of the model co efficients. The results of that test indicated that the model
was statistically significant, χ2 (16, N = 207) = 57.45, p < .001. The results of the logistic
regression were also evaluated with regard to how accurately the model classifies accident
and nonaccident flights. In this case, the logistic regression model resulted in an overall
classification success rate of 77 percent (accurately identifying 121 of 135 nonaccidents
and 38 of 72 accidents) compared to 65 percent without the model.
The significance of individual variables within the model was assessed using the
Wald chi-square statistic, which tests the unique contribution of ea ch variable. The Wald
statistic is interpreted like the chi-square co mparisons presented earlier. Of the variables
entered into the model, planned length of flight, pilot age at first certificate,
accident/incident history, and instrument rating were identifi ed as unique and statistically
significant predictors of accident involveme nt. Table 1 includes the Wald statistic and
significance for all variables, and levels of variable, in the logistic regression.
For each predictor variable, the binary l ogistic regression an alysis also produced
odds ratios that reflected the re lative accident risk associated with different levels of the
variable. An odds ratio is interpreted as a comp arison of the risk asso ciated with each level
of a variable with the selected reference. For example, if a variable has two levels (for
example, yes and no), and the comparison of one level of the variable (for example, yes) to
the level chosen as the reference (for exampl e, no) results in an odds ratio of 2.0, the risk
of involvement in a weather-related GA accident for a flight with that characteristic is
twice as high as a flight that does not share that trait. Th e rightmost column of table 1
indicates the odds ratios associat ed with each level of variable.
Flight Leg Length
In comparison to a flight of 50 nautical miles (nm) or less, pilots on flights of
50.01 to 150 nm were 2 .9 times more likely to be i nvolved in a weather-related GA
accident, and those on flight s of more than 300 nm were 4.7 times more likely to be
involved in a weather-related accident.
Chapter 5 33 Safety Study
Table 1
Chapter 5 34 Safety Study
Age at Private Certificate
Pilots with the lowest risk of accident involvement we re those who received their
private certificates at or before age 25. Pilots who received th eir private certificates between
age 25 and 35 were found to be at 4.5 times greater risk than those in the youngest group of
being involved in a weather-rela ted accident, and those who recei ved their certificates between
35 and 45 were at 4.8 times greate r risk. Pilots who received thei r licenses after age 45 were 3.4
times more likely than the youngest group to be involved in a w eather-related accident.
Prior Accident/Incident Involvement
Pilots with a history of a ny type of accident or incide nt were found to be 3.1 times
more likely to be represented in the accid ent group than pilots with no such history.
Instrument Rating
Not having an instrument rating was asso ciated with significantly higher accident
risk. Specifically, pilots who did not hold an instrument rating were found to be 4.8 times
more likely than instrument-rated pilots to be involved in a weat her-related accident.
Chapter 6 35 Safety Study
Chapter 6
Discussion
This study compared GA
accident flights that occurred in
reduced visib ility conditions
(weather-related accidents)
with a corresponding set of non-
accident GA flights and
identified several pilot- and
flight-related fa ctors that are
associated with increased risk of
accident involvement. This sec-
tion focuses on several factors
that distinguished accident from
nonaccident pilots including their
age, flight training, and perfor-
mance history. It also discusses
the sources of weather informa-tion used by all GA pilots in this
study. The recommendations
issued as a result of the studyfindings focus on three specific
issue areas:
1. Ensuring a minimum level of proficiency for all pilots to recognize and safely
respond to hazardous weather situations.
2. Identifying and providing additional support for pilots whose performance
history indicates an increased ri sk of weather-related accidents.
3. Providing GA pilots with additional guidance regarding s ources of preflight
weather information.
Age-Related Differences
The relationship between pilot age and accident risk has been the subject of
research and debate, especially as it relates to the mandatory retirement age for Part 121
flight operations. St udies of age-related risk are subject to me thodological choices and On January 19, 2004, about 0106 eastern standard time, a
Piper PA-28-181, N298PA, impacted trees, a power line, and
the ground while on an instrume nt landing system approach to
the Saint Lucie County Airport, Fort Pierce, Florida. The
certificated flight in structor and the private pilot-rated student
were fatally injured, and a private-pilot rated passenger
received serious injuries. Instrument meteorological
conditions prevailed at the time and an instrument flight rules
flight plan was filed for the 14 CFR Part 91 instructional flight
(MIA04FA045).
36 Safety Study
analytical issues that can substant ially affect the resulting conclusions.47 As a result, prior
studies of age-related aviation risk have pr oduced remarkably different findings, including
a linear increase in risk with age,48 a decreasing risk with age,49 or no relationship between
age and accident risk.50
The analysis in this study identified signi ficant differences be tween accident pilots
and nonaccident control group pilo ts with respect to age, but the groups were not found to
be significantly different in e xperience as measured in years as a pilot or total flight hours.
The combination of these results indicated that pilots in the nonaccident group started
flying earlier in life, on average, than acci dent pilots. This fi nding was supported by the
chi-square analysis finding that showed a significant differ ence between the accident and
control groups with respect to age at certific ation, and the logistic regression analysis that
identified pilot age at private certificate as a signifi cant predictor of accident risk. Based
on the results of these analyses, the Safety Board concludes that pilots who start flying
earlier in life are at lower risk of being involved in a weather-related GA accident than
those who start flying when they are older, and age at first cer tificate is a better predictor
of future accident involvement than age at time of flight.
The changes in cognitive and physical f unctioning that occur with aging are well
documented.51 Conditions typically asso ciated with age-related performance decrements,
such as visual impairment or decreased mobility, are genera lly considered to begin about
age 60, and an age-related increase in driv ing risk has been identified after age 75.52
However, this study identified a significantly lower risk for pilots who began flying at age
25 or younger. Therefore, the Safety Boar d concludes that th e observed connection
between age and accident risk in this study is not likely due to physi cal aging issues, but to
other factors associated with the age at which a person starts flight training.
47 For a discussion of the specific methodological issu es facing studies of age- related aviation risk, see
the following: (a) D. Broach, Methodological Issues in the Study of Airplane Accident Rates by Pilot Age:
Effects of Accident and Pilot Inclusion Criteria and Analytic Strategy , DOT/FAA/AM-04/8 (Washington,
DC: FAA Office of Aerospace Medicine, 2004); and (b) G. Li, Pilot-Related Factors in Aircraft Crashes: A
Review of Epidemiologic Studies, Aviation, Space, and Environmental Medicine , 65 (1994): 979-85.
48 D. Broach, Methodological Issues in the Study of Airplane Accident Rates by Pilot Age: Effects of
Accident and Pilot Inclusion Criteria and Analytic Strategy , DOT/FAA/AM-04/8 (Washington, DC. FAA
Office of Aerospace Medicine, 2004).
49 E.J. Kay, D.J. Hillman, D.T. Hyland, R.S. V oros, R.M. Harris, and J.D. Deimler, Age 60 Rule
Research, Part III: Consolidated Data Base Experiments Final Report, DOT/FAA/AM-94/22 (Washington,
DC: FAA Office of Aerospace Medicine, 1994).
50 G. Li, S.P. Baker, J.G. Grabowski, Y . Qiang, M.L. McCarthy, and G.W. Rebok, Age, Flight
Experience, and Risk of Crash Involvement in a Cohort of Professional Pilots, American Journal of
Epidemiology, 157 (2003): 874-880.
51 For reviews of aging literature, see the following: (a) A.D. Fisk and W.A. Rogers (eds.), Handbook
of Human Factors and the Older Adult (San Diego, CA: Academic Pres s: 1997); (b) D.J. Hardy and R.
Parasuraman, Cognition and Flight Performance in Older Pilots, Journal of Experimental Psychology:
Applied 3(4) (1997): 313-348; or (c) T.A. Salthouse, Adult Cognition: An Experimental Psychology of
Human Aging (New York: Springer-Verlag, 1982).
52 G.H. Li, E.R. Braver, and L.H. Chen, Fra gility Versus Excessive Crash Involvement as
Determinants of High Death Rates per Vehi cle-Mile of Travel Among Older Drivers, Accident Analysis
and Prevention, 35(2) (2003): 227-235.
Chapter 6 37 Safety Study
Although there may be several reasons for th is finding, one of the likely differences
between pilots who begin flying at different ages is their mo tivation for learning to fly and
what they plan to do with their flying privileg es. These factors can affect initial choices
about the type of flight training pilots pursu e and also have implic ations for the type of
flying environment and oversight they will encounter over the long-term.
Many pilots who invest the time and money to learn to fly during their late teens or
early twenties may do so with the intention of pursuing a car eer in aviation. In contrast,
pilots who start flying in thei r thirties or later may be mo re likely to pursue flying for
pleasure or personal transportatio n rather than as a potential career path. Support for this
suggestion comes from this st udys findings that nonaccident pilots had higher levels of
certification, were more likely to be c onducting paid flight operations, and were more
likely to be flying an airplane belonging to someone else rather than their own airplane.
The differences between pilots pursuing a career in aviation a nd those who fly for
recreation or personal travel extend beyond fl ight hours and the equipment they operate,
and those differences may explain the study findings.
Flight Training Differences
Many persons who start fl ight training with the intent of becoming paid
professional pilots engage in full-time flight training that typically result s in a regular
schedule for practicing and testing knowledge and skills , regular oversight, and an
immersion in the aviation envi ronment. A typical professiona l pilot curriculum culminates
with a commercial pilot certif icate, multiengine rating, and either flight instructor or
instrument flight instructor certificate. Fo r these pilots, milest ones like the private
certificate or instrume nt rating are steps leading to th e higher levels of experience and
certification necessary for employment. Pilo ts who go on to find em ployment in aviation
are subject to additional scrutiny and require ments from third parties such as their
employers, the aircraft owners, their customers, and passengers.
Conversely, persons not traini ng for a career in aviation ma y be more likely to train
part-time with instructors at local airport FBOs or flight schools, and to have longer
intervals between training sessions. Persons who pursue flying for re creation or personal
travel may view the private pi lot certificate or instrument rating as a finalnot a first
step in flight training. Pilo ts who do not pursue higher levels of certificati on are expected
to maintain and improve th eir skills and knowledge on th eir own through regular flight
activity. Unlike the direct and indirect oversight of most paid flight operations, pilots
engaging in personal or business flight operations are required to fly with instructors again
only to satisfy the fl ight review require ment of 14 CFR 61.56.53 Consequently, the safety
of personal flight operations may be more de pendent on the skill, ability, and judgment of
individual pilots.
53 Although nonregulatory, many aviation insurance co mpanies stipulate additional recurrent training in
order to maintain policies and/or favorable rates.
Chapter 6 38 Safety Study
Periodic training is an impor tant part of maintaining and increasing knowledge and
skills, and this seems to be particularly tr ue for weather-related information and flight
operations. For example, a su rvey study of the general we ather knowledge of GA pilots54
found no differences related to total hours of experience or experience during the previous
6 months after correcting for the highest level of pilot tr aining/certification. These find-
ings prompted the study author to conclude the following:
It appears that pilots generally requir e formal training to obtain weather
knowledge and cannot be expected to acqu ire it on their own as they simply gain
more flight experience.
To obtain any pilot certificat e, applicants are required to demonstrate aeronautical
knowledge and skills related to identifying ha zardous weather, obtaining and interpreting
weather information, and perf orming associated decision-ma king tasks. After initial
certification, the only specific weather-related requireme nt applies to instrument-rated
pilots, who must maintain a minimum level of flight activity in or der to exercise the
privileges of that rating. The results from this study and pr evious research suggest that
flight activity alone may not be sufficient to enable pilots to maintain or improve their
ability to avoid hazardous weather conditions . The Safety Board c oncludes that periodic
training and evaluation may be necessary to en sure that pilots maintain weather-related
knowledge and skills.
The Safety Board therefore recommends th at the FAA add a specific requirement
for all pilots who do not receiv e weather-related recurrent trai ning, that the biennial flight
review include the following: recognition of criti cal weather situations from the ground
and in flight, procurement and use of aeronautical we ather reports and forecasts,
determination of fuel requireme nts, and planning for alternat ives if the intended flight
cannot be completed or delays are encountered.
In addition, pilots who were require d to demonstrate a minimum level of
proficiency in flight by reference to aircra ft instruments for certification should maintain
that minimum proficiency. Th erefore, the Safety Board al so recommends that the FAA
should, for pilots holding a private, commercial , or airline transport pilot certificate in the
airplane category who do not receive recurrent instrume nt training, add a specific
requirement that the biennial flight review include a de monstration of control and
maneuvering of an airplane so lely by reference to instrument s, including straight and level
flight, constant airspeed climbs and des cents, turns to a head ing, and recovery from
unusual flight attitudes.
54 B. Burian, General Aviation Pilot Weather Knowledge and Trainin g, final report of the FAA, grant
#00-G-020 (2002).
Chapter 6 39 Safety Study
Testing, Accident, and Incident History
A history of mishaps and/or violations has been shown to be an indicator of an
individuals future accide nt risk in both driving55 and aviation,56 and is a key component for
the actuarial calculations used to set ve hicle insurance rates and demerit-based law
enforcement programs. Less is known about the relati onship between airman test
performance and aviation accident risk; however, a 2004 Safety Bo ard accident investigation
report suggested that multiple test failures are an indicator of poor pilot proficiency.57
Airman Knowledge and Practical Test Requirements
To obtain pilot certificat ions and ratings, airmen must pass knowledge and
practical tests. Knowledge te sts are designed to assess an applicants understanding of the
information that is necessary to exercise the privileges of a particul ar certificate or rating.
Using a computerized test ing system, the FAA admini sters over 100,000 knowledge tests
each year at hundreds of testing centers. Three knowle dge teststhe private pilot
airplane, commercial airplane, and airplane instrument ratingaccount for almost half of
all knowledge tests administered. Once airm en have passed a wr itten test, received
appropriate flight training, and obtained an in structors endorsement, they are eligible to
take the corresponding practical test. Practical tests, typically administered by FAA-
designated pilot examiners, ar e designed to evaluate pilots ability to apply their
knowledge and skills in the actual flying environment.
As an example of FAA airman certification requi rements, pilots who wish to add an
instrument rating to their private or commerci al certificates must pass both the instrument
rating airplane knowledge test and the practi cal test. In 2004, th e FAA administered 13,794
knowledge tests for the instrument airplane ra ting, and 93 percent of these tests received a
passing score.58 The test has 60 mult iple-choice questions59 representing multiple
knowledge areas, such as general aeronauti cal information, navigation using instruments,
and aviation weather. The applicant must r eceive a score of 70 perc ent or higher to pass.
55 L.S. Robertson and S.P. Baker, Prior Violation Records of 1,447 Drivers Involved in Fatal Crashes,
Accident Analysis and Prevention, 7 (1975), 121-128.
56 (a) D.R. Hunter, Retrospective and Prospective Validity of Aircraft Accident Risk Indicators,
Human Factors, 43(4) (2001): 509-518; (b) G. Li and S.P. Baker, Crash and Violation Experience of Pilots
Involved in Prior Commuter and Air Taxi Crashes: A Historical Cohort Study, Aviation, Space, and
Environmental Medicine, 66, 1131-35; and (c) G. Li and S.P. Baker, Prior Crash and Violation Records of
Pilots in Commuter and Air Taxi Crashes: A Case-Control Study, Aviation, Space, and Environmental
Medicine, 65 (1994), 979-85.
57 National Transportation Safety Board, In-Flight Engine Failure and Subsequent Ditching, Air
Sunshine, Inc., Flight 527, Cessna 402C, N314AB, Abou t 7.35 Nautical Miles We st-Northwest of Treasure
Cay Airport, Treasure Cay, Great Abaco Island, Bahamas, July 13, 2003, Aircraft Accident Report
NTSB/AAR-04/03 (Washington, DC: NTSB, 2004).
58 See <http://av-info.faa.gov/data/teststat/04volume.htm >. Pass rate data were not available for the
private pilot instrume nt practical test.
59 The 60 questions are selected from a test bank of over 700 questions.
Chapter 6 40 Safety Study
FAA practical test standards for obtaining an airplane in strument rating require that
pilot applicants be assessed on eight areas of operation that include topics such as
preflight procedures, flight by reference to instruments, a nd emergency operations. Within
those areas of operation, pilots must demonstrate proficie ncy by performing specific tasks,
including an instrument cockpit check, basi c instrument flight maneuvers, several
instrument approach procedures, and emergenc y operations. In addi tion to demonstrating
competency in each required task, applicants must successfully res pond to an oral test,
show overall proficiency, and exhibit good use of aeronautical decision-making.
Applicants who fail the know ledge or practical test receive a not ice of disapproval
that specifies the knowledge ar eas with incorrect answers or areas of operation in which
tasks were not performed sati sfactorily. According to 14 CFR 61.49, before re-taking a
failed knowledge or practical test, applicants must re ceive additional training and
endorsement from an authorized instructor stat ing that they are proficient to pass the test.
Test Performance and Accident Risk
In this study, pilots were grouped into high- and lo w-pass-rate groups depending
on their past performance on kn owledge and practical tests. Overall, higher percentages of
accident pilots were represented in the lo w-pass-rate group for both the knowledge and the
practical test. Results of the chi-square analyses indicated a significant difference
between the accident and nonaccident groups, suggesting a relationship between test
performance and subsequent accident invol vement. Therefore, the Safety Board
concludes that knowledge and pr actical test fa ilures are both associat ed with a higher risk
of a pilot being involved in a weather-related GA accident.
The Safety Board also notes that, unlike the practical test standards in which failure
of one area of operation is grounds for failure of the entire test, no minimum number of
questions must be answered correctly within a given knowledge area on the knowledge
test. For example, an averag e of 12 out of 60 questions on the private pilot certification
knowledge test are weather-related.60 A pilot could answer all 12 questions incorrectly and
still receive a score as high as 80 percent, which is well a bove the minimum passing score of
70 percent. The Safety Board concludes that a pilot can incorrectly answer all questions
relating to weather on an airman knowledge test and still receive a pass ing score on the test.
The Safety Board believes that a basi c understanding of aviation weather is an
important prerequisite to obtaini ng any pilot certificate or rating. Therefore, the Safety
Board recommends that the FAA establis h a minimum number of weather-related
questions that must be answer ed correctly in order to pa ss FAA airman knowledge tests.
The establishment of such requirements will further ensure that pilots who pass a
knowledge test will have de monstrated a basic understa nding of aviation weather.
For pilots who repeatedly fail knowledge or pr actical tests, fa ilure limits may be
appropriate. The Safety Boar d suggested the potential need for such a limit during the
60 According to the FAA Airman Testing Standards Br anch, there are 10 versio ns of the private pilot
airplane knowledge test, and the number of questions pertaining to weather range from 11 to 13.
Chapter 6 41 Safety Study
investigation of a fatal acci dent on July 13, 2003, involvin g a Cessna 402C operated as an
Air Sunshine commuter flight that ditched in the Atlant ic Ocean following an in-flight
engine failure.61
A review of FAA records revealed that over a 15-year period, the pilot involved in
that accident had failed nine pr actical tests. Specifically, the pilot had failed two practical
tests for his private pilot certi ficate, one practical test for his instrument pilot rating, one
practical test for his airline transport pilot certificate, tw o practical test s for his flight
instructor certificate, and three practical te sts for his instrument instructor certificate.62
The Safety Board concluded th at the pilot had a history of below-average proficiency
before the accident flight, in cluding numerous failed FAA fl ight tests, which contributed
to his inability to maintain maximum fli ght performance and r each land after the right
engine failed.
As a result of the Air S unshine accident, the Safety Board recommended that the
FAA:
Conduct a study to determine whether the nu mber of flight checks a pilot can fail
should be limited and whether the existin g system of providing additional training
after a notice of disapproval is adequate for pilots who have failed multiple flight
checks. On the basis of the findings of the study, establish a flight check failure
limit and modify the recheck training requirements, if necessary. (A-05-02)63
The Air Sunshine acciden t is an example in wh ich the pilots poor test
performance may have indicated an underlying sk ill or knowledge deficit. If failure limits
or other measures had been in place when the Air Sunshine pilot undertook the
certification process, he might have been identified fo r remedial training or prevented
from flying.
Accidents and Incidents
During the same timeframe that data we re being collected for the 72 accidents
included in this study, the Safety Board investigated a total of 1,129 GA accidents,
involving 1,144 pilots. In the 10 years from 1993 through 2002, an average of about 1 in
every 330 active pilots in the United States was involved in a GA accident annually. A far
greater number of pilots are involved in incidents each year that have less severe outcomes
but represent similar la pses in safety. Typically, pilots involved in accidents and incidents
are not seriously injured, and most continue to fly after being involved in these events.
61 NTSB/AAR-04/03.
62 Cumulative pass rate, 36 percent.
63 Recommendation A-05-02 was issued on January 27, 2005, and its status is OpenAwait
Response.
Chapter 6 42 Safety Study
Pilot History of Accidents and Incidents
This study used FAA records to determ ine whether pilots with a history of
accidents or incidents ar e at an increased risk of future weather-related accidents. Analysis
of the accident and incident records for all study pilots s howed significant differences
between the accident and incide nt histories for the accident and nonaccident pilots. Not
only were accident pilots si gnificantly more likely to have had prior accidents and
incidents, several accident pilo ts had been involved in more than one accident or incident
before the study. Results of the logistic regression indicated that pilots involved in prior
accidents or incidents were about 3 times more likely to be involved in a weather-related
accident than pilots with no re cord of accidents or incidents. This finding is particularly
interesting because the increased risk was not limited to prior accidents and incidents
involving weather. Therefore, the Safety Boar d concludes that a hi story of accident or
incident involvement is associated with a higher risk of being involved in a future
weather-related GA accident.
Tracking a Pilots Testing a nd Accident/Incident History
The results of this study demonstrate th at both poor test pe rformance and prior
accident/incident involvement are linked to future accident involve ment among GA pilots.
Prior accident and incident involvement ma y indicate a pattern of risk and operating
performance, whereas knowledge and practi cal testing performance may reflect an
airmans overall trai ning, knowledge, skills, and proficie ncy. These findings suggest a
possibility for reducing accidents by identifying pi lots at increased risk for weather-related
accident involvement.
Currently, no formal requirements exis t for tracking and reviewing GA pilot
performance histories. However, a 1996 law esta blished such a system for air carrier pilots
after a series of Safety Board recommendations,64 and ultimately the action of United
States Congress, led to the Pilot Records Improvement Act (PRIA) .65 The intent of the
PRIA was to make operators aware of the performance hist ory of potential employees so
that they could make appropria te hiring decisions about appl icants who have exhibited a
pattern of performance problems or regulatory violations. Pilots with a history of testing
failures or violations are not restricted from engaging in flight operations by the PRIA, but
they may be subject to more scrutiny by poten tial air carrier empl oyers than applicants
with better performance histories.
Once pilots are hired, the PRIA prec ludes employers from using pilots
preemployment records. However, as a means to identify pi lots who may be at risk of
accident involvement, certai n companies have established oversight programs that allow
them to identify and track pilots who have demonstrated perfor mance deficiencies or
failures in the training envir onment. The Safety Board highl ighted the importance of such
programs during its investig ation of the December 18, 2003, accident involving Federal
64 A-88-141 (Superseded), A-88 -145 (ClosedAcceptable), A- 89-004 (ClosedAcceptable),
A-90-141 (ClosedUunacceptable), A-90-144 (Close dAcceptable), A-93-014 (ClosedUnacceptable).
65 Section 502 of Public Law 104-26.
Chapter 6 43 Safety Study
Express Flight 647, 66 and recommended that the FAA re quire all Part 121 air carrier
operators to establish similar programs. 67
Although there are currently no formal effo rts to identify and track pilots with
patterns of performance that indicate they are at increased risk of weather-related GA
accidents, the results of this study suggest that it would be possible to use existing records
to develop such a progr am. Because of the associations between testing performance, past
accident and incident involve ment, and future accidents, the Safety Board recommends
that the FAA develop a means to identify pilots whose overall performance history
indicates that they are at fu ture risk of accident involve ment, and develop a program to
reduce risk for those pilots.
Pilot Weather Briefing Sources and Methods
In addition to the statistical analyses condu cted as part of this study, the numerous
nonaccident pilot interv iews provided an opportunity to identify other weather-related
safety issues. One of those is sues was the use of preflight weather information. A common
assumption is that pilots involved in weat her-related accidents did not obtain adequate
preflight weather forecast info rmation for their route of fli ght and were therefore unaware
of the weather conditions they encountered. In addition to at tempting to determine if the
pilot had obtained weather forecast inform ation before the fli ght, study managers also
tried to discover what sources the pilots used to obtain th is information and the number of
sources they consulted before th eir flights. The intent of co llecting this information was to
determine if the weather brie fing activities of accident a nd nonaccident pilots varied
systematically and if certain activities were more predic tive of accident involvement.
Information about accident pilots use of preflight weather was obtained by ASIs
as part of their normal invest igations. When pilots were fatally injured, investigators
relied on documented briefings to determ ine if the pilots had obtained weather
information before their flights. Although th e FAA maintains records of FSS transactions,
pilots use of other weather information provided via televi sion, Internet, or satellite
sources is not typically doc umented. In 66 percent of the 72 study accidents, ASIs
reported that the pilot had rece ived preflight weather informat ion. However, it is likely
that some of the remaining 34 percent of acc ident pilots sought we ather information that
could not be identified after their accidents.
Weather briefing information for nonaccid ent pilots was gathered during their
post-flight interviews, usually within 72 hours of a given flight, increasing the likelihood
that the pilots memories of th eir preflight activiti es were accurate. According to their
reports, 94 percent of nonaccident pilots ha d checked at least one source for weather
66 National Transportation Safety Board, Hard Landing, Gear Collapse, Federal Express Flight 647,
Boeing MD-10-10F , N364F A, Memphi s, Tennessee, December 18, 2003, NTSB/AAR-05/01 (Washington,
DC: NTSB, 2005).
67 Recommendation A-05-014 was issued on May 31, 2005, and its status is OpenAwait Response.
Chapter 6 44 Safety Study
information before their flights, and 57 pe rcent reported checking two or more sources.
FSS was the most widely reported source of weather information, used by 58 percent of
nonaccident pilots. Other sour ces, by order of use, were co mmercial satellite services
such as WSI or Meteorologix68 (28 percent), DUATS (25 percent), various Internet
websites (21 percent), automa ted services like ATIS, ASOS , or AWOS (14 percent), and
television (11 percent).
These data are consistent wi th survey research suggesti ng that pilots use a variety
of sources and products to gather info rmation about weather before a flight.69 O n t h e
surface, the nonaccident pilots may appear to be more likely to gather preflight weather
information. However, for the nonaccident pi lots interviewed, 42 pe rcent used only those
sources that may not have left any record in the case of an accident. This proportion is
even higher than the 34 percent of accide nt pilots for whom no weather briefing
information was found. Knowi ng this, study managers concl uded that current methods do
not provide a means for determ ining if accident pilots ar e more or less likely than
nonaccident pilots to seek weather informati on before a flight. Furthermore, because of
the adverse weather conditions that were comm on to all flights in this study, both accident
and nonaccident pilots may have been more motivated to check weather and to check
multiple sources.
A majority of accident a nd nonaccident pilots in th is study had contacted FSS
before their flights, but mo st nonaccident pilots also re ported checking other sources of
weather information to supplement their FSS briefings. For exam ple, many nonaccident
pilots reported that they rout inely use Internet and satellit e services to obtain graphical
depictions of current and forecast conditions . This information was consistent with
previous surveys of GA pilots, which indicated that the pilots rely on a vari ety of weather
sources including sources that are not part of a standard FSS briefing.
Pilots may choose to supplement the standard FSS briefing with weather
information from alternative sources for a number of reasons. With the proliferation of
websites and electronic sources that provide weather inform ation, pilots can now easily
access detailed weather images that can be printed for use in the cockpit; they can also
access interactive tools that pr ovide a more detailed repres entation of in-flight weather
conditions than would be available through or al or textual briefings. The Safety Board
concludes that GA pilots rout inely consult alternative s ources of aviation weather to
obtain information that is not currently av ailable from a standard weather briefing.
Part 91 regulations do not sp ecifically require the use of any particular sources of
weather information for GA pilots, but do requi re that all pilots familiarize themselves
with weather and weather for ecast information before beginni ng a flight. The instructions
given to pilots about how to obtain a good preflight weather brie fing in FAA Advisory
68 WSI and Meteorologix are examples of commercial providers of weather products. Many airport
FBOs provide GA pilots with data connections to such services.
69 D. B. Beringer and R. Schvaneveldt, Priorities of Weather Information in Various Phases of Flight,
Proceedings of the Human Factors and Ergonomics Society 46th Annual Meeting (2002).
Chapter 6 45 Safety Study
Circular AC00-45E, Aviation Weather Service, are limited to FSS briefings and DUATS
and do not provide guidance on how to evalua te or select other sources of weather
information, and the FAA Aeronautical Info rmation Manual identifi es FSS and DUATS as
primary weather sour ces. Because FSS and DUATS ar e currently the only sources of
weather information endorsed by the FAA for use by GA pilots, it is important to ensure
that the information a nd services provided by t hose sources meet the needs of the GA pilot
community.
In February 2005, the FAA announced that it had selected a co ntractor to operate
the agencys FSS system, which comprises 58 stations throughout the United States. At
the time of this report, the c ontractor is expected to take over the FSS operations in late
2005. The Safety Board believes that the transition to new management for the FSS
system is an opportune time for the FAA and its contractor to eval uate FSS methods for
providing weather information. For example, the FAA should seek to understand why
some pilots choose other sources over FSS br iefings and to consider whether the FSS
system would benefit from an updating and/or broadening of the serv ices and products it
provides. A better understandi ng of pilots weather needs would allow the FAA and its
contractor to improve the weat her services and products they provide to pilots. The Safety
Board therefore recommends that the FAA determine optimal info rmation presentation
methods and delivery syst ems for FSS weather informat ion briefings, including the
possibility of supplementing or replacing some port ions of the current standard weather
briefing with graphical data.
The results of study interviews suggest that many pilots use other sources to obtain
weather data not included in a standard brie fing and then contact FSS or DUATS to fulfill
a perceived regulatory obl igation. This creates the potentia l for pilot misinterpretation or
confusion if weather inform ation gathered from various sources appears to be more
detailed than the FSS information. In some cases, the FAA and NW S contribute to this
potential confusion by providi ng detailed graphical weathe r products with disclaimers
indicating that the products are not suitable to meet the briefing requirement.70
Without specific guidance, some pilots may hesitate to c onsult electronic data
sources or may use sources that are not adequate to meet the intent of 14 CFR Part 91.103.
Therefore, the Safety Board recommends that the FAA revise guidance materials
associated with pilot weather briefings to incl ude guidance for pilots in the use of Internet,
satellite, and other data sources for obtaini ng weather information suitable for meeting the
intent of 14 CFR Part 91.103 a nd subsequently inform the aviation community about this
change.
70 For example, the standard briefing section of NWS/FAA site www.aviationweather.gov contains
all of the information cited in AC00-45E as constituti ng a standard briefing, as well as additional graphical
weather products, yet it includes a disclaimer stating that it should be used for advisory purposes only.
46 Safety Study
Summary
This study used the case control methodol ogy to determine factors associated with
an increased risk of being involved in a w eather-related GA accident. Within the context
of this study, the identified factors resulted in a statistically significant model of weather-
related accident risk. Although study analyses and the recommendations that resulted
from those analyseswere limited to the dom ain of weather-related GA accidents, it is
possible that study findings woul d generalize to the larger population of GA operations.
Historic accident data s uggest that reduced-visibil ity weather represents a
particularly high risk to GA operations. However, rather than being unique, weather may
simply test the limits of pilot knowledge, training, and skil l to the point that underlying
issues are identified. Additional research is necessary to determine whether the identified
safety issues extend beyond weather, but it is likely that implementation of study
recommendations will improve the safety of GA in general.
47 Safety Study
Findings
1. Pilots who start flying earlie r in life are at lower risk of being involved in a weather-
related general aviation accident than those who start flying when they are older, and
age at first certificate is a better predictor of future acci dent involvement than age at
time of flight.
2. The observed connection between age and acc ident risk in this study is not likely due
to physical aging issues, but to other factors associated wi th the age at which a person
starts flight training.
3. Periodic training and evaluation may be ne cessary to ensure that pilots maintain
weather-related knowledge and skills.
4. Knowledge and practical test failures are both associated wi th a higher risk of a pilot
being involved in a weather-re lated general aviation accident.
5. A pilot can incorrectly answer all que stions relating to weather on an airman
knowledge test and still receiv e a passing score on the test.
6. A history of accident or inci dent involvement is associated with a higher risk of being
involved in a future weather-related general aviation accident.
7. General aviation pilots rout inely consult alternative s ources of aviation weather to
obtain information that is not currently av ailable from a standa rd weather briefing.
48 Safety Study
Recommendations
As a result of this safety study, the Na tional Transportation Safety Board makes the
following recommendations to the Federal Aviation Administration:
Add a specific requirement for all pilots who do not receive weather-
related recurrent training, that the bi ennial flight re view include the
following: recognition of critical weather situati ons from the ground and in
flight, procurement and use of aeronaut ical weather reports and forecasts,
determination of fuel requirements, and planning for alternatives if the
intended flight cannot be completed or delays are encountered. (A-05-024)
For pilots holding a private, commer cial, or airline transport pilot
certificate in the airplane category who do not receive r ecurrent instrument
training, add a specific requi rement that the biennial flight review include a
demonstration of control and maneuv ering of an airplane solely by
reference to instruments, including straight and leve l flight, constant
airspeed climbs and de scents, turns to a head ing, and recovery from
unusual flight attitudes. (A-05-025)
Establish a minimum number of weathe r-related questions that must be
answered correctly in order to pass Federal Aviation Administration airman
knowledge tests. (A-05-026)Develop a means to identify pilots whose overall performance history
indicates that they are at future risk of accident involve ment, and develop a
program to reduce risk fo r those pilots. (A-05-027)
Determine optimal information presen tation methods and delivery systems
for flight service stat ion weather information briefings, including the
possibility of supplementing or repl acing some portions of the current
standard weather briefing wi th graphical data. (A-05-028)
Revise guidance materials associated with pilot weather briefings to
include guidance for pilots in the use of Internet, satellite, and other data
sources for obtaining weather informatio n suitable for meet ing the intent of
14 Code of Federal Regulations Part 91.103 and subseq uently inform the
aviation community about this change. (A-05-029)
49 Safety Study
Appendix A
Previous National Transpor tation Safety Board Safety
Recommendations
A search of the Safety Boards Recomm endations Database revealed that 82
recommendations on the safety of flight in IMC or visibi lity-related weather conditions
have been issued since 1968. Recommendations have addresse d a variety of topics, which
may be grouped into three broa d areas: 1) the collection a nd dissemination of weather
information, 2) pilot training and operations , and 3) air traffic control issues.
Collection and Dissemination of Weather Information
A focus on the collec tion, measurement, presentation, and commu nication of weather
information is apparent in the 82 recommendation s the National Transporta tion Safety Board has
made since 1968. Early reco mmendations called on the Federal Aviation Administration (FAA)
to increase the number of weathe r observation sites and to deve lop or improve measurements of
cloud heights, runway visual range (RVR), and slant visi bility range. Ot her recommendations
called for audio-recorded preflight weather information, a system for providing en route flight
advisories, and data-linked informat ion on precipitation and turbulence.
Although not all of the recommendations related to weathe r information were
implemented at the time they were issued, most of the concerns raised in these
recommendations have been addressed over the years by system-wide improvements in
the way that weather informat ion is gathered, presented, a nd disseminated. Pilots are now
able to access an array of up-to-date mete orological information from the telephone, the
Internet, and avionics that provide graphical weather information.
Pilot Training and Operations
Recommendations concerning the safety of flight in adverse we ather have targeted
both pilot training and flight operations. Initial training-re lated recommendations called
for increased emphasis on aviation meteorology in pilot training. In the mid-1970s, the
Safety Board recommended that the FAA
1. increase the emphasis of meteorology a nd its applications in pilot training,
2. specify a minimum number of instructi onal hours dedicated to meteorological
instruction,
3. require written examinati ons to assess a pilots me teorological knowledge, and
4. require pilot license applicants to de monstrate their abil ity to procure and
utilize weather information.
50 Safety Study
In response to these early recommenda tions, FAA increased its emphasis on the
use of weather products and took action to encourage comm ercial pilots to obtain
instrument ratings. However, the FAA did not set a mini mum number of instructional
hours dedicated to meteorological instruction, nor did it require any assessment of a pilots
knowledge or use of meteorologi cal products. Since that time, the FAA has modified the
curriculum for the private pilo t certificate to include info rmation on gathering and using
weather information. The examination for th e private pilot certificate also includes
questions on these topics; however, due to th e way the exam is scor ed, pilots who answer
all of these questions incorrectly still may achieve a passing score.
The curriculum for the priv ate pilot certificate also includes material on basic
control of an aircraft by reference to inst ruments, and in 1997, the FAA began requiring a
minimum 3 hours of instrument f light training as part of the experience requirements for a
private pilot certificate. This instruction is intended to train pilots to recognize, avoid,
andif necessaryescape instrument meteor ological conditions (I MC) if encountered.
Additional instrument flight tr aining is required for the commercial pilot license and for
the instrument rating that allows pilots to fly in IMC.
Air Traffic Control
Recommendations directed to air traffi c control (ATC) pers onnel have focused on
informing pilots about cha nges in weather and providing assistance to pilots when
visibility is limited. In response to Sa fety Board recommendations, FAA improved ATC
distribution of center weather advisories (CWA) and hazardous inflight weather advisory
service (HIWAS) information to pilots. The FAA also instituted procedures for tower
supervisors to communicate deteriorating weather conditi ons to pilots via approach
controllers. FAA has also de veloped training for controllers to assist VFR pilots who are
caught in IMC. Finally, FAA has emphasized that controllers should adjust traffic flow to
accommodate low altitude en route traffic when visibility is poor.
Table of Safety Recommendations
Please note that in table A1, the followi ng abbreviations are used to denote the
current status of the safety recommendations.
CAA: ClosedAcceptable Action
CAAA: ClosedAcceptable Alternate ActionCNLA: ClosedNo Longer ApplicableCR: ClosedReconsidered
CUA: ClosedUnacceptable Action
OAA: OpenAcceptable ResponseOUA: OpenUnacceptable ResponseAppendix A
51 Safety StudyTable A1. Previous NTSB Recommendations on the Safety of Flig ht in IMC or Visibility-Related Weather Conditions.
Rec.
Number Addressee Year Status Status Date Summary of Recommendation
A-02-8 FAA 2002 OAA Revise any restrictions and prohibitions that currently reference or address "night" or "nighttime" flight
operations in mountainous terrain so t hat those restrictions and prohibiti ons account for the entire period of
insufficient ambient light conditions, and ensure that it is clear to flight crews when such restrictions and
prohibitions apply.
A-01-35 FAA 2001 CR 07/16/2002 Amend FAA order 7110.65, "air traffic control," paragraph 10-2-5, "emergency situations," to include as
emergencies (1) inadvertent entry into instrument meteor ological conditions (IMC) by a visual flight rules
aircraft and (2) in-flight failure of attitude instruments needed to operate safely in IMC if the affected aircraft cannot remain in visual meteorological c onditions for the remainder of its flight.
A-01-36 FAA 2001 OAA Develop and ensure that air traffic controllers re ceive academic and simulator training that teaches
controllers to quickly recognize and aggressively respond to potential di stress and emergency situations in
which pilots may require air traffic control (ATC) as sistance, including but not lim ited to (1) recognition of
situations in which visual flight rules aircraft may be encountering in strument meteorological conditions; (2)
an understanding of common aircraft system failures t hat may require ATC assistance or special handling;
and (3) the application of specific tec hniques for assisting pilots that enc ounter such weather difficulties and
aircraft system failures. Further, this training should be ba sed on actual accidents or incidents, include a
comprehensive review of successful flight assists and the techniques used, and be reviewed annually to
ensure that the training materials remain current and effective.
A-01-58 FAA 2001 OAA In cooperation with the National Weather Service, ens ure that Center Weather Service Units are adequately
staffed at all times when any si gnificant weather is forecast.
A-01-71 NWS 2001 CAA 01/27/2003 In cooperation with the Federal Aviation Administrati on, ensure that Center Weather Service Units are
adequately staffed at all times when any significant weather is forecast.
A-01-59 FAA 2001 OAA Modify automated weather systems to accept runway visual range (RVR) data directly from RVR sensors.
A-01-56 FAA 2001 OUA Incorporate, at all air traffic control facilities, a near -real-time color weather radar display that shows detailed
precipitation intensities. This display could be in corporated by configuring ex isting and planned Terminal
Doppler Weather Radar or Weather Syst ems Processor systems with this cap ability or by procuring, within 1
year, a commercial computer weather program currently available through the Internet or existing stand-
alone computer hardware that displays the closest single-site Weather Survei llance Radar 1988 Doppler
data or regional mosaic images.
A-01-72 NWS 2001 OAA Eliminate the Automated Surface Observing Sy stem lockout feature as soon as possible.
A-97-27 FAA 1997 CAAA 01/14/1999 Require, under the Standard Terminal Automation R eplacement System (STARS) Program, that minimum
safe altitude warning (MSAW) alerts on instrument flight rules (IFR) aircraft be duplicated at a position in the
operational quarters designated for super visory personnel and that the supervi sor determine the validity of
the alert and whether appropriate corrective action has been initiated or is required.
A-96-049 FAA 1996 CUA 07/08/1998 Require that hazardous in-flight weather advisory service (HIWAS) broadcasts consistently include all
pertinent info contained in weather reports & forecast s, including in-flight weather advisories, airmans
meteorological info (AIRMETs), significant meteorol ogical info (SIGMETs), & center weather advisories
(CWAs).Appendix A
52 Safety StudyA-96-050 FAA 1996 CAA 08/20/1997 Encourage principal operations inspec tors (POIs) & operators to reemphasize to pilots that hazardous in-
flight weather advisory service (HIWAS) is a source of timely weather info & should be used whenever they
are operating in or near areas of pot entially hazardous weather conditions.
A-95-41 FAA 1995 CAA 06/09/1997 Amend FAA order 7110.65, Air Traffic Control, Chapter 2, General Control, section 6, weather information,
paragraph 2-115, reporting weather conditions, to require the tower supervisor to notify tower & radar approach control facility personnel, in addition to the National Weather Service observer, of the deterioration
of prevailing visibility to le ss than 3 miles. Additional, require the cont rollers to issue the visibility value to
pilots until the info is broadcast on the ATIS & the pilots have acknowledged receipt of the information.
A-89-112 FAA 1989 CAA 05/04/1992 Provide radar control trainees in the en route option at the radar training facility an emergency situation in
which a VFR pilot is asking for assi stance, is caught in weather, but is below the minimum IFR altitude.
A-89-113 FAA 1989 CAA 05/04/1992 Provide radar control trainees in the terminal option at the radar training facility an emergency situation in
which a VFR pilot is asking for assistance, is caught in weather, but is below the minimum vectoring altitude.
A-88-20 FAA 1988 CUA 01/10/1989 Amend chapter 2, section 6, paragraph 2-101, of the air traffic control handbook, 7110.65e, to require air
traffic controllers to frequently broadcast the signifi cant hazardous weather reports that are in effect.
A-88-21 FAA 1988 CAA 06/21/1989 Amend chapter 4, section 7, paragraph 343, subparag raph (1)(h) of the airmans information manual to
include airborne weather radar as an item of equipment whose complete or partial loss of capability should
be reported to air traffic control.
A-88-22 FAA 1988 CAA 06/21/1989 Amend chapter 2, paragraph 2-7 of the air traffic co ntrol handbook, 7110.65e, to include airborne weather
radar equipment as an in-flight equipment loss or malfunction covered by this paragraph.
A-88-23 FAA 1988 CAAA 08/16/1991 Expedite the implementation of the hazardous in-flight weather advisory service program in all air route
traffic control centers within the conterminous unit ed states, before the summer convective weather season
of 1988.
A-86-68 FAA 1986 CUA 10/07/1987 Include a message on the automatic terminal informat ion service broadcast whenever weather conditions
conducive to thunderstorm or microburst development exist in the terminal area or when such actual
conditions have been observed or reported.
A-86-69 FAA 1986 CAA 07/28/1987 Amend Federal Aviation Handbook 7210.3g, Facility O peration and Administration, to require the
observation of lightning or existence of cumulonimbus and towering cumulus clouds as items to be included
on Automatic Terminal Information Service broadcas t when that information has been included in the
remarks section of official weather reports.
A-86-71 FAA 1986 CUA 10/27/1988 Develop a position in major terminal facilities, to be st affed with national weather service meteorologists or
Federal Aviation Administration personnel trained for me teorological observations, to be the focal point for
weather information coordination during periods of convective weather activi ty that adversely affects aircraft
and air traffic control system operations.
A-86-74 FAA 1986 CAA 12/16/1986 Issue a general notice to all en route and terminal fa cilities emphasizing the ph raseology requirements for
describing weather areas as stated in Feder al Aviation Administration Handbook 7110.65d.
A-86-76 FAA/NWS 1986 CUA/CAAA 10/13/1987, 2/22/1994 Develop procedures to require that center weather se rvice units are attended constantly during operation so
that information concerning hazardou s weather conditions, such as thunderstorms, wind shear, icing, and
turbulence, either occurring or expected to occu r, receives prompt, appropriate dissemination.Appendix A
53 Safety StudyA-86-77 FAA/NWS 1986 CUA 10/13/1987,
07/22/1988Develop procedures to require the center weather serv ice unit meteorologist to disseminate information on
rapidly developing hazardous weathe r conditions, such as thunderstor ms and low-altitude windshear, to
Federal Aviation Administration terminal radar appr oach control and/or tower facilities immediately upon
detection of the conditions.
A-86-78 FAA/NWS 1986 CAA 07/04/1991, 07/22/1988 Expedite the implementation of equipment to upgrade all center weather service units to the state of the
technology in data acquisition and display capability.
A-86-80 NOAA 1986 CAA 07/22/1988 Require that all offices that have a weather radar display or displays and an aviation weather warning
responsibility to airports have those airports clearly located on a usable map on each weather radar display.
A-86-81 NOAA 1986 CUA 04/15/1998 Develop definitive aviation weat her warning criteria based on radar weather echo intensities and the
proximities of radar weather echos to airport appr oach and departure corridors, and implement a means to
communicate this information immediately to Feder al Aviation Administration terminal radar approach
control and tower facilities.
A-84-111 FAA 1984 CAA 01/17/1986 Postpone nationwide implementation of the hazardous infl ight weather advisory service program at air traffic
control centers until the broadcasting procedures are improved and program information is disseminated
widely.
A-84-112 FAA 1984 CUA 01/17/1986 Designate communication frequencies within the 118-135 mhz band for each air route traffic control center
to broadcast hazardous inflight weather advisory service information.
A-84-113 FAA 1984 CUA 01/17/1986 Develop procedures similar to those currently used in terminal areas for automatic terminal information
service, for flight crews to monitor an individual facilitys hazardous inflight weather advisory service
frequency and to inform the controller/facility on init ial contact that the flight has the current HIWAS
information.
A-84-114 FAA 1984 CAA 01/17/1986 During a transition period following the further impl ementation of hazardous inflight weather advisory
service, require air traffic controlle rs to advise flight crews when critic al safety information is being made
available through HIWAS. For example, ARTCC controll ers should be required to advise flights upon initial
contact "significant weather in formation available on HIWAS."
A-83-17 FAA 1983 CAA 04/02/1985 Require that automatic terminal information service advisories be amended promptly to provide current wind
shear information and other information pertinent to hazardous meteorological conditions in the terminal
area as provided by center weather se rvice unit meteorologists, and that al l aircraft operating in the terminal
area be advised by blind broadcast when a new automatic terminal information service advisory has been
issued.
A-83-18 FAA 1983 CAAA 01/23/1996 Evaluate methods and procedures for the use of current weather information from sources such as radar,
low level wind shear alert systems, and pilot reports as criteria for delaying approach and departure
operations which would expose the flight to low altitude penetration of severe convective weather.
A-82-045 NWS 1982 CAA 10/05/1982 Establish a policy of trans mitting all nonscheduled airport terminal forecasts and amendments to the Federal
Aviation Administration weather message switching cent er in Kansas City, Missouri, for distribution on
service a circuits.
A-82-17 FAA 1982 CAA 07/15/1982 Issue an emergency airworthi ness directive specifying the installation of the dual vacuum pump accessory
kit in all Cessna model 210n aircraft equipped with dei cer boots as a requirement for flight into known
instrument meteorological conditions.Appendix A
54 Safety StudyA-82-30 FAA 1982 CUA 01/07/1983 Take action to amend 14 CFR 91.116 to provide that takeoffs cannot be initiated or an approach continued
past the final approach fix or into the final approac h segment of an instrument approach procedure unless
the latest weather report for that airport issued by t he U.S. National Weather Service, a source approved by
that service, or a source approved by the administrator, reports the visibility to be equal to or more than the
visibility minimu ms prescribed for that procedure.
A-81-162 FAA 1981 CAA 01/28/1985 Require all holders of an instrument rating and a multi engine rating to demonstrate their ability to operate a
multiengine aircraft under normal and em ergency conditions by reference to flight instruments only as a
prerequisite to exercising the privileges of an instrument rating in multiengine aircraft.
A-81-23 FAA 1981 CAA 05/24/1984 Publish procedures in air traffic control handbook 7110. 65b covering the control handling of center weather
advisories.
A-81-8 FAA 1981 CAA 01/06/1984 Develop and implement a priority message-handling proc edure to assure the immediate delivery of urgent
weather messages to all weather circuits that orig inate from the weather message switching center in
Kansas City, Missouri.
A-81-94 FAA 1981 CAA 12/10/1981 Audio-record all weather briefings provided by F SS personnel and retain such records for a reasonable
period of time.
A-81-95 FAA 1981 CAA 12/10/1981 Take steps to ensure that all FSS personnel who provi de weather briefings comply with the weather briefing
procedures published in fli ght services handbook 7110.10.
A-80-115 FAA 1980 CAA 06/19/1985 Expedite the delivery of NWS weather radar color remote displays to all air route traffic control centers
center weather service units.
A-80-117 FAA 1980 CAA 10/15/1984 Expedite the development of appropriate graphic mappi ng techniques for correlation of the NWS weather
radar color remote display and the air traffic controllers radar display presentation.
A-80-118 FAA 1980 CAA 07/02/1993 Expedite the development of an integrated weather radar/ air traffic control radar single video display system
capable of providing multiple weather echo intensity discrimination without derogation of air traffic control
radar intelligence.
A-80-119 FAA 1980 CAA 06/19/1985 Require air route traffic control centers to make maxi mum use of the existing national weather service radar
sites as inputs to the color remote displays at their facilities.
A-80-135 FAA 1980 CAA 05/24/1982 Require that flow controllers and supervisory personnel assess the potential effects of hazardous weather
on low-altitude en route traffic and use the evaluat ion to adjust air traffic flow as necessary.
A-80-136 FAA 1980 CAA 06/08/1982 Require that the effect of precipitation-induc ed attenuation on x-band airborne weather radar be
incorporated into airline training programs and t hat airborne weather radar manufacturers include
attenuation data in radar operators handbooks.
A-80-109 FAA 1980 CUA 07/23/1986 Require that the effect of precipitation-induc ed attenuation on x-band airborne weather radar be
incorporated into airline training programs and t hat airborne weather radar manufacturers include
attenuation data in radar operators handbooks.Appendix A
55 Safety StudyA-79-055 FAA 1979 CAA 10/29/1979 Revise air traffic control handbook 7110.65a so that a VFR aircraft issued an altitude assignment or
instruction is provided terrain protection comparable to th at received by an IFR aircra ft. However, sufficient
latitude should be provided in the handbook so that the controller may approve a request of a pilot who
wishes to exercise the provisions of the exceptions to 14 CFR 91.79.
A-78-34 FAA 1978 CUA 10/26/1988 Install an alerting feature on all existing and new equipm ent for disseminating essential weather information
in all air traffic control facilities, at positions that require timely inform ation and at positions that are required
to issue current weather information as a pa rt of their air traffic control functions.
A-77-63 FAA 1977 CAA 06/18/1990 Expedite the development and implementation of an aviation weather subsystem for both en route and
terminal area environments, which is capable of providing a real-time display of either precipitation or
turbulence, or both and which includes a multiple-intens ity classification scheme. Transmit this information
to pilots either via the controller as a sa fety advisory or via an electronic data link.
A-77-65 FAA 1977 CAAA 08/15/1978 Transmit SIGMETs more frequently on NAVAIDs so that pilots can receive more timely information about
hazardous weather.
A-77-66 FAA 1977 CAA 08/15/1978 Code, according to geographic applicability, seve re thunderstorm watch bulletins and tornado watch
bulletins issued by the nati onal severe storms forecast center so that they may be transmitted to appropriate
air traffic control faci lities by the FAA weather message switching c enter; thus, air traffic control facilities can
relay the earliest warning of severe weather to flight crews.
A-76-123 FAA 1976 CR 05/22/1980 Institute procedures which require air tr affic controllers to release an aircra ft from all airspeed restrictions at
least 3 to 4 miles outside of the outer marker on all ILS approaches when the reported weather is below
basic VFR minimums.
A-76-29 FAA 1976 CAAA 04/07/1977 Amend that portion of 14 CFR 91.33 applicable to instrument flight rules to require a source of energy for the
rate-of-turn indicator separate from that used to power the bank and pitch indicator.
A-75-70 FAA 1975 CAA 12/22/1978 Change the wording of the restriction on the national oc ean survey instrument approach charts for locations
where night approaches are not authorized so t hat the restriction is clearly understood.
A-75-71 FAA 1975 CAAA 01/27/1977 Advise pilots arriving in a terminal area on an IF R flight plan whenever the published instrument approach
procedure is not authorized for night operations. Inst ead of an approach clearance, issue a clearance to
cruise at the appropriate minimum en route altitude/ minimum obstruction clearance altitude (MEA/MOCA) at
night.
A-74-19 FAA 1974 CAA 09/14/1978 Issue an advisory ci rcular which describes the RVR equipment and emphasizes that the RVR value is a
sampling of a small segment of the atmosphere, usually near the touchdown point. it should also be
emphasized that RVR value does not nec essarily represent actual runway visibility conditions near the
touchdown point and includes a signific ant time delay before reaching the cr ew. This information should also
be placed in the airmans information manual.
A-74-127 FAA 1974 CR 10/01/1975 Amend 14 CFR 91.105 to require the same weather mini mums outside controlled airspace as are required
within controlled airspace.
A-74-14 FAA 1974 CAA 05/24/1984 Implement, in cooperation with the national weather service, a system to relay severe thunderstorm and
tornado warning bulletins expediti ously to inbound and outbound flights when such bulletins include the
terminal area.Appendix A
56 Safety StudyA-74-30 FAA 1974 CUA 01/11/1979 Install conventional transmi ssometers or other visib ility measuring devices in the approach areas of
instrument runways.
A-74-32 FAA 1974 CUA 01/11/1979 Combine the approach zone transmissomet er or visibility readings with the ceilometer readings to produce
estimates of "thres hold contact height."
A-74-33 FAA 1974 CUA 01/11/1979 Employ runway observers to take cloud height and visi bility observations in the approach area whenever the
prevailing visibility becomes 1 mile or less at those locations hav ing minimums of less than 1 mile. These
observers should be used until the instruments descr ibed in 1. Above are installed in the approach area.
A-74-67 FAA 1974 CAA 12/03/1975 Amend 14 CFR 141 to increase the required minimum of 35 hours of classroom instru ction given to private
pilot trainees, and specify the number of hour s of meteorological instruction required.
A-74-68 FAA 1974 CUA 12/03/1975 Require that written meteorology examinations be designed to measure an applicants knowledge of the
practical application in addition to technical aspects of meteorology.
A-74-69 FAA 1974 CAAA 03/10/1977 Amend 14 CFR 61.57(b) to require a demonstration of the applicants competence to procure and utilize
weather information which will enable him to exercise safely the privileges of his pilots certificate.
A-74-70 FAA 1974 CR 10/01/1975 Amend 14 CFR 61.125 aeronautical knowledge (a) airplanes, to require an applicant for a commercial pilot
certificate to present evidence of meteorological know ledge in addition to the other areas of aeronautical
knowledge now specified, similar to the requirements of 14 CFR 61.125 (b)(2), (c)(3), (d)(5), or (e)(3).
A-74-71 FAA 1974 CAA 06/04/1975 Increase the emphasis on aviation meteorology and weather limitations of pilots through its general aviation
accident prevention program.
A-74-72 FAA 1974 CAA 06/04/1975 Take priority action in order to adhere to the propos ed 4-year implementation plan for the en route flight
advisory service (Flightwatch) program.
A-74-73 FAA 1974 CAA 12/03/1975 Implement, at least on an experimental basis at sele cted high general aviation ac tivity locations, the audio
recording of preflight weather briefings.
A-74-74 NOAA 1974 CAA 12/03/1975 Accelerate efforts to update, publish, advertise, and di sseminate the document entitled Aviation Weather for
Pilots and Operations Personnel.
A-74-75 NOAA 1974 CAA 02/20/1975 Accelerate the expansion of the evaluation staf f to its proposed complement of one evaluations
meteorologist per state and include in his responsibilities the implementati on of a quality control program for
aviation weather observations.
A-74-76 NOAA 1974 CNLA 02/20/1975 Accelerate efforts to improve the presentation of aviation weather products.
A-72-105 FAA 1972 CAA 01/01/1975 Visibility and separation from cl oud distances should be assessed cons ervatively in VFR operations, and
that VFR flight should be continued only when visibility is unquestionable.
A-72-139 FAA 1972 CUA 12/03/1975 Combined efforts and resources of government and indu stry be applied toward the expeditious application
and use of technological advances in the field of all weather flight navigation and approach/landing systems.
Although it cannot be stated unequivocally that this ac cident would not have occurred if an instrument
landing system or more advanced landing system had been in use, it is the boards belief that the probability
of the occurrence of such an accident would have been greatly reduced.Appendix A
57 Safety StudyA-72-198 FAA 1972 CAA 11/30/1973 The FAA ensure widespread dissemination of information to pilots in all segments of aviation regarding the
potential hazards associated with weather conditions characterized by a partial obscuration of the sky
caused by a shallow layer of dense fog.
A-72-43 FAA 1972 CUA 06/08/1972 Issue an advisory circ ular incorporating excerpts of this r eport, including the findings, stressing to all
instrument and airline transport pilots the need for c ontinuous surveillance of flight instruments when
operating in instrument meteorological conditions.
A-70-63 10 airlines
and private
flying organizatio
ns1970 CNLA 03/22/1971 The Boards review of these and other approach and landi ng accidents show that they occur with needless
regularity and that they are not confined to a single segment of aviation. It was emphasized that there were
a number of factors which must be considered in pr eventing this type of accident, namely, airborne and
ground equipment, procedures and piloting techniques and judgment. Safety should be in continually
improving the quality and increasing the quantity of landing aids and weather reporting facilities and
services. it was further emphasized that all pilots should be thoroughly instructed in the hazards associated
with shallow fog penetration and be prepared to make the missed approach decisi on and execute it without
delay whenever the alternative course is required. Illusions tending to disorientate the pilot during the final
phases of the approach and the limitations of various sy stems are but two of the areas bearing upon such a
decision which must be recognized.
A-69-1 FAA 1969 CAA 02/17/1969 1. Amend sections 91.117 and 121.649 of the federal aviation regulations "to prohibit any approach below
200 feet above field level unless the pilot has the runway threshold in sight and require that he continue to
have same in sight during the remainder of the approach." 2. "Bring to the attention of all instrument pilots
the hazards associated with shallo w fog penetration." 3. Include as ma ndatory items in airline training
programs and FAA-approved instrument flight school cu rriculums "information on shallow fog penetration,
the effect upon the guidance segment, and the potential illusions that can be created." 4. "Pursue as
expeditiously as possi ble" its research into instrumentation which would provide slant visual range
information. 5. set standards and specifications and encourage the development of "r ealistic" low-visibility-
approach flight simulators. 6. progra m improved approach zone lighting covering "at least the last 1,000 feet
of the approach" for "installation on a priority basis," when and if "financ ial conditions permit," at airports
prone to frequent heavy fog.
A-69-32 FAA 1969 CUA 11/12/1973 (1) That section 91.116 of the FAR be changed to agr ee with the provisions of section 121.653 and the
similar requirements of parts 123 and 135 in order that t he approach be restricted as well as the landing. (2)
That section 91.117 be amended to the effect that in no event shall descent below 200 feet be performed
unless landing minimums are present. (3 ) that while section 91.116(b) clear ly states that a landing may not
be made unless the visibility is at or above the landing minimum required, nevertheless, in the interests of
safety and in order to insure proper interpretation , all conditions requiring a missed approach should be
contained in section 91.117 (b). Accordingly, an additi onal condition should be added to section 91.117(b) to
the effect that if landing minimums cannot be maintained, a missed approach must be executed.Appendix A
58 Safety StudyA-68-06 FAA and
ESSA1968 CAA/CNLA 01/01/1975 1. Increase the number of aviation weather observing sites. 2. Vigorous program of quality control of
aviation weather observations should be developed. 3. Cloud-height measuring equipment should be
provided at all aviation weather observing stations. 4. Additional efforts should be made to standardize the
location of weather instruments at airports. 5. Me thods should be developed for measuring and forecasting
low level wind shear in the terminal area. 6. In or der to insure more accura te visibility observations,
adequate visibility reference markers (p articularly nighttime markers) sh ould be provided for the guidance of
observers. 7. Continued efforts should be made to exp and the upper-air observing network and to increase
the number of rawinsonde ascents to four per day. 8. The weather radar network should be expanded,
particularly west of approximately 100 degree west longitude, and weak, obsolete, war-surplus equipment
should be replaced with up-to-date, long-range weather r adar sets. 9. It is recognized that it is generally
impractical to base a staffing plan on the "bad weather" situation. It appears, however, that some revisions
or expansions are required, so that a continuous w eather watch could be maintained and improved pilot
briefing services provided at those locations m anned by one person during certain hours. 10. Continue the
expansion of the runway visual range (RVR) pr ogram including the multiple installations of
transmissometers. 11. A means of measuring slant vi sibility or slant visual range which a pilot would
experience on an approach to landing would certainly enhance air safety. 12. In view of the enthusiastic
support by the users of the pilot-to-forecaster experim ental programs at Kansas City and Washington, it is
suggested that serious considerati on be given to establishing an operat ional program on a national basis.
13. The transcribed weather broadcasts (TWEB) netwo rk should be expanded to provide coast-to-coast
coverage. 14. The pilots automatic telephone weather answering service (PATWAS) should be greatly
expanded to provide its service to many additional areas, particularly those areas where live weather
briefing may not now be available. 15. There is a nee d for more pilot weather briefing facilities. 16.
Substantial improvements in weather briefings could be realized by the provision of facsimile equipment for
all weather briefing facilities. This would also assist in the desired standardization of pilot weather briefing
procedures. 17. Additional efforts should be made to im prove and standardize weather briefing displays. 18.
Provision should be made for additional telephone lines to weather briefing fa cilities. 19. In order to assist
the safety board in accident investigations and for ESSA/FAA quality control purposes, audio recording of
pilot weather briefings is advocated. 20. Aviation st ands to benefit from information derived from weather
satellites. Accordingly, it is cons idered that special efforts should be made to devise refined techniques and
procedures for providing aviation-or iented weather satellite informat ion on a national basis. 21. There
continues to be a need for improved delineation of avia tion forecast area boundaries. A revision of the
present system of delineation should be considered in order to define more precisely the area boundaries
perhaps a reassignment of areas of forecast respon sibility to make the boundaries contiguous with state
boundaries. 23. We adhere to the belief that a centrali zed Clear Air Turbulence (CAT) forecasting center
should be established, similar to the Severe Local Storms (SELS) unit. Certai nly safety, efficiency, and
economy would be enhanced by such an establishment. 24. Continued efforts should be made to improve
the procedures for obtaining and disseminating in-flight weather information. 25. We are concerned with
instructions to forecasters regarding the modifiers to be used for In-flight Advisories (SIGMETS) containing
Clear Air Turbulence (CAT) forecasts. Forecasters are directed to use the phrase moderate or greater in
CAT forecasts and may only use severe or extreme in CAT reports . These instructions (in Chapter D-22
of the Weather Bureau Operations Manual) appear to be cont rary to preceding instructions (in that Manual )
which call for SIGMETS to be issued when (among ot her things) severe or extreme turbulence are
expected. Unfortunately, moderate or more turbulence includes all intensities except light. Furthermore,
it seems unfair and certainly not very helpful to the pilot not to be apprised of the forecasters thinking and intent in regard to the category of turbulence to be anticipated. 26. There has always been a requirement for more accurate aviation weather forecasts, particular ly for the terminal area, and research into improved
forecasting methods should continue to be pursued. Research should also be conducted to develop
objective methods for measuring or foreca sting the intensity of icing and turbulence.
A-68-24 FAA 1968 CAA 09/04/1969 It is recommended that the installation of adverse weather warning devices (light) be encouraged at all
airports where official weat her observations are taken.Appendix A
59 Safety Study
Appendix B
Basic VFR Weather Minimums
Airspace Flight Visibility Distance from Clouds
Class B 3 statute miles Clear of Clouds
Class C 3 statute miles 500 feet below
1,000 feet above
2,000 feet horizontal
Class D 3 statute miles 500 feet below
1,000 feet above
2,000 feet horizontal
Class E
Less than 10,000 feet MSL 3 statute miles 500 feet below1,000 feet above
2,000 feet horizontal
At or above 10,000 feet MSL 5 statute miles 1,000 feet below
1,000 feet above
1 statute mile horizontal
Class G
1,200 feet or less above t he surface (regardless of
MSL altitude)
Day, except as provided in section 91.155(b) 1 statute mile Clear of clouds
Night, except as provided in section 91.155(b) 3 statute miles 500 feet below
1,000 feet above2,000 feet horizontal
Class G
More than 1,200 feet above the surface but less than
10,000 feet MSL
Day 1 statute mile 500 feet below1,000 feet above2,000 feet horizontal
Night 3 statute miles 500 feet below
1,000 feet above
2,000 feet horizontal
Class G
More than 1,200 feet above the surface and at or
above 10,000 feet MSL 5 statute miles 1,000 feet below
1,000 feet above
1 statute mile horizontal
60 Safety Study
Appendix C
Evaluation of M atching Procedure
Interviews were conducted with 135 pilots of matching nonaccident flights. The
mean number of matches iden tified per accident was 2.2 for accidents involving IFR
flights, and 1.5 per accident for those involving VFR flights. There were no statistically
significant differences between the accident and nonacciden t groups on the variables used
for matching, and the absenc e of any differences confirmed the soundness of the
procedures used to match acc ident and nonaccident flights.
Aircraft Types
Of the study accidents, 74 pe rcent involved single-engine aircraft and 26 percent
involved multiengine aircraft. The nonaccident aircraft group included 72 percent single-
engine and 28 percent multiengine airplane s. In comparison, estimates from the 2002
General Aviation and Air Taxi Activity (GAATA ) Survey1 indicated that single-engine
airplanes accounted for 69 percent of aircraft involved in GA operations, and multiengine
airplanes accounted for 13 percent. Additi onally, 93 percent of study accidents involved
piston-powered airplanes, 6 percent invol ved turboprops, and 1 percent involved jet
aircraft. The nonaccident gro up included 90 percent piston-pow ered aircraft, 7 percent
turboprop aircraft, and 2 percent jet aircraft.2 GAATA Survey estimates indicated that
piston-powered airplanes acc ounted for 77 percent of th e GA fleet, while turboprops
accounted for 3 percent, and jets fo r 4 percent. See figures C1 and C2.
1 U.S. Department of Transportati on, Federal Aviation Administration, General Aviation and Air Taxi
Activity Survey, Calendar Year 2002 (Washington, DC: FAA, 2004), availa ble at <http://api.hq.faa.gov/pubs.asp>.
2 Values do not sum to 100 percent due to rounding.
61 Safety Study
Figure C1
Figure C2Distribution of Study Aircraft
by Number of Engines
Multi Single0%20%40%60%80%100%
Accident Nonaccident
Distribution of Study Aircraft
by Type of Engine
Piston Turbo-jet Turbo-prop0%20%40%60%80%100%
Accident NonaccidentAppendix C
62 Safety Study
Appendix D
Guided Interview
Departure Point T i m e
Destination T i m e
En route Altitude
N - n u m b e r Aircraft Manufacturer Aircraft Model
Do you rent or own other: ?
N o t e s :
Purpose of flight
Personal Travel Proficiency Flight to or from work, or work-related
Pleasure Paid to fly (Instructor, observation, delivery)
Regulation
Part 91 Part 135 Part 121
Leg of Trip
Inbound Outbound Intermediate Local
Were you familiar with the airplane? Yes No
Were you familiar with the area? Yes No
Flew in the region in the last 6 months? Yes No
What flight rules were you operating under? VFR VFR with flight following
VFR to IFR (filed en route) IFR
if VFR, did you file a flight plan or use flight following? Yes No
Was there more than 1 pilot aboard the aircraft? Yes No
N o t e s :
63 Safety Study
Weather Briefing
Did you get a briefing prior to departure? Yes No
If yes, what source? DUATS Flight Service Television
ATIS/METAR Internet Other:
Did you get more than one briefing? Yes No
Approximate time of last briefing:
Did you request weather en route? Yes No
How would you describe the weather you encountered on that flight?
Basic Weather: IFR MVFR Convective
VFR – degraded visual reference
Was it better or worse than the forecast? Better Worse Same
Did you have to deviate from your planned route or altitude? Yes No
What type of approach did you do at your destination?
Visual ILS Localizer VOR NDB GPS
Basic aircraft equipment VFR only IFR certified
Navigation Avionics VOR DME RNAV/LORAN
ADF GPS (VFR-only) GPS (IFR en route)
GPS (IFR approach) Moving Map Glideslope
Marker Beacon Other:
Comments:
Weather equipment Weather Radar Storm scope/Strike finder
Structural Anti-ice/De-ice Other:
Is your aircraft certified for known icing? Yes No
Comments:
Aircraft Control Equipment Wing-leveler Autopilot (2 axis)
Autopilot (2 axis, coupled) Altitude Hold Yaw Damper
Other:
Comments:
Additional Avionics Altitude alert Radar altimeter
Multifunction Display(s) GPWS TCAS
Other:
Comments: Appendix D
64 Safety Study
Pilot Details
Sex
Male Female
Certificates: None Student Private Commercial
ATP DATE(S):
Ratings: SEL MEL SES MES O t h e r :
Instrument Ratings: Airplane Other: DATE(S):
Instructor Rating(s)? None Airplane Multi-engine Airplane
Instrument Airplane Other:
Date of most recent flight review
Date of most recent flight instruction
Date of most recent instrument competency check (if applicable)
Date of most recent instrument flight instruction
Date of last medical exam
Medical Certificate First Class Second Class Third Class
Waivers/Limitations No Yes If yes, what?
Certificate number (this will be de-identified)
Pilot Date of Birth
Flight Activity
The rest of my questions have to do with the number of flight hours you have, and you
may not have all of the information immediately available. If not, is there a time that I can call you back?
All A/C In Type
A/CNight Actual
Instrument Simulated
Instrument
T o t a l T i m e
P I C
Instruction Received
L a s t 9 0 D a y s Appendix D
65 Safety Study
Appendix E
Supplemental Accident Form
NTSB General Aviation Weather Supplemental Form
1. Accident/Incident Information
Date Cit y State
Aircraft N-number NTSB#
Basic Weather IFR MVFR VFR – with reduced visual reference Convective
2. Purpose of Flight (Check one)
Persona l Travel Proficienc y Paid to f ly (e.g. Instructor, obser vation, de livery)
Pleasure Flight to or from work, or work-re lated Other:
Comments:
Pilot’s connection to the accident aircraft: Own (full or part) Rent Other:
3. Leg of Trip
Outbound leg Intermediate leg Inbound leg Local Flight
Comments:
4. Weather Briefing
4.1 Pi lot recei ved a weather briefing prior to departure Yes No Record
4.2 Pilot recei ved more than one weather briefing Yes No Record
4.3 Time of last weather briefing (UTC)
4.3 Pi lot requested updated weather while en route Yes No Record
4.4 Comments:
5. Pilot Experience
5.1 Date of most recent f light re view Could not be determined
5.2 Date of most recent instrument competenc y check (if app licable ) Could not be determined
5.3 Pi lot was fami liar with the area Yes No Could not be determined
5.4 Pi lot was fami liar with the accident aircraft Yes No Could not be determined
5.5 Comments:
6. Pilot Instrument Training
6.1 Tota l hours of simu lated instrument instruction recei ved Could not be determined
6.2 Total hours of actua l instrument instruction recei ved Could not be determined
6.3 Date of most recent f light instruction Could not be determined
6.4 Date of most recent instrument f light instruction Could not be determined
6.5 Comments:
7. Basic Aircraft Instruments (Check one)
VFR-on ly IFR certified Unkno wn / Destro yed
Comments:
8. Navigation Avioni cs (Check all that apply)
VOR DME RNAV/LORA N ADF
GPS (VFR-on ly) GPS (IFR en route) GPS (IFR approach) Moving Map
Glideslope Marker Beacon Unkno wn / Destro yed Other:
Comments:
9. Weather Equipment (Check all that apply)
Weather Radar Storm scope/ Strike finder Anti-ice/De-ice Equipment
Unkno wn / Destro yed Other:
Was the accident aircraft certified for kno wn icing? Yes No Unkno wn
Comments:
10. Aircraft Control Equipment
Wing-leveler Autopi lot (2 axis) Unkno wn / Destro yed
Autopi lot (2 axis, coup led) Altitude Ho ld Yaw Damper
Other:
Comments:
66 Safety Study
11. Additional Avionics
Altitude a lert Radar a ltimeter Multifunction Disp lay(s)
GPWS TCAS Unkno wn / Destro yed
Other:
Comments:
12. Did the pilot declare an emergency, communicate a problem, or request help?
Yes No
if Yes, brief ly explain
13. Additional Information (Use this space to add any addi tional information that did not fit in the above spaces)
14. Administrative information
Investigator Date
Please comp lete and return within 30 da ys of the accident. Appendix E
67 Safety Study
Appendix F
A Comparison of Forecast and Ac tual Visibility Conditions for
Study Accidents
Accidents were included in this study if they took place in IMC or marginal VMC
or if they were preceded by a loss of visual reference or en counter with weather. One step
in the analysis was to determ ine if pilots could have an ticipated the weather conditions
they experienced based on the weat her forecasts that were availa ble before their departure.
Aviation weather forecasts provide se veral types of information, including
horizontal visibility, ceiling heights, cloud coverage, pr ecipitation types, wind, and
convective activity. The decisi on that a forecast is consistent with actual conditions is
somewhat subjective, and a difference that is acceptable to one person may not be
acceptable to another. For this study, the FAAs legal definition of VFR minimums was
used to classify the forecast conditions into categories of VFR or IFR. VFR minimums
dictate the minimum ceiling and visibility c onditions for a pilot to fly without an IFR
flight plan. While the threshold between VF R and IFR varies some what depending on the
type of airspace and time of day, VFR conditions are gene rally present when the cloud
ceiling1 is 1,000 feet or greater a nd the horizontal visibility is 3 miles or greater. IFR
conditions exist when either horizontal visibili ty or cloud ceilings are less than these
minimums.
To determine the forecast conditions for each accident , Safety Board
meteorologists obtained Sierra AIRMETs that were released immediately before the
accident airplanes departur e time for the region surrounding the accident location. (The
Sierra AIRMET describes IFR c onditions and mountain obscurations.) Staff then plotted
the Sierra AIRMETS and determined whether the accident site was located within its
boundaries. When an accident site was with in the boundaries of the IFR AIRMET, the
forecast was classified as IFR; ot herwise it was classified as VFR.
The actual conditions at the time and location of the accid ent were determined
from the Safety Boards factua l report. As part of the fa ctual report, the investigator
assigns basic weather conditions as either IMC or VMC.
Table F1 provides an overview of the forecast and actual conditions for the 72
accidents in the study. In terms of forecast quality, act ual conditions ma tched forecast
conditions in 61 percent of case s. Within the 39 percent of cas es in which the forecast did
not match the actual conditions, the visibility conditi ons were better than forecast (that
is, forecast IFR, actual VFR) in 39 percent, and were wor se than forecast (that is,
forecast VFR, actual IFR) in 61 percent of cases.
1 Ceiling in this case is defined as the lowest cloud layer of broken (BKN) or greater coverage.
68 Safety Study
Table F1
Using FAA definitions as criteria for evaluating the accuracy of the forecasts
suggests that in approximately 25 percent of all accident cases, th e actual conditions were
worse than predicted. A few of the many pos sible explanations include the following:
1. Forecasts of IMC or VMC are issued from the perspectiv e of someone on the
ground. In-flight conditions also de pend on the operating altitude.
2. The weather conditions may have been isolated to an ar ea so small that an
AIRMET was not warranted. A Sierra AIRM ET is issued for ceilings less than
1,000 feet and/or for visibility less than 3 miles affecting over 50 percent of an
area at one time, or for extensive mount ain obscuration for an area of at least
3,000 square miles.
3. Using this verification methodology, a fairly small deviation between the
forecast and actual conditi ons could result in a cha nge in classification. For
example, if the ceiling was forecast to be 1,000 feet, and the actual ceiling was
900 feet, it would be clas sified as a change from VFR to IFR conditions.
The term marginal VFR or MVFR is someti mes used to refer to conditions that are
near the VFR minimums. MVFR refers to condi tions in which visibility is between 3 and
5 miles or when ceiling height is between 1,000 and 3,000 feet. At pr esent, the FAAs use
of the term marginal VFR is limited to the section of an area forecast corresponding to the
period of time more than 6 hours af ter the time of forecast issuance.
A closer look at the forecast and actual data reveals that in 4 of the 23 cases where
VMC was forecast, and in 6 of the 49 cases where IMC was fo recast, the actual conditions
fell into the definition of MV FR. In these cases, actual c onditions were closer to the
forecast than the dichotomous categori zation of IFR and VFR would suggest. Actual
VFR IFR Total
VFR 6
(8.3%)17 (23.6%)23(31.9%)
IFR 11
(15.3%)38(52.8%)49(68.1%)Forecast
Total 17
(23.6%)55
(76.4%)72 Forecast and Actual Conditions
for the Study AccidentsAppendix F
69 Safety Study
Appendix G
Study Accidents
NTSB A ccident
Number D ate Lo cationAircraft
Registr ation M ake ModelAccident
Severity
1 IAD03FA069 Aug 02, 2003 G alion, OH N5 77SK Piper P A-34-200T F atal
2 CHI03F A246A ug 05, 2003&ODUN¿HOG01 N4577T Piper P A-28-180 F atal
3 NYC03F A176 A ug 06, 2003 Ple asantville, P A N28788 Gru mma nAA-5B F atal
4 IAD03FA070 Aug 08, 2003 F actoryville, P A N6373C Piper P A-32-300 F atal
5 CHI03L A267 A ug 19, 2003+DPPRQG01 N1812S Beech BE- 76 Non-fatal
60,$)$ Aug 21, 2003 Cle arwater, FL N93DC Piper P A-31 F atal
7 CHI03F A296A ug 28, 2003*UDQG0DUDLV01 N285V Beech 58P F atal
8 CHI03F A291 Sep 01, 20038QLRQGDOH,1 N8018 J Beech B3 6TC F atal
9 CHI03L A315 Sep 22, 2003 Ch anute, K S N122CC Piper P A-24-250 Non- fatal
10 NYC03F A205 Sep 27, 2003&RQFRUG0$ N963LP Cessn a 182TF atal
11SEA04LA001 Oct 0 2, 2003%ODNHO\,VODQG:$ N2695S Cessn a 340 Non- fatal
12SEA04FA009 Oct 20, 20036HDVLGH25 N136SB Beech A36 Fatal
13ATL04F A027 Oct 26, 2003 Spartanburg, SCN 7799Y Piper P A-30 F atal
14 CHI04F A025 Nov 0 2, 2003 Hutchinson, K S N6107Z &RPPDQGHU 114TC F atal
15)7:)$ Nov 16, 2003 Tol ar, TX N 777S G Cessn a0 Fatal
16 LAX04GA051 Nov 21, 2003 Bi g Bear City, C A1 : Piper P A-28-180 F atal
17SEA04FA022 Nov 25, 2003:DUUHQ25 N10BX Beech S35 F atal
180,$)$ Nov 27, 2003 Jacksonville, FL N 698X SwearingenSA-26-ATF atal
19)7:)$ Dec 04, 2003+DUULVRQ$5 N350JL Beech S35 F atal
20 LAX04LA058 Dec 04, 2003 San Diego, CA N15C Cessn a 525 Non- fatal
21ATL04F A045 Dec 04, 20030RUODQ*$ N85BK Beech B 200 F atal
22 LAX04FA061 Dec 0 7, 2003 Chino Hills, C A N16264 Piper 3$5 Fatal
23ATL04F A049 Dec 10, 2003 Vest avia Hills, AL1: Cessn a 441 F atal
24ATL04F A051 Dec 11, 2003 Greeneville, TN N159 2T Cessn a 414 F atal
25SEA04LA026 Dec 11, 20030F0LQQYLOOH25 N29CV Upri ght 59$ Fatal
26 LAX04FA066 Dec 14, 2003 Cl aremont, CA N6887L Cessn a 421C F atal
27ANC04F A015 Dec 14, 2003 Tonop ah, NV N4 674A Cessn a P210N F atal
28 LAX04FA081 Dec 15, 2003 Coron a, CA N61303 Cessn a 150J Fatal
29)7:)$ Dec 15, 20030HQD$5 N7929D Beech H-35 F atal
30 CHI04F A044 Dec 1 7, 2003 Brooklyn, I A 1: Piper P A-32-260F atal
31 CHI04F A043 Dec 1 7, 2003 D aytona Beach, FL1%0 Piper P A46-500TP F atal
32ATL04F A056 Dec 17, 20036DQIRUG1& N9562L Cessn a2 06HF atal
33 LAX04LA074 Dec 20, 2003 Angwin, CA N20480 Beech B55 Non- fatal
34SEA04LA029 Dec 22, 20030LVVRXOD07 1$0 Beech 58P Non- fatal
35 LAX04FA077 Dec 24, 2003 Avalon, CA N3747U Piper P A-34-200T F atal
36SEA04LA030 Dec 25, 2003 Elk City, ID N13 63U Cessn a0 Non-fatal
70 Safety Study
176%$FFLGHQW
Number D ate Loc ationAircraft
5HJLVWUDWLRQ 0DNH 0RGHO$FFLGHQW
Severity
37SEA04FA032J a n 01, 2004*ODVJRZ07 N3171S Cessn a 182GF atal
38)7:)$ Jan 01, 2004 D allas, TX N4104B Bell anca 17-30A Fatal
39SEA04FA031 Jan 01, 2004&DYH-XQFWLRQ25 N53505 Piper P A-44-180 F atal
40)7:)$ Jan 02, 2004 Be aumont, TX N39 6HP Piper 3$5 Fatal
41 CHI04L A052J a n 02, 2004 Fishers, IN N 21992 Piper 3$57 Non-fatal
42 DEN04F A035 Jan 03, 2004 Cortez, CO 165 5RFNZHOO 690A Fatal
43 CHI04F A055 Jan 10, 2004%DXGHWWH01 N5787J Cessn a 182PF atal
44 LAX04LA092J a n 11, 2004 V acaville, C A N77S Cessn a 140 Non- fatal
450,$&$ Jan 14, 2004)RUW0HDGH)/ 10& Cessn a2 10L Non- fatal
46)7:)$ Jan 15, 2004:LOORZ3DUN7; N7252X Beech B3 6TC F atal
47 LAX04FA096J a n 19, 2004 Gr ass Valley, CA1 /5 Cessn a 172KF atal
480,$)$ Jan 19, 2004 Fort Pierce, FL N 298PA Piper P A-28-181 F atal
49SEA04LA038 Jan 28, 20045LJE\,' N734CB Cessn a 172N Non- fatal
500,$)$ Jan 31, 2004+RPHVWHDG)/ N75GC Beech C90 F atal
51 LAX04FA113 Jan 31, 2004 L aupahoehoe, HI N5 637C Cessn a 414A Fatal
52 DEN04F A043 Feb 0 6, 2004:DOFRWW:< N5119 2 Cessn a T206HF atal
53 CHI04L A064 Feb 0 7, 2004/RQH5RFN:, N5039 S Piper 3$5 Non-fatal
54ATL04F A075 Feb 11, 2004 Alma, GA N6473J Piper P A-28-180 F atal
55ANC04L A022 Feb 14, 2004 Crescent City , FL 1 :% Thorp T-18 F atal
56 CHI04F A071 Feb 1 6, 20045R]HO.6 15' Cessn a 182QF atal
57 CHI04F A069 Feb 1 7, 20045LFK+LOO02 1: Piper P A-28-180 F atal
58)7:)$ Feb 22, 2004 V alley Spring, TX N 6175Y Cessn a2 10N F atal
59ATL04F A077 Feb 23, 2004 Arlington, AL N9103Z Piper P A-46-310P F atal
60 LAX04FA139 Feb 27, 20043LQH0WQ&OXE&$ N7687J Piper 3$5 Fatal
61 CHI04L A085 0DU Dubuque, I A N105F S Cessn a5 Non-fatal
620,$)$ 0DU Spring Hill, FL N8148G Cessn a 182PF atal
63 LAX04FA1620DU Los Angeles, C A N1148V 0RRQH\ 0. Fatal
64 NYC04F A0920DU Harlan, KY N81 73U Piper 3$5 Fatal
65ATL04L A0870DU %D\0LQHWWH$/ 1$0 Alarus CH 2T Non- fatal
66 NYC04F A100 Apr 02, 2004 H arrietsto wn, NY N4 686J Piper 3$5 Fatal
67 LAX04FA177 A pr 04, 2004 Uki ah, CA N4130D Piper 3$57 Fatal
68 DEN04F A057A pr 06, 2004 Burlin gton, CO N35 24Y Cessn a 182FF atal
690,$/$ Apr 10, 20041/DXGHUGDOH)/ N916LJ Cirrus 65 Non-fatal
70ATL04F A093 Apr 12, 2004 North Augusta, SC10( Cessn a 182S Fatal
71SEA04LA071Apr 17, 2004 Gibbonsville, ID N5XK Keller Prospector FK1 Non- fatal
72ATL04L A097A pr 23, 20042DNODQG71 1:' Durr L ancair Legacy F atalAppendix G
Fly GACA is an independent educational platform. It is not affiliated with, endorsed by, or operated by the General Authority of Civil Aviation (GACA) or the Government of the Kingdom of Saudi Arabia. The official and authoritative source for all civil aviation regulations, publications, and aeronautical information is always GACA. Always verify against the latest official GACA publication at gaca.gov.sa.