A study of ASRS reports involving general aviation and weather encounters

(NASA-CR-166212) A STUDY OF ASRS REPORTS N81L.27063 INVOLVING GENERAL AVIATION AND WEATHER ENCOUNTERS InteEi| RepoEt (Battelle Colunbus Labs., Ohio.} 57 p HC AOq/_F A01 Unclas

(

31102

cscA_o !.c G3/_o 3

A STUDY OF ASRS REPORTS INVOLVING GENERAL AVIATION AND WEATHER ENCOUNTERS By Thomas H. Rockwell, Darrell E. Roach, and Walter C. Giffin June 26, 1981 Distribution of this report is provided in the interest of information exchange. Responsibility for the contents resides in the author or organization that prepared it.

Prepared under Contract No. NAS2-10060 by BATTELLE COLUMBUS LABORATORIES ASRS OFFICE 625 Ellis Street, Suite 305 Mountain View, California 94043 for AMES RESEARCH CENTER NATIONAL AERONAUTICS AND SPACE ADMINISTRATION REPRODI_ED OY NATIONAL TECHNICAL INFORMATION SERVICE U.S.DI_PAIITU[III OF COMMERCE, SPRINGFI£LD, ¥A. 22161 TABLE OF CONTENTS Page i SUMMARY . • • • • • • • • • • • • • • • • • • • • • • • • • B • • • • I INTRODUCTION ............................

STUDY OBJECTIVES ..........................

APPROACH ...............................

CHARACTERIZATION OF THE SINGLE IFR AIRCRAFT WEATHER-RELATED INCIDENTS ..............................

i0 Generalizations ........................

SINGLE VFR AIRCRAFT IN ADVERSE WEATHER ...............

ii MULTIPLE AIRCRAFT WEATHER-RELATED ASRS REPORTS ...........

Introduction ..........................

IFR/IFR Incidents ........................

Weather Factors ......................

Problem Nature . ; ....................

Problem Causes ......................

Recovery Factors ..................... 3O Generalizations ...................... 3O Contributing Factors ...................

Recovery Factors .....................

WEATHER INFORMATION SERVICES ....................

iii

Preceding pageblank

TABLE OF CONTENTS (Continued) Page WEATHER-RELATED INCIDENTS INVOLVING BALLOONS .............

LIMITATIONS TO THE DATA .....................

SPECIAL ISSUES ............................

Role of Increased Workload in General Aviation Problems With Obtaining Clearances in Flight ..........

Weather Involvement and General Aviation Operations in the Vicinity of Airports ...................

APPENDIX A. ACCESSION NUMBERS USED .................

iv A STUDY OF ASRS REPORTS INVOLVING GENERAL AVIATION AND WEATHER ENCOUNTERS by Thomas H. Rockwell, Darrell E. Roach, and Walter C. Giffin* SUMMARY This report describes a study of material in the ASRS data base dealing with weather-related incidents in General Aviation. Factors leading to such incidents are discussed and analyzed. Consideration is given to the nature and characteristics of problems involving dissemination of weather informa- tion, use of this information by pilots, its adequacy for the purpose intended, the ability of the air traffic control system to cope with weather-related incidents, and the various aspects of pilot behavior, air- craft equipment and NAVAIDS affecting flights in which weather figures. It is concluded from the study that skill and training deficiencies of GA pilots are not major factors in weather-related occurrences, nor is lack of aircraft equipment. Major problem causes are identified with timely and easily inter- preted weather information, judgement and attitude factors of pilots, and the functioning of the ATC system.

INTRODUCTION Analysis of general aviation aircraft accidents indicates that pilots The National often fail to deal effectively with weather-induced problems.

*The authors are faculty members at Ohio State University and serve as con- sultants to the ASRS program.

Transportation Safety Boardrs tabulation of 1978 general aviation accidents,*

published in its annual review of accident statistics, indicated that

weather-related accidents led all other types of causes. There were 150

fatal accidents resulting from pilots encountering bad weather; whether this

is the result of inaccurate forecasts, deteriorating weather, or poor pilot

judgement, constitutes an important safety problem for general aviation.

The question is often posed about the ability of general aviation pilots

to cope with weather problems. What is not known is the extent to which such

problems are due to inadequate weather information systems, inadequate air-

craft capability, lack of sufficient training for general aviation pilots, or

lack of responsiveness of the air traffic control system.

STUDY OBJECTIVES

The ASRS data base on General Aviation weather-related incidents was

examined to shed light on how humanor systems problems link with weather to

create unsafe conditions. The overall objective of this study was to deter-

mine if patterns emerge which would explain more fully current accident

reports and to suggest appropriate counter-measures. The ASRS data were

examined to ascertain:

I. The nature of the weather problem in the related

incident. Factors considered included the weather condi-

tion at the time of the incident, the pilot's use of

information systems available to him, the adequacy of the

weather information for planning purposes, and the abil-

ity of the air traffic control system to provide weather

information.

2. The characteristics of the weather-related incident. Did

an unsafe condition result, and how serious was it? What

types of aircraft are most often involved?

3. The types of pilot and system problems contributing to

the incident.

*As reported in AOPA Pilot, December1980, p. 88.

4. The nature of pilot and system attributes which contri-

buted to the problems, e.g., pilot experience.

APPROACH

The basis for this study was all reports of incidents from the ASRSsys-

tem covering the period from May 1978 to September 1980. The search pro-

cedure involved selection of all incidents which met the following criteria:

a. The incident was weather related, i.e., weather was con-

sidered a contributing factor in the cause or consequence

of the incident; weather, for purposes of this report, could be instrument meteorological conditions (IMC),

i.e., visibility below three miles and/or ceilings below

i000 ft. (agl) in controlled airspace; marginal VMC, i.e., visibility 3-5 miles and/or ceilings 1000-3000 ft.,

visual meteorological conditions (VMC), i.e., visibility

greater than 3 miles and ceilings above 3000 ft. (agl) in

controlled airspace and visibility one mile and clear of

clouds in uncontrolled airspace. Mixed IMCand VMC (MXD)

could also be a factor as will be shown later.

b. The incident involved at least one aircraft which was not

an air carrier or military aircraft. This embraces all

general aviation aircraft from small aircraft (SMA-less

than 5000 ibs.) to small and medium transports. In the

case of a potential conflict between two aircraft, mili-

tary and air carrier could be involved if one of the air-

craft involved "General Aviation" operations. This term

embraces all flight operations except air carriers and

military, and includes pleasure, business, corporate, and

air taxi.

One hundred seventy-seven reports met the above criteria.

It was evident at the outset of this study that these reports were not

homogeneous with respect to their characteristics, e.g., weather, type of

aircraft, location, etc. Thus, it was decided to classify these reports into

six categories as follows: a. Single aircraft under IFR flight (27 reports).

b. Single VFRaircraft in adverse weather (52 reports).

c. Less than standard separation (LTSS) or potential con-

flicts (or in the extreme case, near midair collision

events) between two aircraft operating on IFR flight

plans (IFR/IFR); (34 reports).

d. LTSSor potential conflicts between two aircraft, one of

which was presumed to be VFRand one of which was usually

on an IFR flight plan (mixed flight rule operations).

(43 reports).

e. Weather services which, while not involving a specific

in-flight event, were included because of complaints

about the weather information system or the inadequacies

of the FSS system and ATCsystem (14 reports).

f. Balloons (5 reports).

Because of multiple reports for the same incidents, there were 27

incidents for the single IFR aircraft, 52 for the single VFRaircraft, 23 for

the IFR/IFR and 42 for the mixed flight rule two aircraft incident. Thus,

the 177 reports actually described 163 separate incidents for all six

categories.

Setting aside the balloon and weather service reports, the basis for the

four major classifications can be seen in Figures 1 to 7. Figure I shows

that the four categories are quite different in terms of the involvement of

small general aviation aircraft (less than 5000 Ibs.) which was one of the

main interests in this analysis. In the IFR/IFR cases, small general avia-

tion aircraft were virtually non-existent. Yet, for the single VFRaircraft

incidents, 90 percent were small general aviation aircraft. Figure 2 indi-

cates that over half of the single IFR incidents occurred under 1000 ft. agl

while virtually none (4 percent) of the IFR/IFR incidents occurred at or

below this altitude. Where humanerror was assigned by the ASRS analyst,

Figure 3 shows that flight crews were indicated in 80 percent of the mixed

flight rule incidents as compared to 50 percent of the single VFRand 35 per-

cent for both the single IFR and IFR/IFR incidents. Figure 4 indicates that

IMCconditions existed in 75 percent of the single IFR incidents, but in only

45 percent of the single VFRincidents. Here marginal VMC and MXD weather

was reported in 41 percent of the incidents. Even the time of year varies

across these categories as shown in Figure 5. Note that 26 percent of the

single IFR incidents involved summermonths, while 50 percent of the two-

aircraft incidents occurred in the summermonths. This fact supports the

::.:...

Key: b?-.: r,\ i!?:i!i!

I-J

.: .o.

80 i'£o".

!;.<.; Single Single IFFI/ Mixed C/: ..,'2 IFR VFR IFFI Flight ':'..:i:, Rule .'°j.

o.°.

o.- %':: 55 ....'- #::T:::::::: $i$i$i: 50 50-- .,':.;"2 •....... • ;.:.:.:+:< ¢D ;'.;% ::::2:::::: .... i .;.x.:.:.: P .'::.* ;;iii!;!;iZ!; " 40 40- :.:.:.:.:.:.: .....

.o. _ :.>:.:.;.:.: :.°.

i_;i;iiil;ii;i ::o':: :::::::::::::; 30- i!_!iil;ii_!ii (Ib .;t°'j: P :::::;:::::::: I.-': :. " ::::::::::::: 20 .....

52..'. " 20 - ::::::::::::: ..... $;:;:i:i:!: !. "..::.

..o ::::::::::::: :.::5'."

, .....

i.. -..

::::::::;:::: 10 - 4 = :::::*::::::

. EJ

, ....

"y:: iiiiiiiiiiiii

K:'-:-.!

FIGURE i. PERCENT SMALL FIGURE 2.

PERCENT OCCURRING AIRCRAFT (SMA) BELOW i000 FT A.G.L.

8O 80 - i:i:i:i$i:l 70 - :!:i:i:i:i:i!

6O ::::5::::.:: 60 - :'!:15i$: 6O ....° iliiiiii_i:! ,:+:.•:.:.: :::::::::5::: ::-i:i:i:i:!: ::::::::::::: 50--

...- i!ii!i!!ili!

;::::;:;::::: • -. _j :i:$i:i:!: i:!:i:!:i:il ,.,\-., :i:i:i:iri$ ...., ,:.:.:.::+ .... • | ,,\\ ,.....,....

40-- ..:::'-:.:: ::-::2::::: 35 I ,,\\ 40 :::::::::::::: :i::::| 35 :....<>: 'II,IIl L+:+: _..,.., :- .:.. ": :!:i:i:i:!:l ::.:::::::::: . "'4 %\_ :.....• :!i!ii!

i:i:i$i:i: I _. 30 :.°.., ...-....

' ; '. ". t %%% .........

.o:1 'o ;;.o-. :i::::::::: %%% .o•_o- i!iiiiiiiii| ...Oo] ::..--:'...

f:.:::::::: 20 ::+:<.

i i iiiiiii!

.:._':-'.. <::::::::::: ,.........

.....

";t:-.!

::::::::::::: ..-...-.

:!:!:i:i:i| :::i:i:!:[:i: ::::::;:::! 10 "::::i_!

:i$[:i:!::: :::::2:::::

-.: ... ':.":1v:'_,: -'": -:. _" _. ;" ": "I",',",,,, ,.,,"., .. ,"

_!i!i!i!i!i!

...::: :.:: ..::::::::: .<::+:: ": :':.'F_ ; :::f::.

FIGURE 3.

PERCENT _{ERE FLIGHT FIGURE 4.

FLIGHT WEATHER CREW WAS ASSIGNED CONDITIONS AN ERROR PERCENT IMC 60 - 57 60-- 50 - 5O 45 _-., \\ \\ ,,... .,.,,...., .......,.

\\ ..,....,.

40 :i:i:i:i$i: \\ ...........

\\ :;;:::;:::::; ............

\\ :::::::::;::: _g :::::::2::: aD :::::2::::: u 30 30 \\ ;:;:;:;:::::; ¢g ::;::::::::::: ...........,.. \\ ;::::::;:::::: 0,.

:.:.:.:.:+: \\ 2o - .}.::_:':" :i$;:18i:i :.:+:+;.:. \\ .;...

:,:,:+:,:.: i'.:'." ::;::::::::::; \\ :.:.:.:.:.:.: i:.:,:+:.:.: \\ lo !';')!i?: .:,:,:,:+:, ...,.,,., .,..,....., :::::::::::t:

1A 'O

:!:i:i:!:i$ ..,, ....,..

\\\ ...,,,..,, ..,.,,.v,.

FIGURE 5. PERCENT INCIDENTS FIGURE 6.

PERCENT THUNDERSTORMS OCCURRING DURING JULY, AS FLIGHT ENCOUNTERS AUGUST, AND SEPTEMBER Key: ,............

DDZ

Single Single IFR/ Mixed IFR VFR IFR Flight Rule 7P 70-- :,:,:.;.:+: :i:!$1:i:i: 60-- iiiii!iiiiii!

iiii!iiiiiiii_ 50-- m iiii!iiiiii!i •- 40-- e-

iiiiiiiiiiii!

£. 30- iiiiiiiii_iiii 22 o: i:i:i$1:i:i: ::::::::::::: ;,:.:+:.:: 20- z !iiiiiiiiiiii ::::;:::::::: :::::;;:::::: iiiiiiii!iji; ::::::;::::::: 10 - :;;:;:;::::::; ::.:::1::::::: !i!i_ii!_!!!i .,....,..,.

FIGURE 7.

PERCENT OF CASES WITH TOTAL FLIGHT TIME ABOVE 1500 HOURS (OF THOSE REPORTED)

differences found in the in-flight weather encounters. More than half of the

IFR/IFR incidents reported thunderstorms, while less than 15 percent of the

other three categories involved thunderstorms. (Figure 6).

Where the pilot flight time was noted, approximately 75 percent of the

single IFR and mixed flight plan pilots had more than 1500 total hours, while

this figure drops to 22 percent for the pilots in the single VFR aircraft

incidents. (See Figure 7).

With the above pattern differences, it was clear that adding statistics

from these categories would have produced erroneous conclusions. Hence, the

subsequent analysis treats each of the basic categories independently.

Later, in the conclusion section, attempts will be madeto draw inferences

among these categories.

CHARACTERIZATION OF THESINGLEIFR AIRCRAFT WEATHER-RELATED INCIDENTS

Twenty-seven of the 163 total weather-related incidents examined

involved a single IFR aircraft. Of these incidents, 45 percent were small

general aviation aircraft (less than 5000 ibs.). About half of the incidents

occurred in the period of Noon to 6:00 P.M.; 75 percent were reported in day-

time conditions; 55 percent occurred under 1000 feet (AGL) and were involved

with approach, landing, or take-off phases of flight. Unlike other

categories of incident reports of GAaircraft in weather, which were over-

represented in the summermonths, the winter months were over-represented in

single IFR incidents. Over 50 percent of all single aircraft IFR incidents

were reported in Decemberthrough March. For these incidents, the weather

was usually IMC (75 percent of the cases).

It should be noted that except for four or five cases, the weather

should not have been a significant factor in the incidents. IFR flight sys-

tems are designed to permit safe and efficient air transport in poor weather, e.g., 200-feet ceilings and I/2-mile visibility for most ILS approaches.

Thus, it might be argued that, in single aircraft IFR operations, weather

should not pose a serious problem. The few cases of enroute icing and turbu-

lence are the exceptions in this class of incidents.

Since 55 percent of the incidents involved small transport (SMT) or

larger aircraft, it is not surprising that pilot qualifications (in the cases

where this was reported) showedthat 75 percent of the pilots had more than

1500 hours. The nature of the single IFR aircraft problem is best shown in

Table I. Note that the 27 incidents were widely scattered over 8 problems.

Landing below published minimums, severe weather encountered, and failure to

accept or receive clearances represent the most reported incidents.

TABLEI. SINGLEAIRCRAFT IFR INCIDENTS

Failure to secure or accept clearance • • • 5

Vectors into severe weather

(or failure to receive vectors

In assessing human errors associated with these incidents (Table 2), it should be noted that in many cases the error is not always certain -- the classic case is landing when the weather is near published minimums.

Although the weather may be just below minimums, it can change quickly. In one incident, the pilot argued that the weather was well above minimums rather than below. The errors behind five incidents were caused by pilots breaking instrument flight rules regarding clearances and landing minimums.

Most of these incidents involved small transports with very experienced pilots. In one of these incidents a crew acknowledged landing when the weather was below published minimums.

TABLE2. ERRORS ASSOCIATED WITH SINGLEAIRCRAFT

IFR WEATHER-RELATED INCIDENTS

CommunicationProblems ............. II

Communications problems, of the eleven most involved a communication

breakdown between pilot and controller (six instances). These included alti-

tude assignment misunderstandings, problems in securing clearances, and

weather deviation. Surprisingly, there were many cases of communication

problems between combinations of pilots, flight service stations, tower con- trollers, approach controllers, and enroute controllers, as typified by the following narrative: "Myself, a journeyman CTLR, and the FAC chief (working for a CTLR on sick leave) were on duty, day shift, a VFR TWR. WX autowriter was out and FSS agreed to supply RAP- CON with current WX. ILS rwy 3 in use. Apch called with SMT A 20 south, ILS 3. I asked if he had the latest WX with I/4 mile vls. He said no. I gave him the WX, adding that the RVV on 34 was 5/8 and recommended that rwy. RAPCON CTLR responded OK, I'll tell him. Snow removal was in progress on all rwy, I told gnd cntl to clear the plows off 34 and turned the apch lights on.

SMT A reported the outer marker, was clrd to land. After idg, and taxiing is not visible from the TWR, it was resolved SMT A had made the ILS 3 apch and landed on 3 with a large snow sweeper in the center of the rwy. A turned off short of that point and never saw the vehicle.

I thought the conditions automatically dictated 34. The apch CTLR merely gave the info to the pilot who continued for 3. There was no further discussion."

Five incidents involved lack of proper information on runway conditions.

These resulted in diversions due to closed runways, flareouts in deep slush, landing excursions, and a near-accldent with a snow sweeper. There were five cases of poor pre-flight preparation by pilots - two involving decision to fly into known icing conditions.

Table 3 depicts some of the factors behind the observed errors. Weather interpretation, lack of knowledge of the IFR system, lack of airport informa- tion, and unprofessional attitude, as inferred by the ASRS @nalyst, dominate these data.

TABLE 3. CONTRIBUTING FACTORS IN HUMAN ERRORS IN SINGLE AIRCRAFT IFR WEATHER-RELATED INCIDENTS Weather Interpretation ..........

Knowledge of the IFR System Weather Induced Workload .........

Miscellaneous ..............

Total Generalizations For the single IFR aircraft, the 27 weather-related incidents could almost be considered substandard performance in the IFR system. Severe weather abnormalities were infrequently involved. Most incidents involved normal IMC. The incidents were less involved with General Aviation small aircraft than with larger General Aviation transports, suggesting that pilot experience and aircraft equipment were probably not key factors in the incidents. If weather, experience, and aircraft capability were not key

causes, what were the bases of the incidents? Four major causes appear to be

prevalent in the data: I. Violations of instrument flying rules.

2. Communicationbreakdowns between FSSspecialists or air

traffic controllers and pilots.

3. Poor pre-flight planning.

4. Weather interpretations, i.e., the recognition of when

conditions have gone from VMC to IMC.

Violations of instrument flight rules include pilots bending the rules

or breaking them altogether -- this ranges from landing when the weather was

below published minimumsto refusing to accept clearances. Reports of com-

munication confusions were expected; such problems are always with us. What

was not expected was the lack of transmission of airport and runway informa-

tion for winter flying. Poor pre-flight planning includes not only failure

to secure necessary weather information, but also taking chances with icing

in aircraft not equipped to cope with it. Weather interpretation involved

rake-offs into IMCwithout IFR clearances.

SINGLEVFRAIRCRAFT IN ADVERSE WEATHER

A total of 52 incidents involving VFR pilots who encountered adverse

weather conditions were identified in the ASRSdata file. It should be

pointed out that these incidents may not be representative of typical general

aviation weather encounters. In most of these incidents the pilot was in

contact with an ATC facility or flight service station. None resulted in accidents (to be expected; ASRS does not process accident reports) and only one involved an off-airport landing. Thus, the GA pilot who extricates him- self from a potentially dangerous situation and does not report to anyone is not represented, nor is the pilot involved in an accident. It is worthwhile, however, to examine the reported incidents to ascertain if patterns exist which might lead to improvements in weather reporting, the ATC system, or pilot training.

Ii

The analysis which follows resembles a flight log including pre-flight

weather briefings, the flight phase of the weather encounter, the pilot's

action in response to the encounter, the nature of assistance requested, if

any, the recovery factors, and other contributing factors.

Almost all (49 out of 52) of the incidents in this set involved small

aircraft. The other three incidents involved small transports. The majority

(43) were reported by pilots. The others were reported by controllers (7)

and observers (2). The two observers reporting were flight instructors

reporting on situations encountered by their students.

While, as Table 4 shows, incidents occurred in every month of the year,

they were most frequent during the summermonths. This is probably the

result of a higher level of general aviation activity during the summer.

TABLE4. REPORTED MONTH OF OCCURRENCE

Month I Incidents Month Incidents

6 I0

January July

5 I0

February August

March 3 4

September

5 October I

April

1 November 2

May

June December I

Total - 52

Little information concerning pilot ratings was available from the ASRS

reports. In 41 of the incidents the pilot's rating was reported as unknown.

It may be inferred that these were largely non-lnstrument rated pilots.

Pilots of all experience levels were involved in these weather

incidents. Although the largest group consisted of relatively low time

pilots (19 reported less than 500 hours), as Table 5 shows, there were 14

pilots with more than 1000 hours. This raises some questions as to the role

of experience in avoiding VFRflight in adverse weather, and whether more

experienced pilots are willing to take more chances. Perhaps kinds of

experience, attitude, and judgement all interact to contribute to the pilot's

involvement. It does not appear that experience alone, as indicated by

reported flight hours, is the major contributing factor.

TABLE5. REPORTED EXPERIENCE OF PILOTS

INVOLVED IN VFRFLIGHTIN IMC WEATHER

[Note: One pilot reported I year of experience]

Flight Time Incidents Flight Time Incidents

3000 Hours or More 8

200 - 299 Hours

1000 - 2999 Hours

6 i00 - 199 Hours

500 - 999 Hours

99 Hours or Less 3

300 - 499 Hours 3

Unknown

Total - 52 Tables 6 and 7 show the weather conditions at the time of the incident as related by the reporter. In most of the incidents the weather was reported as IMC or mixed. Only one incident was reported as VFR and that involved a situation of reduced visibility created by rain showers. Fog and/or rain was present in 16 of the reported incidents, and there was a bro- ken or solid overcast in 26 of the situations. Visibility of less than 3 miles was reported in 15 incidents and in 2 incidents the visibility was less than one mile.

What led these pilots into weather encounters? The obvious initial question relates to their preflight planning. Did they check the current and forecast weather for their departure, destination, and enroute stations?

TABLE6. REPORTED FLIGHTCONDITIONS AT THE

TIME OF THEWEATHER INCIDENT

TABLE7. REPORTED WEATHER CONDITIONS AT TIME OF INCIDENT

Obscuration Sky Cover Visibility Overcast 20 7 Miles 1 Fog 8 Rain 8 Broken 6 3 Miles I Obscured 3 Haze/Smog 4 2 I/2 Miles 2 Snow 3 Other 2 2 Miles 2 r Unknown 29 Scattered 2 1 I/2 Miles 5 Unknown 19 1 Mile 4 1/2 Mile or Less 2 Unknown 35 52 52 52 Table 8 shows the number of pilots reporting a preflight weather briefing and the forecast they received.

TABLE8. WHETHER PILOT RECEIVED A PRE-FLIGHT WEATHER

BRIEFING,SOURCE, ANDFORECAST WEATHER

r Pre-Flight Briefing Forecast Weather Yes: Flight Service 22 VFR 14 Yes: (Source Unknown) I0 Marginal VFR 6 Yes: (Military) 1 IMC 4 No 3 Fog 2 Unknown 16 Unknown 26 Total 52 Most pilots reported receiving a preflight weather briefing from a flight service station, while others reported receiving preflight weather, but did not mention the source. Probably most of these were also from flight service stations, although some may have used PATWAS, TWEB, TV weather, or other sources. Three pilots reported no weather briefing. Two of these pilots had flown earlier in the day and had experienced favorable VMC at that time. The third pilot took off to practice night landings at the local air- port and encountered ground fog.

Most of the pilots who commented reported the forecast weather as VMC.

However, it cannot be determined from the reports how much above VFR minimums the weather was forecast to be. Was it barely VMC, e.g., 1200 and 4 with a change to 800 and 2 in rain showers? In such a case, it would not require much weather deterioration for the pilot to find himself in IMC conditions.

If the forecast was for clear and unrestricted visibility, the pilot would certainly not expect weather problems. In any case, these pilots initiated their flights either because they expected only weather problems with which they could cope, or they were willing to take the risk of possible deteriora- tion because of "get homeitis" or other factors.

In 6 of the incidents the reported forecast weather was marginal VFR

conditions. Not muchdeterioration was needed for the pilot to find himself

in IFR conditions. Of the four pilots who took off in IMC conditions, one

requested a special VFR(SVFR) clearance, and three took off VFRin IMC con-

ditions.

It appears that many of the pilots involved in these incidents did check

the weather, but were surprised by unforecast weather or decided to take a

chance on the weather not deteriorating. Several pilots (12) reported

requesting weather updates in flight. No weather updates were indicated by 9

pilots, and in 24 of the incidents, it could not be determined whether or not

updated weather was requested. There were also 7 incidents in which weather

updates were not practical, since the pilot encountered IMC shortly after

take off.

As shown in Table 9 most of the weather encounters (32) occurred while

.... the pilot was enroute. The weather was encountered at the cruise flight

level (23) and often involved operating above a cloud deck (VFRon top - 12).

Eight encountered IMC conditions immediately after takeoff or climb out.

Several others did not encounter the weather problem until they arrived at

their destination airport and found that the airport was IMC. There were two

incidents in which the IMCwere not encountered until the pilot was approach-

ing to land. One of these involved a pilot who cancelled his IFR flight plan

and landed at the destination airport without realizing that the field was

IMC.

The pilots who reported the weather as underneath or encountered upon

descent were involved in situations in which ground fog or haze existed at

the destination airport.

The nature of the weather encounter is shown in Table I0. The most fre-

quent encounter was reduced visibility (20 incidents). Rain showers (6

incidents) and ground fog (5 incidents) also resulted in reduced visibility either enroute or when attempting to land.

TABLE9. FLIGHTPHASE ANDLEVELOF WEATHER ENCOUNTER

Level of

Flight Phase at

Weather Encounter Encounter

Cruise 23

Enroute 32

Takeoff and Climb 8

On Top 12

Destination I0 Climb Out 8

Underneath 3

Approach or Landing 2

Approach and

Landing 3

Descent 2

Not Applicable 2

Total 52

Several reporters (I0) indicated that clouds had closed beneath them and

two reported that they were between merging layers. Only 2 pilots reported

problems trying to stay below an overcast. There were three incidents

involving thunderstorms. One involved a pilot attempting to fly between two

cells when they closed; one involved a pilot encountering cells and making a

precautionary landing; and the other involved a pilot penetrating a res-

tricted area after becoming unsure of his position while deviating around

thunderstorms. Thunderstorms do not appear to constitute a frequent problem

factor in weather encounters involving VFRpilots. Thunderstorm activity may

appear so threatening to the VFRpilot, (e.g., dark sky and lightning) that

he immediately deviates or lands, and consequently avoids this weather condi-

tion. Of the five incidents of flight into clouds, two involved pilots who

were on SVFRclearances and encountered clouds on climb out, but proceeded to

climb through the layer to VFRon top. Two others involved deteriorating

weather in which the pilot continued while contacting ATCor FSS. The last

TABLEI0• NATURE OFWEATHER ENCOUNTER INDICATED BY REPORTERS*

Broken or Solid Undercast ....... I0

Rain 6

• • • • • • • • • • • • • • • • • • Saw or Heard of] *Totals over 52 since some pilots encountered more than one condition.

involved a pilot trying to stay underneath a lowering ceiling and, becoming concerned about obstructions, climbed into the clouds before contacting ATC.

The actions taken by the pilots upon encountering the adverse weather conditions are shown in Table Ii. During training VFR pilots are frequently advised of the value of a "180 degree turn" when encountering adverse weather• It is apparent that many of them waited too long to execute a 180 degree turn. In some cases, such as ground fog at the departure airport, a return to VMC was not a feasible option. As shown in Table II, the most fre- quent actions reported were contacting ATC (ii) or landing at an enroute air- port (9). Some pilots reported climbing or descending through cloud layers in an attempt to get out of the adverse weather condition• Incidents where the flight was reported as "continued" represent a variety of situations• They include requesting SVFR at destination airport,

TABLEii. REPORTED PILOT ACTION UPON WEATHER ENCOUNTER

Contacted ATC................ II

Contacted FSS • . • _ ........... _ 5

Made180 Degree Turn or Returned

Contacted ATCand Requested IFR Clearance . . 3

Cancelled IFR and Requested SVFRto

finding VFRon top, descending through a hole in the overcast, and encounter-

ing momentary patches of IMCconditions.

In most of the incidents it appears that the pilot waited too long

before taking action. Whenaction was taken, the pilot requested help from a

ground facility, landed at an enroute airport or, perhaps because of get

home-iris or a belief that things would get better, took steps to continue

the flight.

Most pilots (44) did not make a precautionary 180 degree turn when

encountering adverse weather. Table 12 shows reasons for their decisions.

TABLE 12. REASONS FOR NOT MAKING A RETURN TO VMC Unknown ............. I0 Requested IFR or Other .............. II In I0 of the incidents the reason could not be determined. There were several isolated reasons (II), such as lost student pilots, pilots not real- izing the airport was below VFR minimums, and breaking into VFR conditions.

The most frequent reasons were that the 180 degree turn was not feasible, due to adverse weather closing in or ground fog (7), landing at an enroute air- port (6), or requesting IFR or special VFR clearances (5). Apparently in these incidents the failure to make the recommended 180 degree turn was jus- tified by the situation or the pilot's action. Many of the pilots requested assistance from a ground facility when encountering the adverse weather con- dition. Analysis of the narratives indicated that 33 pilots requested assis- tance, 16 did not ask for help, and, in 3 incidents, it could not be deter- r, mined whether they requested assistance. Table 13 shows the type of service requested and provided.

In two-thirds of the requests for assistance, service was provided by the ground facility. The majority of the services consisted of radar vectors to the destination airport, an alternate airport, or VFR weather. In one- third of the requests, however, the pilot was unable to obtain assistance.

This was often due to inability to contact the desired facility (4 incidents) or the difficulty of obtaining an IFR or SVFR clearance (4 incidents). This latter situation may have been related to the controller's workload.

The recoveries or outcomes of these incidents are shown in Table 14.

The most frequent recoveries were the pilot landing at an enroute airport (I0), or receiving vectors to his destination or enroute airport (9). In three incidents the weather was such that the pilot had to be given an ASR or

TABLE13. WHETHER SERVICE PROVIDED ANDTYPEOF

SERVICE GIVENORDENIED

Service Provided Reason Service Not Provided

Vectors to Airport or Could Not Contact

Landing 12 ATCor FSS 4

Vectors to VFR

Request for IFR or

Weather or Deviations 4

SVFRDenied 4

IFR or SVFR

Too Low for Radar Coverage 1

Clearance Issued 2

EmergencyClimb or

Delay in Granting SVFR

Descent Through Clearance 1

Clouds 3

Weather Update 1 ATCWould Not Provide

Clearance

Total 22

II PAR approach. There were 6 incidents in which flight was continued to the destination airport because the pilot broke out into VMC or continued in the poor weather. The 7 incidents in which no recovery factor was indicated involved such things as pilots entering ATAs without clearance, landing at their destination below VFR minimums, and misunderstanding clearances.

In many of the incidents the situation or the pilot's action resulted in an unsafe condition, as shown in Table 15. The most frequent unsafe condi- tion was the non-instrument rated pilot flying IMC. In some cases the con- cern was whether the pilot could control the aircraft and follow the controller's instructions. In other incidents, however, the pilot was not in contact with ATC which, in addition to the aforementioned control hazard, posed a threat to IFR traffic. Six incidents involved pilots flying in or through clouds and fog, and not in contact with an ATC facility. Penetration of airport traffic areas, approach corridors, and other controlled airspace without clearance (7 incidents) also produced a threat to other aircraft, especially under conditions of restricted visibility.

TABLE14. RECOVERY FACTORS OF VFRPILOTS IN IMC CONDITIONS

Landed Enroute Airport ......... i0

TABLE 15. UNSAFE CONDITIONS CREATED BY VFR FLIGHT IN ADVERSE WEATHER ro VFR in IMC (In contact with ATSS) .......... I0 p VFR in Clouds or Fog Penetrated ATA or Other Controlled Airspace Without Clearance • • 7 Landed Without Reporting: Non-instrument Rated None ................... I0 In four incidents ATC assists were required for an ASR or PAR approach, or vectors to the runway. While not necessarily a hazard to other aircraft, the safety of the flight in such cases is contingent upon ability of the

pilot and the controller. There were several ineldents (8) in which no

threat to safety was involved. In several of these the pilots saw the

deteriorating weather and madea 180 degree turn or landed at an enroute air-

port when the weather started to deteriorate while the airport was still

under VMC. In two of these incidents the pilot heard on the radio frequency

he was monitoring that weather ahead was deteriorating and landed before

encountering the weather.

Somepilots (I0) reported taking off in IMC. Seven of these obtained

SVFR clearances. There seems to be some confusion about SVFR clearances

amongthe reporters. Somepilots, finding the weather below VFR minimums,

asked for a SVFRto depart and continued to their destination in IMC, or at

best, marginal VMC. (See the "Special Issues, Special VFRIncidents" subsec-

tion). Three pilots apparently departed under VFRin IMC.

There were only two pilots who reported poor or incomplete weather

briefings. In someof the other incidents the briefings may have been incom-

plete, but this is not apparent from the narratives. There were two

incidents in which the pilot did not check the weather. One of these

involved a pilot shooting night landings. The other involved a pilot who had

flown earlier when the weather was VMC,so he did not check weather for his

next flight.

In most of the incidents (31) the pilot was able to get himself out of

the situation by returning to VMC,landing at an enroute airport, or finding

VMC. In 16 incidents, however, the pilot required assistance from ATC or

flight service for landing or finding VMC.

Generalizations What insights can be gained from this examination of incidents involving VFR pilots in adverse weather? The typical scenario is as follows: the pilot checked the weather with flight service and the forecast was for VFR.

He took off and either encountered low visibility conditions or a cloud deck closing in beneath him. He continued in IMC or marginal IMC before electing

to execute a return to VMC. He then landed at an enroute airport or received

an assist from ATCfor vectors to his destination airport or an alternate.

Meteorology is not an exact science. Sometimes weather will be better

than forecast and sometimesit will be worse. Somepilots may place too much

faith in the forecast and believe that the deteriorating weather is just a

local condition and/or that good VFR weather is Just down the road. Even when the forecast is for good VFR a pilot should have an alternative course of action in case the weather turns sour.

There is need for pilots to become more familiar with the mechanics of weather observation, and the nature of the forecasts. To prepare for a safe flight, and to be aware of options available in case of an encounter with adverse weather, pilots must understand both the value and the limitations of weather information dissemination systems.

Another problem involves the issue of legal versus practical VFR. A pilot may legally fly in controlled airspace when the reported visibility is as low as three miles, yet he cannot see an airport until he is within the airport traffic area. In such situations the slant visibility may be even less than three miles. At this pont he may become a potential conflict with IFR traffic.

One of the most important issues is the need for more timely weather information to be available to pilots. More PIREPS should be encouraged, even in good weather, and especially for enroute areas where no weather reporting facilities are available. There needs to be better coordination between ATC and FSS for relaying weather information. Many PIREPS are reported to enroute, approach, and departure control facilities that may not be relayed to flight service for dissemination to pilots. Unfortunately, when the weather begins to deteriorate is when the controllers are the busi- est but also when such reports are most needed.

Pilots flying under VFR, especially on long cross-country flights, are often uncertain of the frequencies they should monitor in order to be alerted to deteriorating or other weather changes. Should they monitor Center, FSS,

or EFASfrequencies? Pilots should be better indoctrinated on the sources of

such information.

MULTIPLE AIRCRAFT WEATHER-RELATED ASRSREPORTS

Introduction One might expect that most reported events involving multiple aircraft would address potential conflicts. This expectation is upheld; 55 out of 65 reported incidents in the non-commercial weather data set being considered are classed as potential conflicts. (Due to multiple reporting, these 65 incidents are described in 77 separate ASRS reports).

A second expectation is that potential conflicts may be closely related to weather problems, due to reduced visibility and heavier pilot workloads.

That expectation is not supported by the data. The percentages of potential conflict and near mid-air collision reports contained in the weather-related subset are nearly identical to the percentages in the entire ASRS data base.

The weather factor does not appear to raise significantly the frequency of potential conflicts, as noted in Table 16.

TABLE 16. POTENTIAL CONFLICT FREQUENCY Total File (21,871) Weather Subset (175) I.....

PotentialConflicts 6977 (32 percent) 67 (38 percent) Potential Conflicts-SMA 3308 (15 percent) 34 (19 percent) Near Mid-Air Collisions 1799 ( 8 percent) 16 ( 9 percent) Near Mid-Air Collisions-SMA 1284 ( 6 percent) 12 ( 7 percent) If multiple reports of the same event are deleted, the potential con- flict weather subset percentage (34 percent) is even closer to the total data base percentage.

It is also interesting to note in Table 16 that small aircraft (SMA) do not appear to be any more heavily involved in potential conflicts in weather-related reports than they are with all reports of potential con- flicts.

Typical scenarios for multiple aircraft events differ significantly between those in which both aircraft are under ATC jurisdication and those in which only one or neither is talking to ATC. When both aircraft are on IFR flight plans, all participants are well identified and the details of the encounter are well documented. Encounters under mixed flight rule opera- tions, on the other hand, often depend upon the observation of a single reporter, who may be accusing others of misbehavior. The reported facts may not be verifiable by a second party and often appear subject to interpreta- tion. For these reasons, the following analyses separate multiple aircraft events into two sets: one, in which all aircraft are on IFR flight plans and /_ a second, in which one or more aircraft may be operating outside the IFR sys- w tem.

IFR/IFR Incidents Multiple aircraft events in the set of ASRS reports examined here in which all aircraft are operating under IFR flight plans are reported exclusively by controllers; such events often lead to multiple reports.

Among 34 ASRS reports in this category, there are only 23 separate events, all reported by controllers. Because this set of reports is the exclusive contribution of controllers, relatively little is known about pilot qualifi- cations. Of 47 aircraft involved, 10 are assumed to be commanded by an ATP rated pilot and 10 others are commanded by pilots known to hold instrument ratings. The qualifications of the remaining 27 pilots are not reported.

Furthermore, no information is reported about any of the 47 pilots" total flying time or recent flight experience.

The aircraft type involved in IFR/IFR incidents is heavily dominated by

small transports (SMT) with 23 of 47, or nearly 50 percent, being in that

class (SMT-FAR135 limit, 5000-14,000 Ibs.).

Small aircraft (SMA)involvement is minimal with only 2 reported among

the 47 (SMA-less than 5000 Ibs.). The remaining aircraft are scattered among

the heavier classes and are most often air carrier type equipment.

Weather Factors

In spite of the fact that all aircraft involved in this subset of events

were on IFR flight plans, only 12 of the 23 events (52 percent) occurred in

instrument meteorological conditions (IMC). Six of the 23 events (26 per-

cent) actually occurred in good visibility with the remainder in marginal

visual meteorological conditions.

There is no evidence that these events were precipitated by flight

crews' inability to deal effectively with weather problems. However, there

is evidence that these events may be closely related to failure of the total

system to deal effectively with at least one major weather problem, namely

thunderstorms. Of the 23 events, 12 involve thunderstorm encounters. Such

encounters typically lead to weather diversion maneuvers, (i.e., diversion

from flight plan route) which, in turn, culminate in lack of separation

through lack of controller communication or coordination.

The domination of thunderstorm activity in these reports is reflected in

two other statistics: time of day and month of year. Seventeen of the 23

events (74 percent) occurred between 1200 and 1800, and 15 of the 23 (65 per-

cent) occurred during the summermonths of June, July, and August. These

patterns differ significantly from the total ASRSdata base, but are con-

sistent with the heavy involvement of thunderstorms.

Problem Nature

One report narrative in the IFR/IFR subset is paraphrased as follows: "Air Carrier Aircraft A, inbound to the terminal via J60,

was deviating right of course for weather. Whenhe could

finally return to course, coordination was accomplished

with the next sector for a delay vector for sequencing.

This put A's projected path in conflict with GA Aircraft

B, descending through the altitude of A. B was told to

stop above A, but the controller transferred communica-

tions to the next sector without listening for a readback

from B. B continued to descend and passed through the

altitude of A with approximately one mile lateral separa-

tion. Traffic was heavy and complex with sequencing of

inbound aircraft, which were deviating from course due to

weather."

This narrative contains several of the problem elements evident

throughout the IFR/IFR subset. Weather deviations, heavy workloads, communi-

cation difficulties, controller coordination, and mixed operating phases

(cruise by one aircraft and descent by the other in this case) all play a

part in this narrative. Not every narrative is so rich in detail, but nearly

all have one or more of these sameproblem-causing elements.

The nature of the problem is potential conflict between two IFR air-

craft. Seventeen of 23 events are coded as potential conflicts. Five of the

remaining 6 are saved from that designation only because the aircraft

involved happened not to violate less-than-standard separation requirements,

even though the conditions were such that only by chance (or sometimes by

controller vigilance) was it prevented. These intrusions into occupied air-

space most often occur during mixed operating phases, e.g., one aircraft

cruising, while the second descends or climbs. Seventeen of 23 events occur

in such mixed phases. It is also worth noting that only one occurs when both

aircraft are in cruise phase and four occur when both are climbing. Although

13 cases (57 percent) occur between i000 and I0,000 ft. altitude, a signifi-

cant number 8 (35 percent) take place above i0,000 feet. The events are

split nearly equally in terms of the responsible ATC controlling agency.

Eleven occur while under the jurisdiction of an air route traffic control

center, and 12 while controlled by a TRACON or tower facility.

Problem Causes

The generation of nearly every IFR/IFR event reported begins with some

combination of the following three factors: I. Weather diversions by one or more aircraft (56 percent).

2. Difficulties in coordinating handoffs between controllers

(70 percent).

3. Violation of ATCclearances by flight crews (49 percent).

Indeed, as the following paraphrased report indicates, all three factors

may be present at once.

"Aircraft A was climbing to FL280; B was climbing to

FL350 out of 290 under control of ZDC. A few minutes

later, B was handed off to ZID climbing to 350 on J149.

At this point, B was 5 west of EKNVORand A was over EKN

VOR. Aircraft A was cleared to FL350. A rogered and

climbed. ZOB then called and said stop A at FL310.

Meanwhile, ZID controller turned B to deviate around

weather without advising ZDC. Transmissions to stop A at

FL310went unanswered. At a point 20 north of EKN, A

called ZDCand was told to descend immediately to FL310.

About 20 seconds later A reports level at FL350. These

aircraft were together 3 or 4 times before being resolved

by ZOBand ZDC."

j_

This report noted ATCas the primary problem code, even though it shows

elements of both ATCand FLC (flight crew) shortcomings. It all began when

one controller authorized a weather deviation for an aircraft near the boun-

dary of three different sectors without advising the other controllers. The

problem was then complicated by communication difficulties, violation of

altitude restrictions, and an uncoordinated climb. Weather diversion, con-

troller coordination, and violation of ATCclearances all contributed to this

potential conflict.

narratives with the The three causes are mentioned in the ASRS report following frequencies: Weather Diversions 13 (56 percent) Controller Coordination - 16 (70 percent) Clearance Violations I! (49 percent)

However, as noted in the report above, it is often the combination of

these enabling factors which leads to a reportable event. In the present set

of 13 weather diversion cases, i! (85 percent) involve controller coordina-

tion difficulties. This indicates that ASRSreports are often generated, not

so much because of the weather, but rather because of controller coordination

problems arising from attempts to accommodatediversions from flight planned

routes.

Recovery Factors

The 23 reports under discussion all related to potential disasters.

Pure chance appears to have saved at least 9 in which no recovery factor was

noted. In 12 of the remaining cases, controller intervention resolved the problem. Pilot vigilance was responsible for recovery in two cases. There was some evidence in the narratives that auxiliary warning systems, which supplement radar tracking, play an important role in recovery. The activa- tion of a conflict alert alarm was mentioned in 6 cases as the means by which controllers became aware that there were potential problems.

Generalizations The multiple IFR aircraft reports stem from potential conflicts between small transport and larger aircraft. Small aircraft do not play a signifi- cant role. Problems are created by weather diversions, complicated con- troller coordination procedures, and misunderstandings of ATC clearances.

Mixed Flight Rule Events Although 35 of 42 reports (83 percent) involve one IFR aircraft, multi- ple aircraft events in which one or more of the participants are not operat- ing under IFR flight plans differ decidedly from the previously discussed IFR/IFR events. The reporter is more likely to be a pilot than a controller (55 percent pilots, 45 percent controllers), whereas in the IFR/IFR case all reporters are controllers. Often, qualifications of only one pilot are reported (27 out of 42 pilots are identified by rating). Of the 22 out of 27 3O

pilots who report flying experience, as well as ratings, 17 (77 percent)

claim more more than 1500 hours total flying time, and most of them claim

Weknow nothing of the qualifications of

over 30 hours in the last 90 days.

the non-reportlng, unidentified pilots, who are often the targets of the

reports.

The small aircraft types (SMA)tend to dominate mixed flight rule events

with 53 (63 percent) of the 84 aircraft reported being SMA. This is markedly

different from the IFR/IFR case, in which only 4 percent were SMA. Sixteen

small transports (19 percent) are reported, with the remaining aircraft dis-

tributed amongthe heavier classes of equipment.

Weather Factors

A high proportion of the mixed flight rule events were reported to have

occurred in instrument meteorological conditions (IMC): 26 out of 42 (62

percent). Since only 3 (7 percent) were reported to be in VMC, the total

marginal weather plus IMC reports account for 93 percent of the reported

events. This contrasts in an unexpected way with the IFR/IFR events, in

which 26 percent were reported to have occurred under visual meteorological

conditions (VMC).

Contrary to the predominanceof thunderstorm activity in the IFR/IFR

case, the mixed flight rule events show no particular in-flight weather

encounters beyond clouds and precipitation. Thirty-eight out of 42 (90 per-

cent) of the events list no in-flight encounter, and only 3 (7 percent) are

related to thunderstorm activity. The most common obscuration factor is

"fog" with 19 reports (45 percent). This classification includes both ground

fog and cloud encounters. Rain or snow are reported 8 times (19 percent) and

haze is noted 5 times (12 percent). Flight visibility of less than 3 miles

is noted in 14 reports (33 percent). Ceiling of less than I000 feet is noted

8 times (19 percent).

One troublesome aspect of the reported weather statistics is that they

often originate with an IFR pilot who has encountered another aircraft while

flying in conditions which he, the IFR pilot, judges to be below VFR

minimums. In somecases, this is a judgement call, with which the non-

reporting pilot might not agree. In marginal weather, it seemspossible for

two persons of good faith to have honest differences of opinion, with one

Judging conditions to be IMC while the other calls it VMC. The following

paraphrased narrative taken from the ASRS data illustrates the point as

reported by the pilot of aircraft A.

"Aircraft A was northbound on V-23 on an IFR flight plan, flying in snow with a visibility of less than a mile.

Aircraft B passed to my right about 1/4 mile away -- not

at a VFRaltitude, VFRin weather that was definitely IMC

-- only chance prevented a midair. The pilot of Aircraft

B was obviously at fault."

The reporting pilot (IFR) in this case, had 400 hours total flying time.

The non-reporting VFRpilot was also in contact with ATCand was reported to

have delayed an expected climb from the altitude in question in order to

remain in VMC. Lacking corroborative reports from either the controller or

the VFRpilot, there seemsto be room to question whether the conditions were

as poor or the encounter as close as the relatively inexperienced IFR pilot

estimated.

Discounting those events in which an IFR pilot is reporting on other

traffic which he judges to be in violation, there is still a group (9 in 42 -

21 percent) of events which seemto point to failure of pilots to deal effec-

tively with a weather problem. For the most part these involve VFRpilots

proceeding into weather beyond their capabilities and at some point

encountering another aircraft or causing traffic disruption. The following

narrative illustrates this point.

"VFR aircraft in the control zone (IMC) was told to

depart to south. Turned northwest and crossed in front

of IFR aircraft on localizer. IFR aircraft had to be

pulled out of approach. VFRaircraft asked for vectors

back to airport. Tower was unable due to no radar. VFR

finally landed with VFR controller's assistance. IFR

aircraft was delayed.

Problem Nature

The nature of the mixed flight rule events is again that of potential

conflict. Thirty-eight of 42 events (90 percent) are so classified.

Twenty-nine (69 percent) of the events occur in the vicinity of airports with

the remainder (31 percent) occurring during the enroute phase. Thirty-five

(83 percent) of the reports involve an encounter between an IFR and other

(presumed to be VFR) aircraft. Four more (10 percent) involved special VFR

(SVFR) and IFR aircraft encounters for a total of 93 percent of the reported

events occurring when at least one aircraft is under direct control of ATC.

Ninety-five percent of the reported events occur in daylight hours.

The role of the small aircraft (SMA) in mixed flight encounters is

highlighted by Table 17. Note that not only do most of the encounters

involve at least one SMA(83 percent), but also that 40 percent of the events

involve two SMA's.

°.

TABLE17. AIRCRAFT TYPESIN MIXEDFLIGHTPLANENCOUNTERS

SMA SMT OTHER Total (Percent) (Percent) (Percent) (Percent) SMA 17 (40) 8 (19) i0 (24) 35 (83) m SMT 2 (05) 4 (I0) 6 (14) OTHER m 1 (02) 1 (02) An analysis of the narratives does not yield an overwhelming majority for the "typical" scenario, as it did in the IFR/IFR case. Rather, mixed flight rule events seem to stem from a variety of problems. If there are common denominators, they must be: I. Entering control zone without clearance when the weather is below VFR minimums, 17 events (40 percent).

.

VFR flight in (reported) IMC, 23 events (55 percent).

3.

Surprise encounters with aircraft popping through holes in clouds, maneuvering around broken clouds, or cruising near cloud decks being penetrated by IFR aircraft, 12 events (29 percent).

A given scenario may involve one or more of these factors.

Again, the statistics on IMC identification and estimation of encounter distances must be tempered by the knowledge that they originate from single, possibly biased, reporters.

As was the case with the IFR/IFR events, the mixed flight rule encounters most often occurred during mixed flight phases; 24 out of 42 events (57 percent). An additional 11 (26 percent) occurred with both air- craft in the traffic pattern or on approach. Only 4 events (I0 percent) occurred when both aircraft were in cruise mode.

The mixed flight rule events are primarily low altitude phenomonon, with I0 reported under 1000 feet and 26 between 1000 and I0,000 for a total of 86 percent of all reports. The controlling agency is also reflected in the altitude structure, with ARTCC involvement in only 8 of 42 cases (19 per- cent). The other com_unlcation facilities are spread among TRACON, tower, FSS and UNICOM.

Contributing Factors An examination of the narratives reveals the presence of one or more of a variety of contributing factors: I. Weather interpretation - 21 events (50 percent).

2. Controller coordination - I0 events (24 percent).

3. Regulation knowledge and interpretation - I0 events (24 percent).

4. Clearance violations - 9 events (21 percent).

5. Weather encounters beyond a pilot's capabilities - 9 events (21 percent).

As previously suggested, the outcomes of such contributing factors are

to cause a pilot to enter a control zone when the weather is below VFR

minimums without clearance (40 percent), to cause a second pilot to be

surprised by sudden encounters with another aircraft close to clouds (29 per-

cent), or to cause a VFRflight to be operating in IMC (55 percent).

Weather interpretation causes ASRSreports to be filed for the following

reasons:

I. A VFRpilot may misjudge visibilities and actually enter

IMC.

2. An IFR pilot may misjudge visibilities and report an

illegal VFRoperation when, in fact, VMCdoes exist.

3. A VFRpilot may misjudge cloud clearance distances and

actually be too close to clouds.

4. An IFR pilot may misjudge cloud clearance distances and

report an offending VFR operation which is, in fact, legal.

5. Ground observers and a VFRpilot may disagree on prevail-

ing ceiling and visibility, causing a VFRflight to enter

a control zone in reported IMC. For example, the visi-

bility value which is reported is that which prevails

over 51 percent of the 360-degree horizon visible from

the observing point. Thus, an airport may be below VFR

minimumsdue to low visibility values over 51 percent of

the area. A pilot, however, may be approaching the field

from a quadrant where the visibility is better, have the

airport in sight, and does not realize that the field is

IMC. He consequently proceeds to land without concern

for "legality".

Unfortunately, the weather interpretation factor is one in which con-

flicting opinions can only be inferred and, most often, are not available for

explicit analysis. The suspicion that conditions may not be exactly as

reported is raised when a pilot reports in one narrative, "... I was on solid instruments at that altitude", yet advises radar that he has the traffic. It seems plausible that the pilot of the offending VFR aircraft might have believed conditions were VMC and was following visual separation procedures.

Controller/pilot communication factors may appear when adjacent facili- ties have different reported weather. A VFR aircraft may be cleared (legally) out of one jurisdiction, only to enter a second which is reporting IMC and ultimately causes a potential conflict. The following synopsis involving a VFR departure which entered a neighboring control zone reported to be below VFR minimums illustrates such a problem: "LAX was VMC, HHR, adjoining airport IMC. Helicopter A departed LAX VFR. Later light aircraft on approach to HHR reported NMAC with A. A was outside TCA and no longer in contact with LAX tower."

Knowledge and interpretation of regulations may be factors in potential conflicts occurring in control zones which are not served by towers. Con- sider the following excerpts from a narrative, in which the field had a flight service station, but no control tower: "Aircraft A broke out on an ILS approach Just above minimums and observed aircraft landing on same runway.

--He was not in communication with controlling authority in a control zone that was IMC at the time."

_LI" _ The offending VFR pilot was illegally in the control zone, which was below VFR minimums. He did not check with the FSS; either he was not radio equipped or he did not know that a clearance was required when the visibility was less than 3 miles. No communication would have been required if the con- trol zone was in fact above VFR minimums.

A more subtle regulation factor occurs when an IFR approach is made to zj_ an uncontrolled airport not in a control zone at which local operations are being conducted below 700 feet in one mile visibility. Since, at many loca- tions, that is uncontrolled airspace, the local VFR flights may be legal, but may be a surprise to the arriving IFR pilot who breaks out at 700 feet to face another aircraft in the pattern as noted in this narrative: "With a ceiling of 700 ft. A GA Aircraft B was practicing pattern work at what he said was 500 ft. AGL. I consider it unsafe when we break out right at minimums on the cir- cling approach, switch frequencies from approach to UNICOM and hear B say he will do a 360 on crosswind to avoid the jet. -- We all could have been dead right."

Clearance violations are contributing factors whenever failure to follow instructions causes one aircraft to come in close proximity to another. The

most common violation (5 events - 12 percent) seems to be for a VFRpilot to

change altitude or direction while in contact with ATCin marginal weather

without announcing his intentions.

Weather encounters beyond a pilot's capability include those in which a pilot loses separation from clouds and requires special handling to complete his flight. These become mixed flight rule events only when other flights are delayed or diverted while the problem is resolved.

Recovery Factors Pure chance, as evidenced by no recovery factor, appears to be the only reason that 17 of the 42 mixed flight rule events (40 percent) did not lead from potential to actual conflict. Pilots initiated recovery actions in 13 (31 percent) cases. These were most often self-initiated evasive maneuvers.

In the remaining 12 cases (29 percent), controller vigilance played a part in resolving the mixed flight plan problems.

The majority of the mixed flight rule reports involve small aircraft operations (SMA). The most numerous reports involve VFR flight in reported IMC conditions and are filed by IFR pilots. Entering control zones without proper clearance in IMC conditions and operating VFR in close proximity to clouds, are also frequently mentioned. Weather interpretation, knowledge of regulations, controller coordination and clearance violations are factors which lead to the reported problems. A relatively small number of events occur because of obvious inability of pilots to cope with weather encounters.

The data suggest the problem is more serious in MXD weather and when one aircraft is climbing or descending.

WEATHER INFORMATION SERVICES There were 14 complaints about weather information services other than inaccurate forecasts where the pilot encountered IMC weather. The inability to reach flight service by telephone was the complaint in 5 of the incidents.

Reporters relate waits as long as 20 to 45 minutes on hold and then being

disconnected. This problem appears to be more prevalent in busy terminal

areas. Reporters allege that, because of the inability to reach flight ser-

vice, many pilots in their area depart without preflight weather briefings or

take off and contact enroute flight advisory service as illustrated in this

narrative:

"In this area it is practically impossible to obtain a

weather briefing by dialing the local phone numbers for

the FSS. A phone call to the local FSS at any time of

the day, morning or night, and in any kind of weather,

good or bad, will produce the sameresult of receiving a

recording which says that all lines are busy and a wait

of approximately 20 to 25 minutes. On more than one

occasion I have waited 30 to 45 minutes, only to have the

equipment disconnect my call, thereby causing me to re-

dial. My experience is also the experience of other peo-

ple in this area which I believe represents a dangerous

hazard to both VFRand IFR flight due to the inability of

the pilot to obtain a weather briefing. I have heard

pilots inform me that on more than one occasion flights

have been planned without a weather briefing because of

the inability to obtain a briefing on the ground. More

than one pilot has informed me that he will take off and

attempt to get flight watch on 122.0 in an attempt to

obtain a weather briefing to his destination. Simply

passing a law requiring a pilot to obtain a weather

briefing prior to flight and then making it impossible

for him to obtain a weather briefing does not improve

safety of flight and in fact encourages what eventually

must be an illegal operation by the pilot and which

represents a dangerous condition for all. From my own

personal observation the quality of the weather report

received from the FSS personnel is spotty and incon-

sistent with the experienced people giving you an excel-

lent report and the inexperienced people giving you a

broad brush that for all practical purposes is valueless.

• r

Whenthe weather is good, everyone gives you a good

report, but when the weather is bad I have found it is

always necessary to talk directly to the weather person-

nel. I have attended many safety seminars put on by the

FAAwho are aware of the problem and have suggested sub-

mitting these danger reports to NAA for their evaluation.

One pilot reported a systems problem where, after waiting on hold, the

line disconnected when the briefer answered. This appeared to be a recurrent

problem at this particular FSS. Three pilots complained of pessimistic fore-

\

casts issued by flight service. Two of these pilots cancelled flights

because of the forecasts when, in fact, the weather turned out to be VMC.

The other pilot took off in spite of the pessimistic forecast and completed

his flight without incident. In this later case, however, the forecast was

substantially correct and the pilot was fortunate enough to find breaks in

the overcast to climb to VFRon top and descend at his destination. There

were three complaints of incomplete briefings or lack of information. These

complaints involved lack of charts and radar summaries at self-brlefing

desks, automated systems being down, and lack of PIREP information. One

pilot suggested that, at terminal areas with approach and departure control,

every hour a departing aircraft, either GAor air carrier, be designated as a

PIREP aircraft. This aircraft could report cloud bases, tops, freezing

level, turbulence, etc. The information could then be relayed to flight ser-

vice, entered into the computer, and dispersed to FSSthroughout the system.

This could be effective in poor or rapidly changing weather conditions.

There were also three complaints of poor attitudes on the part of FSS or

NWS personnel. One pilot reported that NWS personnel refused to provide

weather information and even stood in front of the radar so that the pilot

could not see it.

The other complaints involved an inaccurate forecast and lack of weather

reporting facilities at the airport.

r_ WEATHER-RELATED INCIDENTS INVOLVING BALLOONS Five incidents involved hot-alr balloons. Four of these occurred because the winds aloft differed from those forecast. One pilot landed to avoid penetrating a TCA; another climbed to altitudes of 16,000-17,500 feet without oxygen and crossed airways to overfly forested areas. The third pilot flew at I00 feet AGL over a housing development to take advantage of surface winds. The last pilot drifted into an airport traffic area and was becalmed over the airport. All pilots and especially balloon pilots should recognize that the winds aloft forecast is a computer generated forecast based on the prevailing and projected pattern of high and low pressure areas

and will often be different from actual conditions. While in most cases this

does not pose a problem for the pilot of a powered aircraft, who maymiss his

ETAby a few minutes, it can be a major problem for balloon pilots.

LIMITATIONS TO THE DATA In the foregoing analysis it should be recognized that there are limita- tions to the conclusions drawn from the ASRS data. Clearly, there is no evi- dence on the representativeness of the data. For example: how many mixed VFR/IFR potential conflicts go unreported? Are these unreported conflicts substantially different in terms of inflight weather, mission phase, type of aircraft, etc.?

It is well known that General Aviation pilots are not the usual report- ers of incidents. General Aviation pilots comprise more than 90 percent of all pilots in the USA and account for 75 percent of the total hours flown and 44 percent of the instrument approaches; yet they figure in only 29 percent of the reported incidents (where GA operations include pleasure, corporate, training, and personal business). Does this suggest that fewer GA pilots are involved in incidents? The answer is probably that air carrier pilots are motivated and encouraged to take an active role in the ASRS system in the interest of increasing safety, they have ready access to reporting forms, and probably are more likely to "cover" their performance inadequacies by an ASRS report to avoid disciplinary action from the FAA. Incidents where pilots confess to altitude deviations probably belong to this category. The same might be said for air traffic controller reports of LTSS of two IFR aircraft.

While this may bias the data, the data still provide the analyst with detailed descriptions of system problems.

Attitudes of pilots toward controllers (and vice versa), of air carrier pilots toward General Aviation pilots (and vice versa), and of IFR pilots toward VFR pilots (and vice versa) can be seen as underlying factors behind some of the incidents reported. Again, this bias is important in understand- ing the development and consequential effects of incidents. An example of this is the following narrative:

"I was on V233 tracking from the Roberts VORto Knox at

5000 ft and was with Chicago center. My first indication of any weather was when I s_W l!BhtnlnB flash several miles ahead and began hearing airli1._ &_kln_ t_ _ivert ..... left of course. I was told to climb to 7000 rignL _ feet and ! asked the controller how the weather looked along my line 6_ _li_nt and was told that I would pass to the right of the storm. In a few minutes I saw that my line of flight was taking me into the area of intense lightning. I then asked the man for an update of my position in relation to the storm. He then very sarcast- ically told me 'I told you the weather was to your left, divert to the right if you want to.' I immediately turned right about 30 degs and was then handed off to another frequency. Explained my situation and asked for assistance. He was very helpful and diverted me farther to the right to miss still another cell. The point of this report is two-fold: No. l-please weed out sarcasm in ATC. No. 2-please--let's be assured of some help from the ground in weather situations. My flight originated in Fort Worth, TX, where I checked with FSS twice on weather conditions and once in flight with Decatur, IIi radio thru the VOR. There was no report of any weather, so I didn't knowingly get into the situation."

A basic issue is the relation of exposure frequency to incident fre- quency. In other words, before one becomes awestruck by a statistic that suggests that icing is a factor in IFR incidents, one must know the number of pilot exposures to icing which were incident free. Thus, in drawing conclu- sions from these data, one must recognize that, for the most part, there does not exist sufficient exposure data to permit comparison of one incident type to another.

Finally, all the incidents reported herein were selected because the ASRS analyst reported that weather was a significant factor in the incident.

Indeed, in the case of the single IFR aircraft incidents, it has been argued that weather should not have been a factor in a large percentage of these cases because it was not unusual, i.e., it was IMC which the IFR system was designed to handle. If the analyst can err on over-emphaslzlng the impor- tance of weather, he can obviously also under-emphasize. In any event, this source of error was probably trivial in comparison to the "exposure problem" or the "representativeness problem" or to the "bias problem."

v_t, d_spite the inability to know the true representativeness of the

ASRS data, it is evident that valuable insight into pilot-controller problems has been gained.

SPECIAL ISSUES Role of Increased Workload in General Aviation Weather Incidents One of the consequences of both deteriorating weather and sudden weather changes is the impact on both pilot and controller workloads. The ASRS analyst codes list excessive workload as a behavior factor which contributes to some incidents. It was anticipated that this factor would be a signifi- cant contributor in the reported General Aviation weather-related incidents, yet there were few cases where this was indicated. In only 6 of the 175 reports was workload cited, and half of these involved air traffic controll- ers. The problem might be the analyst interpretation of workload. One might argue that most VFR pilots (at least low-time pilots) experience increased workload with flight into adverse weather. It is difficult to ascertain how excesslve such workloads might be, however. The VFR pilot in IMC usually gets special handling, i.e., altitude and heading instructions from ATC.

To argue workload for the IFR pilot in IMC would seem to contradict the idea of an instrument rating. A "good" pilot is expected to handle naviga- tion, communication and maintain straight and level flight at the same time.

Icing, admittedly, could lead to stress which could enhance workload. Nei- ther of the infllght icing incidents suggests any workload factor, however.

The fact that the small general aviation aircraft has, typically, less sophisticated avionics, e.g., auto pilots, may be counteracted by the fact that the pilot does not have the schedule pressures of air taxi, air carrier crews, or corporate operations associated with more sophisticated aircraft.

The more likely case for workload-lnduced stress resides in the IFR/IFR cases where weather-lnduced diversions lead to controller workload and coor- dination problems between controllers. In summary, stress, workload, fatigue, and related concepts are difficult to ascertain after the fact.

Pilots and controllers may be reluctant to admit such behavioral factors

because they "suggest" an unprofessional pilot or controller. Perhaps no

post hoe reporting system can ever probe the degree to which weather-lnduced

stress influences a flight.

Problems With Obtaining Clearances in Flight The results of this search of General Aviation pilots and weather- related incidents revealed several instances of instrument-rated pilots reporting difficulties in obtaining clearances in the air. Thls is illus- trate_ by the following narrative: :'After takeoff, ACFT A was still within the airport boun- dary. FgS advised that the airport was now below VFR minimums. I asked for a clearance but was told they didn't have the zone, contact AN Center 119.09, center was contacted and advised I was airborne and needed a clearance. Center would not issue a clearance of any kind (depart, hold, land) because there was an IFR Acft B inbound. Again I advised I was airborne and needed to do something so was circling the airport maintaining VFR awaiting a clearance. Several times I asked for a clear- ance, whereupon _TC Anchorage center gave me an IFR clearance of "get the hell out of there." I continued to circle the airport (at 800 ft with 30 miles visibility) until I was given a proper ATC clearance. Time involved approx 25 mln."

Often IFR rated pilots will go up and "take a VFR look" at marginal weather, and assume that they can always get an IFR clearance if it appears that they will need one. They seem surprised that controllers are often too busy to oblige them immediately. Thls stems from lack of understanding of three considerations: i. If an airport becomes MVMC or turns IMC there will prob- ably be an increase in IFR traffic.

2. The controller's primary job is to provide separation of IFR traffic.

3. If the controller is already overloaded with separation responsibilities, he may be loathe to accept new respon- sibilities (at least, immediately).

The upshot of thls situation is the case of a frustrated IFR rated pilot try-

ing to maintain VFR flight in rapidly deteriorating weather conditions in controlled airspace. By not being under positive control, he then becomes a potential hazard for other traffic.

One solution would be to file IFR even when weather is still reported VMC -- a policy adopted by many pilots. This practice could, of course, over- load the IFR system and, hence, create additional problems. A balanced approach would be to let each pilot decide on the basis of weather and traffic in his area of operation. If he can hold VFR above a cloud deck for an IFR clearance, there is little problem. The problem develops when he is faced with visibility interpretations on arrival or on take off. Here the dilemma of two versus three miles visibility estimation appears. Because of the lack of obvious solutions to this problem, it would appear to be a candi- date for systematic research.

Weather Incidents at Night One of the concerns of aviation safety specialists is the combined effect of weather and night operations, particularly for the VFR pilot. Mar- ginal VFR in daylight is considered by some to be IMC at night where there is a paucity of visual cues for the pilot. Even scattered layers may cause air- ports to go un-notlced at night. Examination of the 177 ASPS reports reveal that only 12 incidents were under night conditions, and for VFR into IMC only 4 incidents or 8 percent of that type were reported as occurring at night.

This does not suggest that the night weather problem does not exist; it may only suggest that such cases are not being reported. Lack of such reports may be due to a combination of factors such as: a) Night flying, especially VFR, is much less frequent than daytime flying (the exposure issue).

b) Pilots who fly at night may have more experience and more sophisticated aircraft, and are less likely to find them- selves in a weather-related incident, or better able to cope with weather.

c) The air traffic control system is less crowded and is better able to accept devlatons without system disruption at night.

d) Pilots, in general, demand better weather for night

operation than for daytime operation.

Special VFR Incidents Eight incidents involved special VFR clearances. In several cases, upon flight continuance, the pilot discovered deteriorating rather than improving weather. Some "specials" were filed to reach satellite airports. The fol- lowing incident illustrates one kind of problem resulting from SVFR clear- ances: "I received permission to depart Wichita Falls special VFR. During climb out (at approx 500-600") I could see at a distance of one mile in front of me. The ceiling going down--I immediately decided this was not for me and decided upon a 180 deg turn back to the airport. How- ever, due to commercial traffic, I decided that the tower should be notified of my decision. Before initiating the turn, I radioed the tower twice, with no response (radios were set up correctly). Next contact attempt was to departure with the same results. Felt an emergency (lost comm) was necessary and continued a 500" climb into clouds as opposed to 180 deg turn or flying lower. A commercial jet was preparing to depart when I departed.

At 3500 MSL I broke on top and 2-3 minutes later Wichita Falls transmitted to me (no apparent problems now). I felt the climb was the safest choice -- also due for IFR checkrlde on (date). Continued flight to Dallas with no problems."

Pilots need to have a better understanding of SVFR. Perhaps the pres- ence of VOR receivers leads pilots to undertake trips in such weather condi- tions where pilotage would be impossible.

Whether special VFR's lead VFR pilots into problems excessively cannot be answered with these data without knowledge of successful VFR pilot use of SVFR clearances. For example, how many SVFR clearances are issued each year, and under what conditions -- localized ground fog or general low visibility conditions? How many pilots return to airports immediately after SVFR depar- tures with special assistance from controllers? How many special VFR clear- ances are denied and for what reasons? Clearly, SVFR clearances can invite some pilots into unsafe conditions. The data suggest that research be done

in this area to ascertain the value of SVFR, the percent of time difficulties

result, and the comparison between the benefits versus the costs of this pro-

cedure to the aviation system.

Weather Involvement and General Aviation Operations in the Vicinity of Airports Operations conducted in the vicinity of airports provide many opportuni- ties to generate potential conflicts among aircraft. The airports are col- lection points for arriving aircraft and dispersion points for departing air- craft. This natural opportunity for conflict is increased when IFR and VFR operations are mixed during periods of marginal weather. Three factors con- tribute to enhanced conflict potential: i. Operating rules, involving weather minimums and communi- cation requirements, differ among airports.

make 2. Ceiling and visibility estimates are difficult to and may differ among observers.

with 3. The transition from VFR to IFR operation may occur no time or space buffers.

Table 18 summarizes the airspace types in 27 reports of mixed flight rule events which originate in the vicinity of airports.

l iN Instrument approaches may be available at four different classes of air- ports. The first class includes the major terminals protected by TCA (Termi- nal Control Area) airspace. These areas are well charted and the equipment and operating requirements to enter that airspace are widely publicized. ATC authorization is required to operate in a TCA, regardless of weather condi- tions. Conflicts in these areas are most likely to be caused by clearance violations, intruders, or controller errors.

The second class includes those airports with operating control towers, but no TCA. Again, the airport traffic area (a cylinder of airspace 5 miles in radius and 3000 feet high) should be well understood by all pilots. No one legally approaches the airport served by the control tower without com- municating with ATC.

TABLE18. ASRS REPORTS FORMIXEDFLIGHTRULE

TRAFFIC IN THEVICINITY OF AIRPORTS

Percent

Number of

of Total

Location

of Airports

(42 Reports)

Terminal Control Area

Airport Traffic Area 19

Control Zone

Airport in Uncontrolled Airspace 5 12

Total in Vicinity of Airports 64

The third class of airports consists of those lying in control zones,

but without operating control towers. This appears to be the group with the

most confusing operating rules. The stated weather minima for VFRoperations

in a control zone are I000 feet ceiling and 3 miles visibility. The confu-

sion arises because pilots are not required to communicate with ATC to

operate at these airports in VFRconditions. This leads the VFRpilot to

believe that he is flying in an uncontrolled environment. Unless there is a

flight service station on the field and the pilot is participating in the

airport advisory program, an arriving VFRaircraft could enter a control zone

which is officially below minimums, communicating with no one, in weather he

judges to be VMC. At the same time an IFR aircraft on approach may be

(falsely) counting on ATCto assure separation from all traffic, since he is

in controlled airspace that is officially below VFRminimums. The conflict

potential escalates.

The fourth class of airports consists of those with approved instrument

approaches, but no control towers or control zones. Controlled airspace at

these airports ends at 700 feet above ground level (AGL). Below 700 feet is uncontrolled airspace in which VFR requirements are 1 mile visibility and clear of the clouds. An approaching IFR aircraft flying in IMC moves from

separation by ATCfrom all other legal traffic to no protection, beyond

visual separation, as it descends below 700 feet. The potential for con-

fllct, with perfectly legal VFRoperations just below the ceiling, is obvi-

ous. Marion, Ohio is typical of such airports. Marion has a VORapproach

with a minimumdescent altitude of 490 feet AGL, yet controlled airspace

there ends at 700 feel AGL. There are no time or space buffers between the

IFR and VFRworlds at such locations.

Weather interpretation and non-unlform operating rules combine to

increase the potential conflict hazard in mixed IFR/VFR operations at air-

ports in classes 3 and 4 to a greater degree than in classes 1 and 2.

Any airport with an operating control tower (classes 1 and 2) has a

qualified weather observer who has direct communication with all aircraft

legally entering the airspace in the vicinity of the airport. All players

should know the ground rules. Current weather is, for example: "800 feet

and 2 miles" -- because the tower says it is.

Airports in control zones, but without control towers, may also have

qualified weather observers present, but they may not be talking to all the

participants. The IFR pilot knows the rules because he is in communication

with ATC and has been told that the weather is below i000 feet and/or 3

miles. The VFRpilot, on the other hand, must either contact the flight ser-

vice station, if one is on the field, or rely on his own judgement to deter-

mine whether or not VFRminimumsexist. There is no mandatory communication

in VMCweather. Honest differences lead to a violation of "VFR in IMC" and a

potential eonfllct with IFR traffic.

Airports with instrument approaches which lle in uncontrolled airspace

(below 700 feet AGL) and lack qualified weather observers rely entirely on

the Judgement of the participants to establish the operating ground rules. A

VFR pilot can legally operate at such airports as long as he estimates the

v_slbility to be greater than | mile and he remains clear of the clouds. An

arriving IFR flight might disagree with the visibility assessment of hls VFR

counterpart if both meet on the final approach to landing. The two pilots

are operating under different assumedrules because they lack a common infor-

mation source.

A second weather-related problem, which does not involve conflicting

estimates, occurs at airports in uncontrolled airspace. This is the poten-

tial conflict between an arriving IFR flight which breaks out of the overcast

below 700 feet AGL and a VFR flight operating very close to, but clear of, the clouds. There is no buffer to accommodate the changeover from total reliance on radar separation to a see and be seen doctrine. Perhaps further research is warranted to assess the impact of raising approach minima at such airports, so that the MDA (minimum descent altitude) is always in controlled airspace (e.g., above 700 feet, AGL). That way all participants would be operating under the same rules (VFR) at a well defined altitude. As it is now, an approach with an MDA of 500 feet terminates 200 feet into uncon- trolled airspace. At least if ground contact was required before entering uncontrolled airspace, some buffer would be created to accommodate the changeover from an IFR to a VFR environment.

CONCLUSIONS In general, the analysis of weather-related incidents involving general aviation did not indicate glaring deficiencies in skill or training of the GA pilot. While some pilots may have needed help from a ground-based facility, all but one of the pilots landed safely without damaging the aircraft. There was, also, little indication that the lack of equipment contributed to any of the reported incidents.

The major problems may be ranked as follows: I. Lack of timely weather information, especially in deteriorating weather.

2. Lack of exact weather interpretation, e.g., visibility reports.

3. Questionable judgement and attitude of pilots in regard to weather flying.

4. Pilot attentiveness.

5. Aircraft equipment.

6. Navalds.

7. ATCSystem.

In regard to single IFR aircraft, the problem causes appear to be: I. Violation of rules or lack of precision in IFR flying.

2. Weather interpretation, i.e., the classic issue of when

conditions have deteriorated from VMCto IMC.

3. Communicationsbreakdowns between FSS, controllers, and

pilots.

4. Poor preflight planning by pilots.

The problem causes for the VFRpilot encountering adverse weather appear

to be as follows: I. Weather turns out to be worse than forecast.

2. Pilots departing VFRin marginal weather conditions.

3. Lack of timely weather information or enroute weather

(between stations) e.g., PIREPS.

4. Poor pre-fllght planning by pilots.

5. Failure of pilots to initiate 180 degree turns soon

enough when the weather is worse than forecast•

In the incidents involving multiple aircraft, the major problems were

related to the difficulty of coping with non-routine operations during

weather diversions and the hazard created by two or more aircraft operating

on mixed flight rules and mixed flight phases in conditions of reduced visi-

bility.

These analyses suggest that improvements in the following areas may aid

in reducing future incidents involving GApilots and adverse weather•

i • Pilots a. The need for better weather recognition, especially estimates of visibility.

b. A more professional approach to IFR flying.

c. Better pre-fllght planning, especially in regard to runway information during winter months and planning alternatives in the event of weather changes.

d. Need for fuller understanding of the mechanics of weather observation and forecasting.

. ATC a, Need to improve handling of non-routine events caused by traffic diversions due to weather (sector coordination problems).

be Need for improved asistance to pilots confronting deteriorating weather.

.

Weather Information Services a. Need for more timely dissemination of weather infor- mation.

b. Need to provide pilots with better access to weather information for flight planning purposes.

o Pilot - ATC Relations a. The need to improve communication between pilots and ATC as regards weather-related problems.

b. The need to understand better the attitudes of _7 pilots towards controllers (and vice versa) and how these attitudes contribute to incident causation and associated consequences. For example: if pilots view controllers as policemen as opposed to "helpers", they (the pilots) tend to be easily offended, defensive, non-communicative, and in many cases refuse to confess problems which can affect the safety of others. In turn, controllers may lose their professional composure and ....

The reduction of GA weather-related incidents does not appear to require more experience for the General Aviation pilot, nor better equipment.

Aircraft equipment sophistication endorsed by some aviation safety advo- cates appears not to be a factor _n incident causation. The lessons sug- gested from the analyses provide additional evidence for curriculum design for training both pilots and controllers and, at least minimally, remind all General Aviation pilots of the potential The data suggest better utilization of the current resources of the avi- ation system such as better weather information dissemination, better utili- zation of the IFR system, and better training of VFR pilots to detect, diag- nose, and cope with changing weather patterns.

APPENDIX A.

ACCESSION NUMBERS USED

Preceding page blank

¢ . J, J The accession numbers on which this report is based are as follows: 13456 15733 20013 891o 13608 15751 20014 13806 15900 20015 9188 13840 15964 20016 9539 13887 16136 20048 9847 14008 16156 20198 9865 14019 16206 20260 14162 16230 " 20480 9967 14287 16264 20595 1o056 14439 16318 20667 10200 14532 16392 20743 10268 14582 16423 20782 10281 14588 16504 20954 10286 14589 16514 20986 10328 14599 16661 21084 10629 14858 16832 21086 10714 14903 16916 21284 10759 14964 16948 21391 10769 1/4979 17039 21424 10912 15052 17040 21428 10939 15059 17127 21601 11148 15075 17128 21645 11/468 15138 17343 21650 11634 15196 17533 21723 11673 15252 17639 21734 11744 15312 17654 21735 11806 15314 17700 21955 11876 15315 17728 21965 11911 15323 18068 21966 12136 15350 18121 21976 12166 15358 18223 22058 12211 15359 18277 22072 12215 15369 18496 22101 12216 15417 18571 22176 12447 15438 18584 22186 12484 15462 18672 22200 12533 15465 18785 22245 12589 15499 18786 22246 12714 15506 18888 22333 12726 15576 18928 12843 15610 19011 12.8.57 15655 19196 12876 15665 19214 13130 15670 19578 13288 15694 19586 13399 15699 19657 13419 15732 19734

Preceding page blank'i

I "11_l&_miL _im'm_ -- .-'e,-I "