This paper analyzes international literature on pilots’ behavior and decision-making under adverse weather conditions, particularly focusing on general aviation pilots. It highlights that cognitive biases, inexperience, and insufficient training increase the likelihood of fatal accidents when pilots fly into instrument meteorological conditions. The study suggests that improving decision-making processes through analytical methods and better training could reduce weather-related accidents.

1. The influence of adverse weather conditions on pilots’
behavior and decision-making
Berend Roosendaal
Amsterdam University of Applied Sciences, Honours Programme Engineering
Noord-Holland, Netherlands
Abstract
This paper presents a summary of the international literature regarding pilots’ behavior under
adverse weather conditions. The literature also covers the influences on the decision-making
process of the pilot. The focus on this study is on pilots who fly solely in general aviation.
During the second half of the last century, the amount of private pilots has increased. Along
with this increase is the amount of fatal weather-related accidents. Weather related accident are
most common and consistent in situations where pilots fly from visual flight rules into
instrument meteorological conditions. The objective of this paper is to summarize the current
and previous studies concerning this subject and conclude what actually causes the event.
Current technologies and needs for prevention of the problem will be disucssed. The result of
this paper illustrates the direct and indirect factors that are related to weather related accidents
and how studies have tried to counter and influence this problem. Current studies revealed that
the psychological factor of the pilot during the decision-making process is essential in this
problem. An ultimate decision-making process would be to shift from the intuitive process to a
more analytical process. This will result in broader range of information that is used to perform
an accurate weather assessment, and to make a better decision.
Keywords: weather-related decision-making, debiasing, pilot behavior, general aviation
weather-related accidents
1. Introduction
It is common knowledge that adverse weather influences the aircraft. The safety and the
effective operation of an aircraft will be endangered by flying into adverse weather conditions.
Adverse weather conditions has a significant larger impact on general aviation (GA) than on
commercial flights. This is due to the type of aircraft, number of pilots and certifications of the
pilots. The most weather related cause of GA incidents are a result from an aircraft that flies
from visual flight rules (VFR) into instrument meteorological conditions (IMC) or in other
words into instrument flight rules (IFR). In the US in 2011, 86% of VFR into IMC accidents
were fatal while other incident causal fatality rates were significantly lower. This high
percentage is primarily caused by the fact that most GA pilots have very limited training in
IMC, this limited training focuses on maintaining a minimum level of control of the aircraft,
and being able to return the aircraft to a straight and level condition. When flying into IMC the
pilot must rely upon the aircrafts instruments, this requires a higher level of training and skills.
Pilot that do not have this training and skill lack the psychomotor and cognitive skills necessary
for controlling the aircraft.

2. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
2
Accidents happen when pilots make the decision to fly into IMC instead of changing the,
sometimes, safer option of diverting. It is importing to understand the influences on this decision
so that in the future measures, trainings etc. can be implemented for a better decision-making
process, which results in less hazardous situations. Inaccurate decision making by, in particular,
pilots in GA flying solo are caused by inexperience, bad attitude and cognitive biases. These
influences result in an inaccurate assessment of the weather conditions, a wrong or inappropriate
risk perception and/or motivational factors that can all potentially cause a pilot to fly into
adverse weather.
This paper focuses on the effects of adverse weather on the pilot behavior, particularly on
decision-making. Furthermore, the influences on this decision-making process and current
researches regarding these influences will be discussed. This paper does not consider a
difference between adverse weather conditions and IMC.
2. Literature review
2.1 Accidents
During the last decades, there has been a lot of research performed regarding this subject. Most
of the researches assessed a sample of pilots by examining the pilots in a simulation in which
weather conditions change. An accurate weather assessment is essential for adverse weather
situations, in most researches the weather assessments of the pilots were based on the visibility
and cloud height assessment. The minima for flying VFR according the FAA are three statue
miles flight visibility, a clearance of 1,000 and 500 feet respectively above and below, and a
2,000 feet horizontal distance from clouds (NTSB 2005).
As already discussed, the primary cause for weather-related decision making in accidents is a
combination of in-experience, in appropriate attitude or cognitive bias. One of the core elements
of decision-making is information. Thus, a difficulty in information acquisition, information
interpretation, integration can result in an inaccurate weather assessment.
Study reveals that only 45% of the pilots whom encounter adverse weather, actually
talks to a weather briefer or seek help from air traffic control (ATC) (Hunter, Martinussen et al.
2011). The decision to fly into adverse weather is obviously pilot dependent; study reveals that
there are commonalities in pilot characteristics that avoid adverse weather. Pilots avoiding
adverse weather are more likely to have more instrument time, have the availability of an
autopilot, a longer duration of flights and, a larger use of the telephone briefer.
Unfortunately, there is an unwanted side effect of flying into the adverse weather and
surviving without any problems and consequences for a pilot with a lack of instrument rating.
Every time nothing seriously happens, the pilot may grow overconfident until the pilot
encounters a situation where the demands exceeds their capabilities. Another research
determined that approximately 7.5% of VFR flights into IMC that resulted in an accident were
caused by this overconfidence bias.(Goh, Wiegmann 2001) The overconfidence is one of the
many cognitive biases that influences decision-making. The effects of confidence bias
unfortunately can be catastrophic therefore; the chain effect must be broken.
Another possible effect of the overconfidence bias is the lack in appreciation of the risks
involved in the actions of the pilot. If the pilot do not appreciate the real danger, it could cause
a pilot to continue the flight into adverse weather.
Some psychological characteristics of the pilot, such as overconfidence, are
scientifically proven different for pilots who fly into IMC compared with pilots who decide not

3. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
3
to. Pilots that do fly into IMC have a higher self-rating of skill and judgement, and overestimate
visibility while pilots that do not fly into IMC had a higher estimate of the likelihood that
weather would be the cause of an accident.(Goh, Wiegmann 2001).
Interestingly is that, despite the danger of VFR pilots flying into IMC it remains a problem in
GA. This problem is already known for a long time, initially researches were focused on pilot
demographics, and operational and geographical factors while currently the focus lies more on
the psychological factors since the aircraft and pilot training have become more safe and reliable
(Walmsley, Gilbey 2017)
2.2 Decision-making
Considering the psychological part, inappropriate decision-making can be caused by multiple
cognitive biases. The continuation, anchoring and confirmation bias has been researched as
most influencing in decision-making while approaching adverse weather. The continuation
bias, in other words motivational factor, is unconsciously continuing with the original plan
despite changing conditions. Pilots tend to rely too much on the initial piece of information
(anchor) however, while situations and circumstances change this may result in under-adjusting
the current available information. The confirmation bias is hazardous because humans are
affected by it all the time; pilots tend to favour confirmative evidence that supports the
hypothesis or expectation, instead of looking at all the information available. What the last two
biases have in common is that pilots have the tendency to look at a too narrow range of
information.
Since these cognitive biases could cause a pilot to make a hazardous decision, a lot of
research have been performed to understand the cognitive biases with the goal to prevent them
from happening (debiasing) which could result in a more critical decision-making. The
principle of debiasing lies within changing from mode 1 (intuitive) to mode 2 (analytical)
(Walmsley, Gilbey 2017). Mode 1 tend to use heuristics and is used unconsciously, meaning
automatic and fast. The shortcoming however, is that, cognitive biases can affect this type of
decision-making. The utility of cues is connected to mode 1. A cue is signal that triggers the
pilot to act; this signal aids the memory with retrieving details that are not recalled
spontaneously. The utility of cues result in more time and cognitive resources for other
decisions. (Wiggings, Azar et al. 2014), this study concludes that a higher cue utility result in a
more likelihood of flying into adverse weather. Cue utility in the adverse weather conditions is
rather unusual because of the dynamic conditions, with weather in different forms and, the speed
of the aircraft requires decision-making in short-time period. Mode 2 is used more conscious
and therefore more controlled but this results in a higher cognitive workload. An optimal
balance between these would be a shift from mode 1 to a slow mode 2. Debiasing techniques
aim for this balance.
2.3 Techniques to influence psychological deficiencies
One of these techniques is “consider the alternative,” this technique encourages a person to look
at a broader range of information. This technique has been proven, by research, successful in
reducing overconfidence bias, hindsight bias, confirmation bias and anchoring bias.
Unfortunately, research has also proven the contrary: an unsuccessful debiasing and even a
stronger effect of the initial cognitive bias. A recent study (Walmsley, Gilbey 2017) using this
technique to reduce the anchoring and conformation bias has again showed unsuccessful, this
results implicates that there is a need to explore other debiasing technique. Other techniques

4. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
4
that has proven successful are more intrusive, this however, is again difficult to implement in
the dynamic, high workload environment of a cockpit.
Besides training there are at least three other potential measure described that affect cognitive
biases: proceduralization, automation and decision support tools, and display and information.
Proceduralization is likely to reduce some of the cognitive biases, such as, confirmation bias.
However, the anchoring bias e.g. is not effectively affected by this method. This method also
relies on an accurate perception of weather conditions, and is impracticable to fully implement.
Automation and decision support tools have the purpose to reduce the workload and therefore
result in more capability to focus on mode 2 instead of the fast mode 1 of decision-making.
Display and information has the primary focus on generating a richer flow of information. Not
only improving weather related systems helps but also improving e.g. navigation systems helps
by reducing the workload of navigation and thus, more cognitive resources for other tasks. This
can be accomplished by e.g. a more advanced display. One disadvantage of more advanced
displays is that the pilot may be spending more time looking inside the cockpit instead of the
outside.
NASA just claimed their patent on a new system. This system is capable of displaying
aircraft trajectory and weather forecast with the benefit that the predicted aircraft position is
displayed relative to the predicted weather planning. This system has on top off all option for a
two-dimensional, 3-dimensional and a potential 4-dimensional display (Johnson, Wong et al.
2017).
Besides new system inventions, the weather display symbology also affects the pilot
behavior. Display symbology research concluded that information is better recalled when the
information is displayed in both text and symbols. Along with the introduction of symbols
comes the colour of the symbols. Colour preference is not for all individuals equal. In colour
change however, are better perceived commonalities. Colour change from white to red, for
example, is better perceived than other combinations. Confusion on the other hand, is
disadvantage of the usage of colours, deviation from safety colours and other familiar colour
schemes can cause misinterpretation. Besides that study shows, display symbology, high-
resolution display and others results in closer proximity to adverse weather areas (Ulf Ahlstrom
2015).
Besides the effect of adverse weather on decision-making. The adverse weather also affects the
pilots’ eye behavior. When pilots face unexpected conditions the eye behavior changes. Pilots
tend to look more outside the window instead of inside the cockpit to the instruments due to the
limited training they have had. This may result in missing relevant information. The visual
attention is individual-dependent. No matter the outside conditions pilot focus more times on
outside view than on other areas of interest (AOI). When pilots do focus, on the instruments,
they tend to focus more to the primary flight display instead of the back-up display and pilots
pay more intention to the GPS than other primary displays. When pilots encounter adverse
weather the pilots look more to the outside view than instruments, and they check the PFD more
than the back-up instruments. After encountering adverse weather, they looked more to the
directional gyro (DG). This makes sense since it confirms the direction of the aircraft.
2.4 Limitations
One study is based on a self-report questionnaire; a questionnaire’s reliability is vulnerable to
misinterpretation, wrong recall of events, deliberate distortions. Other simulating studies are

5. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
5
limited by the extend to which a simulation equals the real world-circumstances, such as, use
of static images etc. Generalizability is a problem; does the sample of pilot really represent the
actual pilot behavior?
3. Discussion of the findings from the literature
3.1 General discussion fatality rate GA weather-related accidents
Interesting is the percentage of fatality in all of the weather-related accidents in GA. Between
1990 and 1997 the fatality percentage was 75%, a percentage well above other fatality rates.
This is due to the lower level of training and skill of most GA pilots with IMC. Although
instrument training is a part of the pilot license, there are still some requirements, which must
be fulfilled before a pilot, is authorized to fly into weather conditions below the VFR minima;
the aircraft must be rated for IFR and the pilot must have an IR rating. Thus if a pilot who is
only rated for VFR faces adverse weather and has no other option than flying into the
deteriorating weather conditions, forced to fly IMC, he is unauthorized and untrained. This
discussion raises the question: how is it still possible that a pilot can unauthorized and, more
important, untrained, end up in a hazardous situation that has shown to be fatal? The main
answer to this question is the expensive ratings and trainings. However, this raises another
question: what is more important safety or costs? How much is the safety worth? This ethical
question might be always be unanswered, but it illustrates one of the main problems of this
subject. Reflecting on this part, the overconfidence bias has a specific role in this matter. In the
US many pilot fly only VFR due to the expensive costs for ratings. A huge factor is the thought
about VFR into IMC accidents “this won’t happen to me,” this particular thought is an
overconfidence bias and might be one of the major causes of these accidents.
Even more interesting is the behavior of that percentage over time, found in the literature. In
2007, the percentage was 73% (lowered) while in 2011 the percentage was 86% (risen). This
rise in percentage might come due to the increased cognitive biases among GA pilots described
previously. This is unfortunately because all of the research regarding this subject before 2011
had apparently no effect on the current pilots. This statement is supported by the following fact,
all of the surveyed pilots without an IR seemed able to survive brief encounters with adverse
weather (Hunter, Martinussen et al. 2011). This statement implicates that it does not matter
whether pilots have an IR or not.
3.2 What happens to the pilots whom are facing adverse weather and have to make the
decision?
Pilots who are approaching adverse weather behave different to this unexpected condition. The
pilots’ history plays a role in these situations. The general behavior is different for pilots who
have: already encountered similar conditions once, have instrument rating, and other criteria
such as e.g. age. These factors do not directly help for generating measures that can prevent this
problem. The outcomes of these studies do help by selecting the type of pilots where measures
have the greatest effect.
However, the reliability of the outcomes is important is this matter, with respect to the
generalizability of the results. Typical samples for the researches were 57, 24, 30 and 64 pilots.
In my opinion these sample are quite low, and since in most of the researches pilots volunteer
to participate in the research. Therefore, the outcomes are by definition not generalizable.
Besides that most of the researches and therefore the sample of pilots originate from the US,
the US does represent a huge percentage of the weather-related accidents but obviously not the

6. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
6
entire GA industry. Ethnic difference has an influence on the behavior of humans and therefore
pilots. This is also the part were again, the overconfidence bias is influenced, in the sense of the
magnitude of the bias.
The general behavior of the pilot changes with unexpected condition, research implies that
pilots felt a bit tense and anxious during the event of flying into IMC, and after the event pilots
felt more cautious about weather (Hunter, Martinussen et al. 2011). One of the different
behaviors is that pilot contact weather briefers, which is a good resource to find accurate and
up-to-date information which helps assessing the weather. Unfortunately, only 45% talked to a
weather briefer in adverse weather and 57% near adverse weather (Hunter, Martinussen et al.
2011). If this ratio would be, higher pilots in general could have a more accurate weather
assessment.
Cue-utilization is another effect that the weather has on pilots. Pilots with a higher level of cue-
utilization were more likely to fly into adverse weather. Therefore, logics would say that a
training regarding cue-utilization should be effective. The aim for training would be to identify
the specific cues and “tune” the actions associated with the cue (Wiggings, Azar et al. 2014).
This could result in a higher skill in the acquisition of information, one of the key elements of
decision-making. Connected to cue-utilization is the behavior of the pilots’ eye (Russi-Vigoya,
Patterson 2015). Concluded in this study is that pilots look more to the outside view instead of
inside the cockpit. This means that the cues inside the cockpit are less noted by pilots, provided
that, they fly solely. The method of the study however, questions the quality of the outcome. As
simulator Microsoft Flight Simulator was used in combination with a 46-inch screen, this does
hardly simulate the real environments in a cockpit.
The psychological issues affect the behavior of the pilot in terms of cognitive biases and
risk perceptions. These factors lead to the decision to fly from VFR into IMC. Many studied
these factors but an appropriate effective training is to be found yet. The difficulty with this
problem lies within the limitations in simulation reality, training of cognitive behavior and
practicability within the dynamic cockpit environment.
3.3 Measures preventing the influences on decision making
Most research nowadays focus on broadening the range of information that reaches the pilot for
a better assessment. Besides that, research also focuses on debiasing techniques. Broaden the
range of information can be accomplished by enrichening the information flow, a wider
spectrum of information, this however results in a higher cognitive workload, and removing
cognitive workload from other system results in more cognitive resources for weather related
decision-making (Walmsley, Gilbey 2017). Improving weather related displays will enrichen
the flow of information while improving other systems related displays will reduce those
systems related cognitive workload. Investigating on symbology is a solid method to enrichen
the information without compromising anything, provided that, the symbology will not confuse
the pilot. Debiasing techniques have proven successful in other sectors than aviation but have
also proven unsuccessful.
The big problem I have with the results of researches that investigate biases or debiasing
techniques is that most of the outcomes are similar to one of the following three answers: “this
may be successful”, “this could be effective” or “more research is necessary”. This problem is
related to the already discussed dynamic cockpit environment, difficulty in implementation.
Another new factor influence the outcomes also. Researches imply that they have failed because
the method was incomplete in terms of investigating all possible effective outcomes. E.g.,

7. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
7
(Walmsley, Gilbey 2017) stated that the method could have possibly failed because the
debiasing technique was only introduced brief and passive. I believe that this uncertainty in
outcomes and limitations is a significant factor in unsuccessful researches, in terms of the
wanted outcome.
4. Conclusion
Statistics prove that weather-related accidents are consistent and have a high rate compared to
other type of accidents. Fatality among the VFR into IMC accident is unusual high and is
apparently difficult to lower. The main problem for the accidents to happen is the lack of
experience and skill by pilot who fly seldom and unauthorized into IMC. Before entering
adverse weather conditions, the pilot has to make the most important decision: continue into the
adverse weather or divert with the accompanied consequences. The decision-making process is
influenced by many discussed factors. Most significant factors are cognitive biases, but
eventually it boils down to an inaccurate weather assessment or inappropriate risk perception.
To improve this, pilots must improve information acquisition, interpretation and integration to
form accurate, analytically (mode 2) based conclusions. Current effective methods are enhanced
displays and symbology perfection, these enrichen the information and can reduce the cognitive
workload. Techniques that are focused in countering the psychological “errors” lack a sufficient
proof of effectiveness and practicable implementation options. Unfortunately, as many studies
conclude, there is still a need to for effective methods that counter the deficiencies of the
decision-making process.
5. Relevance for practice and the Aviation academy
This paper summarizes the current state of pilot behavior, decision-making during adverse
weather conditions, and studies regarding the influences on those decision-making, on top of
that the paper discusses current technologies related to pilots’ behavior. Future research may
use this paper to identify the success and failures of the subject. Identifies success can be
investigated why they succeeded and use the result of investigation to improve the failures. The
Aviation Academy may use this paper to be triggered in this subject and may even implement
the associated problems into research or courses.

8. AUAS B. Roosendaal / Weather-related-, decision-making behavior 2017
8
6. References
GOH, J. and WIEGMANN, D., 2001. An Investigation of the Factors that Contribute to Pilots
Decisions to Continue Visual Flight Rules Flight into Adverse Weather. Human Factors and
Ergonomics Society Annual Meeting Proceedings, 45(2), pp. 26-29.
HUNTER, D.R., MARTINUSSEN, M., WIGGINS, M. and O’HARE, D., 2011. Situational and
personal characteristics associated with adverse weather encounters by pilots. Accident Analysis
and Prevention, 43(1), pp. 176-186.
JOHNSON, WONG, D.W., WU, S. and KOTESKY, R.W., 2017. (12) United States Patent.
Washington, D. C: .
NTSB, 2005. Risk Factors Associated with Weather-Related
General Aviation Accidents. Washington, D. C: .
RUSSI-VIGOYA, M.N. and PATTERSON, P., 2015. Analysis of Pilot Eye Behavior during
Glass Cockpit Simulations. Procedia Manufacturing, 3, pp. 5028-5035.
ULF AHLSTROM, 2015. Weather display symbology affects pilot behavior and decision-
making. International Journal of Industrial Ergonomics, 50, pp. 73.
WALMSLEY, S. and GILBEY, A., 2017. Debiasing visual pilot's weather-related decision
making. Oxford: Elsevier.
WIGGINGS, M.W., AZAR, D., HAWKEN, J., LOVEDAY, T. and NEWMAN, D., 2014. Cue-
utilisation typologies and pilots’ pre-flight and in-flight weather
decision-making. Amsterdam [u.a.]: Elsevier.