Skip to main content

Full text of "DTIC ADA297000: Aircraft Icing Over Northwest Europe."

See other formats


AfR WEATHER SERVICE 
TECHNICAL REPORT 105-46 



AIRCRAFT ICING 


OVER NORTHWEST 




JULY 1945 


HEADQUARTERS 

AIR WEATHER SERVICE DTIS QUAIITT INSPECTED 5 

WASHINGTON, DC. 

DISTRIBUTION UNLIMITED: 

APPROVED FOR PUBLIC RELEASE 


2^r\rr\nr\r^ r\r\n 











HQ, 21ST LEATHER SQUADRON 


2 June 1946 


-w\ 



■\ 0 


i, 



PROSPECTUS 

AND PRELIMINARY REPORT 
ON AIRCRAFT ICING 
OVER NORTH-VEST EUROPE 


PROSPECTUS: 

This preliminary study of statistics on aircraft ice 
accretion over Northwestern Europe is the first In a series 
of proposed reports baded on observational data taken by 9th 
Weather Reconnaissance Squadron (Prov). Flying P-51-H and 
P-51-D aircraft, this squadron completed 1340 successful 
tactical weather reconnaissance missions during the period 
3 June 1944 to 3 May 1945. Their routes cover the English 
Channel, Southern N6rth Sea, Northern France, the Low Coun¬ 
tries and Germany west of lS^O* E; flights were made at a ve- 
rage altitudes from 5000 to 15,000 feet, but frequently mea¬ 
surement of thick cloud decks took pilots as high as 30,000 
feet and down to a few hundred feet above the ground. The 
large majority of these observations were- made by experienced 
reconnaissance pilots and provide an accurate, detaild three- 
dimensional description of visual meteorological conditions 
ov c r the area. Elements observed Include measured altitudes 
of low and middle, estimated altitudes of high clouds, visi¬ 
bilities aloft and to the ground, precipitation, icing, tur¬ 
bulence and condensations trails. 

The reconnaissance flights probide a unique source of 
data for meteorological research in sufficiently large quan¬ 
tities to insure moderate statistical accuracy. Several spe¬ 
cific investigations have suggested themselves to the writer 
in addition to the present summary of icing statistics, which 
will be supplemented by a paper on the forecasting of air¬ 
craft icing in Northwestern Europe, describin'; icing condi¬ 
tions to be expected on the basis of synoptic situation, tem¬ 
perature lapse rate and similar Information. Ot ier among 
these topics are: 

1. The delation between actual cloud structure and the 
elements of the radiosonde ascent, base' 1 on^a large number of 
observations in close nroximity to actual balloon runs. 

2. The averag 1 structure of different types of fronts and 
frontal systems occurring over Northwest Europe. 


§ 


/,, / r 5 




3. The structure of decks of "North Sea stratocumulus" 
as related to the vertical distribution of temperature and 
moisture in the air mass. 

4. The orographic effects of the mountains of Western 
Germany on different types of airflow. 

5. A study of turbulence similar to the one on icing. 

6. A study of condensation trails. 

A considerable contribution to the accuracy of some of 
the above work can be gained by coordinating tiie observations 
of 9th Weather Reconnaissance Squadron (Frov) Mrlth higher level 
reconnaissance cl ne over the same area by the Mosquito squa¬ 
dron of 325th Phot Reconnaissance Wing, Eighth Air Forces 
Made or ncipally between 20,000 and 30,000 feet, their flights 
can provide valuable information on altitudes of high cloud 
and tops of stronger frontal systems; they would similarly 
add much to any study of condensation trails or summer icing. 

PRELIMINARY REPORT ON AIRCRAFT ICING 
OVER NORTHWEST EUROPE: 

The data presented below on alrcraf’i ice accretion is 
the result of study of 1340 flights made by.’9th Weather Recon¬ 
naissance Squadron (Frov) June 1944 to May 1945, of which 307 
reported one or more instances of Icing. Breakdown by sea¬ 
sons is presented; a further breakdown into months is not 
considered valuable, because the relatively small number of 
observations detracts from statistical accuracy. 

Certain reservations must be made to the absolute accu¬ 
racy of the figures on type and intensity of icing. These 
observations were made by experienced fighter pilots, flying 
aircraft which cruises at between 290 and 300 mph true air- 
SDeed. The icing usually occurred during a short period of 
flight in cloud or while ascending or descending. The inten¬ 
sities given are necessarily not based on any quantitative 
measurement of accretion and reflect, to the*" extent they have 
not been eliminatied by the writer, the biases of individuals. 
The only significant tendency, If any, is for allots to over¬ 
estimate toe degree of icing. The intensities given conform 
fairly well to definitions apoearLng in Armn Air Forces Train¬ 
ing Manuals on aircraft icing: light Icing is such that it 
can be readily handled by de-icing equipment but neces¬ 
sitates an eventual change of flight plan to avoid the con¬ 
dition, heavy icing is such that it cannot be hadled by de- 
loin^ equi -ment. However, the P-51 has no wing or propeller 
de-lcdrs or anti-icers, and its only safe escape from hazard¬ 
ous icing is to climb or descend rapidly away from the con- 







ditlon. The Gorman training manual Verelsung, D(Luft)1209 
states th->t heavy icing is an increase in the ice coating 
of about one centimeter in five minutes, and this is as close 
to a quantitative estimate as we can come. 

The greatest operational significance of the work of 9th 
Weather Reconnaissance Sou adron (Prov) is that it was done 
entirely in single-engine fighter aircraft. An inordinate dis¬ 
trust of icing conditions seems to have thrown up among many 
fighter allots in the European Theater of Operations, even*to 
the point of avoiding any instrument flight whatever. This 
fear was enhanced bv the statement of theorists that ice ac¬ 
cretion increases in direct orooortion to airspeed, a claim 
which these observations tend oartially to disprove. Desope 
these objections, pilots of toe weather reconnaissance squa¬ 
dron flew instruments on more th n 90 per cent of all their 
1340 missions, yet only throe cases occurred of an aircraft 
being forced to turn back or land because of icing. No losses 
or damage of planes have ever been even indirectly attribu¬ 
table to icing. Furthermore, substantially more cases of flight 
in cloud above the icing level without any icing at all were 
reported than cases with icing. Of all icing encountered, 85.5 
per cent of thee instances were light rime, light clear or 
moderate rime, types which by the statement of pilots them¬ 
selves offer little or no hazard. a 

Certain general conclusions can be seiftly drqwn from \ 
the data presented below: 

1. Icing is worst during Winter and Soring seasons, with 
maxima in November and January. x cing is an almost negligibel 
at altitudes from the surface to lb,000 feet during Summer 
months. This i partly because t e clouds over Europe in Sum¬ 
mer are largely convective type, is olated and easily avoided. 
Also the Summer of 1944 averaged considerably below toe mean 
annual cloud cover. 

2. Rime icing predominates in the ratio of flve-to-one, 
and more dangerous clear icing is almost confined to 
cumulonimbus, bulging cumulus and nimbostratus (freezing rain) 
clouds. This is in agreement with theory that clear icing 
occurs only with large stapercooled water droplets. 

3. Rime icing is especially predominate in Winter, when 
strong convective clouds are rare over land. 

4. In stratocumulus, altocumulus and middle t pe 7 (Ac 
and As) ciouds, rime icing occurs almost exclusively (90- 

97 per cent), with innocuous light and moderate rime predom¬ 
inating in the last two. 

5. Dangerous icing is almost a certainty above tl* 
freezing level in dumulonimbus. The icing danger in cumulus 
increases with the covectlve activity and attendant drop 
size. 



3 






6. There is only one chance in six of getting any icing 
in altoatratus cloud, almost no chance in stratus, and less 
than a 50 oer cent chance in stratocumulus, altocumulus and 
middle tyue 7. Flight through active altocumulus castellatus 
is responsible for the cases of moderate clear and heavy rime 
in altocumulus . 

8. The cloud types in order of decreasing icing hazard 
are: cumulobimbus, nimbostratus (freezing rain), bulging cumu¬ 
lus, cumulus humilis, altocumulus, stratocumulus, middle type 
7 (usually thin), altostratus and stratus. No icing is to be 
expected in the oure ice crystal (cirroforra) clouds, excent 
in cumulonimbus anvil tops. 


TOTAL DATA (June 1944 - May U45) 


Total Missions: 1340 
Icing ReoOrted: 307 - 22.9h 
Uncertain: 49 3.7.2 

No Icing: 984 73.4',' 


Litzht 


Moderate 


Clear 


By Cloud Type: 


Light Rime 
Modt. Rime 
Others:L.C 
M.C 
H.C 
H.R 


Sc 

Cu 

Cb 

Ns 

As 

Ac 

57 

35 

0 

11 

9 

49 

30 

11 

2 

11 

4 

20 

5 

15 

1 


2 

5 


10 


2 

1 

3 


3 

2 

3 



5 

8 

2 

3 


3 


Times aircraft flew in cloud above 
freezing level without icing : 364 

Cloud Type : St Sc Gu Ns Ac M7 As 

Occurrences: 8 114 19 3 88 47 85 


Accession For 

US1S GRA&I 
OTIC TAB 
Unannounced 
Justification- 


Distribu tion/ _ 
Availability i 
(Avail aad 






otal 

MONTH: 

J an. 

Feb 

Mar. 

Aor. May 

June 

Missions: 

85 

113 

142 

149 * 

178 

cing 

Reported: 

34 

30 

49 

52 

13 

Pf 

?rcentnge: 

40/ 

27# 

• 35 5 

35/ 

7 % 



Oct. 

■’ov. 

Dec. 





59 

94 

182 





18 

37 

47 





31/ 

39:5 

3°-t 




KINGr (March, April 

i _ 't, 

1 o 

May 

1945) 


Total Missions: 

302 





Icin 

g Reported: 

109 - 

36. 

1:5 




Uncertain: 

5 

1. 

7 5 




No Icing: 

188 

62. 

O s' 

c. . y 



Rime 

Light 

Moderate 

Heavy 

ob 

19 




9 

Clear 

ii 

10 




2 

By cloud type: 







St Sc 

Cu Cb 

Ns 

As 

Ac 

M7 

Light 

Rime 0 15 

12 0 

2 

2 

30 

8 

Modt. 

Rime 0 7 

7 2 

0 

0 

3 

1 

Other 

: L.C. 2 

7 



2 



M.C. 

8 

1 


1 



H.C. 

1 

1 





H.R. 2 

3 1 

1 


1 

1 


July Aug. Sep,/ 
183 128 76 

7 3 9 

2,A% 12# 


Times aircraft flew in cloud above 
freezing level without icing : 83 


Cloud Type : 
Occurrences: 


St Sc Cu Ns As Ac M7 

1 20 4 1 18 27 12 


SUMMER (June, July, August 1944) 


Total Missions: 

489 

Icing Reported: 

23 - 4. 7# 

Uncertain: 

29 - 5.9/ 

Ho Icing: 

437 -89. 4;/ 



5 






_ Light 

Rime _1 1 

Clear_ 2 

3y Cloud Type: 

3t 3c 

Light Rime 9 6 
Modt. Rime 0 1 
0 the r: L.C. 

M. C. 

H.C. 

H.R. 


Heavy 
0 * 
2 


Cu Cb Ns As Ac 

3 0 2 0 1 

0 0 4 1 1 

2 
1 

1 1 


Moderate 

7 

I 


M7 

1 

0 


Times aircraft flew in cloud above 
freezing level without icing : 117 

Cloud Type : St Sc Cu Ns As Ac M7 

Occurrences: 2 23 6 2 37 32 15 


FALL (September, October, November 1944) 


Total Missions: 229 
Icing Reported: 04 

Uncertain: 4 

No Icing: 161 


28.0 % 
1.7 % 
70.3% 


Rim© 

31 

14 

4 

Clear 

12 

2 

3 


By Cloud Type: 


St 

Sc 

Cu 

Light Rime 0 

10 

8 

T’odt. Rime 0 

4 

0 

Other: L.C. 

2 

4 

H.C 


1 

H.C. 


1 

H. R. 

1 

1 


Cb Ns As Ac M7 

0 2 3 5 5 

0 5 0 6 0 

1 13 1 

1 

1 1 

1 1 


Times aircraft flew in cloud above 
freezing level without icing • 


6 







Cloud Type : St Sc Cu Ns As Ac M7 

Occurrences: 0 28 3 0 20 17 3 


’"INTER (December 1944, January, February 1945) 

Total Missions: 320 

Icing Reported: 111 - 34.7$ 

Uncertain: 11 - 3.4$ 

No Icing: 198 - 61.9$ 


Light _ Moderate _ Heavy 


Rime 08 


.33 



To- 

Clear 5 



3 



i 

By Cloud Tyoe: 







St 

Sc 

Cu 

Cb Ns 

As 

AC M7 

Light Rime 0 

26 

12 

0 

5 

4 

13 16 

Modt. Rime 0 

18 

4 

0 

2 

3 

10 4 

Other: L.C 

1 

2 



1 


M.C. 




1 

1 

1 

H.C. 

• ; 

1 





H.R. 

2 

1 

1 

1 


1 

Times aircraft 

flew 

in 

cloud 

above 


freezing level 

without 

icing 



: 93 

Cloud Type 

: St 

Sc Cu 

Ns 

As Ac M7 

Occurrence 

s: 5 


43 6 

0 

10 12 17 


s/Holt Ashley 
t/HOLT ASHLEY 
1st Lt., AC, 
Weather Officer 


7 



TtCTIICAL 51CTI0M 

HEADQUARTERS, 21 V? v/BATHBR SQUADRON 


1 g 
<y~' 

... J i , l 


: " ••••• - j - 


18 July 1945 


URCRA'TT ICING OVER NORTRVKoT EUROPE 
( y.i ; •*. u,,k a ■ : , 

Part I: Analysis of 1340 Weather Reconnaissance Flights, June 

1944 - May, 1945.) - A ' _ 

.i-y • h r>n' «t e" ; ' ••* r f flKffBHMU99T!*T« , ws*r 

A."' General Remarks. \j ' «MJMS 

The data presented here on aircraft ice accretion are the 
result of a study of flights mide by 9th weather ?Reconnaissance 
Squadron (Prov), of which 307efceported one or more instances 
of icing. Flying r-51-3 and p-bl-D airplanes* this squadron 
completed 1340 succesful tactical v/oathor reconnaissance miss¬ 
ions durxng the period 3 June 1944 to 4 l-'uy 1945. Their routes 
cov-jr the English Channel, Southern Worth Sea, Norther France, 
the Low countries and Germany west of 13° 30* B. Flights were 
made 't average altitudes from 5000 to 15,000 feet, but fre¬ 
quently measurement of thick cloud decks took pilots as high 
as o0,000 fact and down to a few hundred feet above the ground. 

The large majority of these observations were made by exper- 
ienced reconnaissance pilots and urovide a detailed three—di¬ 
mensional dexcrintion of the visual meteorological conditions 
over the area covered. elements observed include measured al¬ 
titudes of low and middle, estimated altlt ;des of high clouds, 
visibility aloft and to the ground, precipitation, icing, tur¬ 
bulence and condensation trails. 

it is felt that several limitations exist to the absolute 
accuracy and general usefulness of t^ese icing data, and these 
should be listed before launching upon a detailed analysis: 

a. The data are for «, period of only one year. Although 
breakdown b.y seasons is presented, it is employed for statistical 
convenience rather then to emphasise differences. Thus, the 
total figures on icing related to fronts, cloud structure, etc, 
are more valuable than monthly or seasonal sums. 

b. The pilots were not weather officers; although their 
experience level was high, their training was limited. At best, 
Pivr icing observations will be affected by personal considera¬ 
tions end are not susceptible of quantitative measurement. 

c. only ohe type of aircraft was emplo ed. Experience 
ith icinc: indicates considerable venation of intensity with 

the rdrfoil. -mis difference may be even more significant^ 
than the relative increase of ice accretion with increase i.n 
;’ir speed. 


6~n 


i 









d. Tne high airspeed of the r~51 (290 to 300 true, 
cruising around 10,000 foot) and the natural desire to avoid 
icing situations combined to Unit the time of flight in any 
particular condition. phis rapid passage through cloud made 


of degree of icing very 
overestimate icing as a 


difficult. Also many 
result of early weather 

three facts tend to en- 


rrood estimates 
pilots tend to 
Graining. 

to balance the above limitations, 
ha.nc e the value of the data presented; 

a. :'ne large number of missions flown in approximately 
the same manner over the s-.me area contributes the advantage 
of high statistical accuracy to the results. For exmaple. If 
all disturbing factors are eliminated aside from pur© chance, 
the probable error in the chances of 307/1340 that any one 
such mission will encounter icing is less than four per cent, 2 

b. The pilots were all tranined in weather by the same 
personnel and according to the sane outline. Standards of ob¬ 
servation were passed from pilot, to pilot by a process of hav¬ 
ing new men fly wing on their first missions. Only relia¬ 
ble, experienced pilots were permitted to lead flights regu¬ 
larly. Furthermore, the writer has tried to adjust some of the 
observations to a more level standard according to h!s personal 
experience with the pilots Involved. 

c. All missions were flown in the same type of nil-craft 
providing the results with internal continuity. 

The greatest operational significance of the work or 9 th 
feather Reconnaissance Squadron (Prov) is that it was done en- 
tirely in singl-cngino fig-ter nrcrnft. An undue fear of icing 
conditions seems to have grown up among many fighter pilots in 
tie Europe an Theater of Operations, even to the point, of avoi- 

;;n5ta n i ( -' l h ,trum 'h, t rll -' t ' Ih g'>var. 3 Despite these objedtione, 
pilots of the weather reconnaissance squadron flew instruments 
on more than 90 per cent of all their 1340 missions, yet^?* 
tiirec cases occurred of an aircraft being forced to turn back 
or land because of icin-t. Fo losses or damage of planes have 
evei ^y el ? -^ n dlrectlv attrlvuted to icing. More cases oc¬ 

curred cm xlying in cloud above the freezing level withou 
icing than cases of icin'. Of all icing encouterd, 85.6 per 
"?n ' ro instances were lignt rime, light clear or moderate 
rime, types wnich b-» the statement of pilots themselves offer 
little or no hazard. 


2 







B. Analysis of Tables* 


A number of general conclusions can swiftly be drawn from 
the tables presented below. Although many of these are obviou 
upon inspection, it is felt valuable to summarize them and 
Voint out, from experience, any especially significant or 
detracting factors involved. The reader is referred to the 
limitations set forth above as a deterrent from blind accep¬ 
tance o'* these results as true in the general case. 

rVom the date on total icing reported and icing by cloud 
typo, the following facts ore a"parent* 

a. The experiences of these pilots indicate that icing 
is worst during winter and spring months, with maxima in 
November, December and January. Icing is least prevalent, or 
rather most easily avoided, in summer. 

b. Rime icing predominates in the ratio of five-to- 
one, and the more dangerous clear icing is almost confined 
to cumulonimbus, bulging cumulus and nimbostratus (freezing 
rain) elo ids. Tills is in agreement with teory that clear icing 
occurs only with large supercooled water drpjkeets* 

c. Rime icing is especially predominate in winter, 
when strong convective clouds are rare over land away from 
immediate seacoasts. 

d. In stratocunulus, altocumulus and middle type seven 
(Ac and As) clouds, rime icing occurs almost exclusively (90- 
97 per cent), with innocuous light and moderate rime prevail¬ 
ing In the last two. 

e. Dangerous ic-Jng is almost a certainty above the freez¬ 
ing level in cunulobimbus; the icing danger in cumulus inereas-” 
es with the convective activity and attendant drop size. 

f. There is only one case in six indicated hehe of getting 
icing in altostratus cloud, almost no char.cc in'stratus, 

and less than a 50 per cent chance In's tr a to cumulus, altocu¬ 
mulus nnd middle type seven(Ac and As), flight through active 
altocumulus castellatus and cloud associated with spring and 
summer high level thunderstorm activity is responsible for the 
cases of moderate clear and heavy rime In altocumulus. 

g. Hie cloud tyres in order of decreasing icing hazard 
are: cumulonimbus, nimbostratus (freezing rain), bulging cumu¬ 
lus, cumulus humilis, altocumulus, stratocumulus, middle type 
seven, altostratus and. stratus. No icing is to be expected in 
the pure ice crystal (eirrofomi) clouds, except in cumulonim¬ 
bus anvil tope 3. 

prom da a on icing in fronts the following facts are 
apparent: 

a. Hie experiences of these pilots Indicate that there 
Is no significant Increase in the icing danger in frontal cloud 
over that in pure air mass cloud, except of course that the 


3 





former Is harder to avoid. Danrerouc, tu 

ate clear, heavy clear) wan enrou tered in' ir 6BVy ri^ ^ 1 ' 3, noder- 

“o HIV }” l T° ntS: li -• —n 14 III cent % 1% 

10 cases in air mass cloud, p C9nt ol tne 

r^wsss :-~£| *«r- 

icing, except possibly inactive nuaal-atatfon!^ ha ^ r ? ous 
encountered were lens a nJn - 0T , 0 i,„ + v'' ■tat_onaries. Cold fronts 

-rom the data on ^ T.LtX ° C £ Ud ? d t ^ 93 ‘ 

facts are apparent: ' '' * 1 to nlr mass the following 

correlation betwee^al^m-a” tn^V 110 ^ indicate a closer 

«*> «« occurrence ° CCUrren<!9 <* ^ing, 

origin, «hloh 1 pr,domlmte t li i! ' 1 i ted t0 ” lr ma3: ‘ es of maritime 
lr , ' * " nic “ Predominate In a largo ratio in Northwest T?m»- 

J«t.« observatlon 13 ln agreement with authors on tl^Tsub- 

c. Dangerous icing is nreseni to n rmi«v, i * i , 

in unstable air masses, with thSl? a”,Delated d *« r »* 

vective clouds. Tire lor- oit ? d strongly con- 

noderato clear and heavy clSnr ;U J^oSt of S l^°f ilQ **Z ^ Iue ' 

?^ a ins?ablo°^w T tabl€? taaritirae tr °Pical with ei^fou^of 
sl S or 7C {COmn ° n OVer lond ln winter ) is third with 

d. Rime icing is the normal situation in stnhi* 

©• Icing in continental niv v» *\s rA , nor , t'Ari fnn 
quently to make general! otlons. T f P d boo infre- 

111 unstable continental tropical of 
in an afternoon cumulonimbus. This 
addition of moisture during a short 
terrenean Sea. 

/: <rhR subscripts in air mass designations are 

i 1 ‘' cc ar> follows: a, of Atlantic origin; b, of continent- 
al European origin; af, or African origin. continent 

,o m c explanation of the table on icing related to stabi- 
iity and relative humidity is in order. For each case it wL 
stympted to find the radiosondes observation best represent- 
tlng tne column of air In which the Icing was encountered The 
iclnr layer was then Investigated to ascertain the avegore- 
lative -tumidity code figure, thus dividing the cases into 
' " W °° nt 0acl1 •>•!»» 00 penmen? and five • 

th r the Lt” 9 ° P° r oont - secondly It was estimated .he- 
" h the icin ® i & ycr had an approximately moist adiabatic 


- ~ J JL H J. X tJ 

i® case of heavy clear icing 
African origin occurred 
air was modified by some 
trajectory over the Medi- 


4 






lapse rate, was definitely more stable, or definitely more un¬ 
stable than moist adiabatic. From this data the following 

facts are a ^ pl ^g 0 ex ^ erlenC0g 0 f these pilots indicate that* although 

saturated atmospheric layers of gre&t instability are 
Northwest Europe, where they do occurs below freezing tempera¬ 
tures dangerous icing is almost a certainty. ^ . 

b. Icing in definitely stable layers is usually not 
hazardous. Only four cases of dangerous icing occurred out of 
GV flights in such air, all of them heavy rime. 

c. Icing danger increases with higher relative humi¬ 
dity, as tdicated by radiosonde observations. Dangerous icing 

is'rare below 90 per cent; only light icing was met below bO 

per caT I 1 ^ , asGOciftion with the r P diosonde investigation, the te- 
perature at which each case of icing took place was determined. 

The cost interesting fact discovered is that oi 4V cases of dan¬ 
gerous icing only fiv" occurred at temperatures below -10 C. 

(14° n», ). Of these oases, four heavy rime and one heavy clear, 
all but one were in strong convective cloud in an unstable 
maritime air mass. 'flic heavy clear was met ne. r -20 C. m 
cumulonimbus of an active cold front type occlusion. These 
facts confirm the statements of some authors regarding relative 
safety 0 from dangerous icing at temperatures far below freezing. 

Two cases of light riine were reported in stra to cumulus over¬ 
casts at temperatures confirmed to be above freezing (between 
0° and 2° C.). These offer proof of IvIcNeal s suggestion that 
li^ht ice can.occup through evaporational cooling in cloud up 
to about 55.6 F. Considering the high cruising speed of the 
P-51, thev indicate that the assumption oi an effective in¬ 
crease in the height of tho freezing level to dynaraiCgWarm- 

ing of air passing over tho wings is probably fallacious. 


C. Tables 


tqt*I, DATA (-Tuns 1944 - l!ay 1945 


Total Missions 
Icing Reported 
Uncertain 
go Icing 


1340 

307 - 22.9 ' 
49 - 3.7 t 

904 - 73.4' 


Rime ] 
Clear 


T7F 


Light 


30 


9 . 2 %) 


Moderate 

“75 t2g .-^y 

16 ( 4.9 



5 




TOTAL DATA (continued) 


By cloud Type: 




St 

Sc 

cru 

Li ght 

Rime 

0 

57 

35 

Modt. 

Rime 

0 

30 

11 

Other 

: L. C . 


5 

15 


M.C. 



10 


H.C'i 



3 


H.R. 


5 

8 


Cb 

Ns 

As 

Ac 

M7 

0 

11 

9 

49 

30 

2 

11 

4 

20 

5 

1 


2 

5 

1 


2 

1 

3 


2 

3 




2 

3 


3 

1 


Times aircraft flew in cloud above 
free sing level without icing 


Cloud Type : St 3c Cu Ns As Ac H7 

Occurrences: 8 114 19 3 85 88 47 


Icing not in Front: 210 (64.4 .) 



Light 

Moderate 

Heavy 

Rime 

— 


IS 

Clear 

2l ~ 

11 

2 

Icing 

in Front: 116 

(35.6;0 



Weak 

C.F. 

L.R. 

9 

Modt • 

C.F. 

10 

Strong 

C.F. 

0 

Weak 

W.F. 

17 

Modt. 

W.F. 

11 

Strong 

W.F. 

1 

Weak 

Occ. 

4 

Modt. 

Occ. 

8 

3 trong 

Occ. 

0 

Stationary 

3 

Totpl! 


63 


M.R. H.R. L.C. 

7 0 1 

2 0 4 

10 0 
4 4 0 

12 2 
0 0 0 

4 10 

6 11 
0 0 0 

0 0 1 

25 8 9 


Dangerous Icing Cases below 


-10° C. 


M.C. 

ii. C . 

Total 

1 

0 

18 

1 

1 

18 

0 

0 

1 

0 

0 

25 

2 

2 

20 

0 

0 

1 

1 

0 

10 

0 

1 

17 

0 

2 

2 

0 

0 

4 

5 

6 

116 

6 out 

of 47 

(13#) 


Heavy Rime 
5 


Modt. Clear 
0 


Heavy Clear 
1 


6 





TOTAL DATA (continued) 
Probable Air Mass 


Unstable 

raPka 

L.R. 

70 

M.R. 

28 

H.R. 

13 

T n 

XJ • % 

12 

M.G. 

9 

H.O. 

3 

Total 

138 

Stable 

mPka 

10 

5 

0 

v 2 

0 

0 

17 

Unstable 

mPwa 

20 

13 

6 

5 

1 

0 

45 

Stable 

mPwa 

41 

18 

2 

4 

3 

0 

68 

Unstable 

mT-a 

21 

8 

O 

3 

2 

4 

40 

Stable 

mT-a 

3 

0 

0 

1 

0 

0 

4 

Unstable 

cPke 

0 

0 

0 

1 

0 

0 

1 

Stable 

9 Pke 

0 

0 

0 

1 

0 

0 

1 

Unstable 

cPwe 

3 

0 

0 

0 

1 

0 

4 

Stable 

cPwe 

1 

0 

0 

0 

0 

0 

1 

Unstable 

cTwaf 

3 

1 

0 

0 

0 

1 

5 

stable 

cTwaf 

4 

0 

0 

1 

0 

0 

5 

Total: 


176 

73 

23 

30 

16 

8 

326 

Icing Related 

to Radiosonde 

r. 228 

(November 

1944 - 

May 1945 

M.A. 

L.R. 

0 

M.R. 

1 

H.R. 

4 

L.C. 

1 

M.O. 

5 

H.C. 

1 

Total 

12 


M.A. 

85 

34 

9 

11 

9 

1 

149 

, 

M.A. 

42 

18 

4 

3 

0 

0 

67 


R.H. 0 

43 

19 

9 

7 

9 

2 

89 


R.H. 9 

56 

28 

6 

6 

4 

0 

100 


R.H. 8 

22 

6 

2 

1 

1 

0 

32 


R.H. 7 

4 

0 

0 

1 

0 

0 

5 


R.H. 7 

2 

0 

0 

0 

0 

0 

2 


Total: 

127 

53 

17 

15 

14 

2 

228 



MONTH: 

Jan. 

Feb. 

Mai’. 

Apr. 

May 

June 

July Aug. 

Sept 

Total Missions: 

85 

113 

142 

149 


178 

183 128 

76 

Icing Reported: 

34 

30 

49 

52 


13 

7 3 

9 

Percentage: 

40^ 

27% 

3 5'’s 



7% 

4% 2,4% 

12 # 


Oct. 

Nov 

Dec. 







59 

94 

122 

-*0nly 

11 

missions flown. 



18 

37 

47 









SPRING (March, Anril, 1-3 May 1945) 


Total Missions: 

302 


Icing Reoorted: 

109 - 36. l/o 


Uncertain: 

5 - 1.7 /4 


No Icing: 

138 - 62.2,4 


Light 

Moderate 

Heavy 

Rime 66 

19 

9 

Clear 11 

10 

2 

Times aircraft 

flew in cloud 

above 

freezing level 

without icing 

: 83 


SIMM R (June, July, August 1944) 

Total Missions: 489 
Icing Reported: 23 - 4.7% 

Uncertain: 29 - 5 . 9,4 

No Icing: 437 - 89.4$ 



Light 

Moderate 

Heavy 

Rime 

-^ 

7 


T) 

Clear 

- 5 - 

-r 


2 

Times 

aircraft 

flew in 

cloud 

above 

freezing level 

without 

icing 

:117 


FALL (September, OctotfoV November 1944) 


Total Missions: 

229 


Icing Reported: 

64 - 4.7/4 


Uncertain: 

4 - 5.9$ 


No Icing : 

161 - 89.4-, 


Light 

Moderate 

Heavy 

Rime 3l 

14 

4 

Clear 12 

3 

3 


Times aircraft flew in cloud above 
freezing level without icing 


I 







WINTER (December 1944, January, February 1945) 


Total Missions: 

Icing Reported: - 34.7$ 
Uncertain: - 3.4$ 

\ No Icing: - 61.9$ 



Light 

Moderate 

Heavy 

Rime _ 

68 

33 

16 

Clear_ 

5 

3 

1 


Times aircraft flew in cloud above 
freezing level without icing : 93 




Part II: The Forecasting of Aircraft Icing. 

Aircraft ice accretion is the greatest weather hazard to 
regularly scheduled flight ooeration. Its primary danger does 
not lie, as blieved by many, in the widespread occurrence of 
icing in dangerous form, but rather in the difficulty of ac¬ 
curately forecasting the type and degree to be expected. 

Part II of this study has been orepared to suoplement and 
integrate the objective results on icing presented in Par I 
The conclusions drawn herein are based in part on the writer’s 
experience in ice forecasting for Northwest Europe, in part on 
facts and suggestions obtained from discussions and reading on 
the subject and in oart on the material summarized in Part I. 

The basic problem of an icing forecast is twofold: first, 
where in the atmosphere will water clouds or precipitation 
occur at temperatures below approximately Oo c.? second, what 
will be the range of droplet sizes and the total quantity of 
condensed moisture in these clouds? The first answer esta-* 
blishes the possibility of an icing situation, the second 
Tlves a good indication of whether or not the icing actually 
will occur and what will be its type and degree. The problem 
of cloud and temperature forecasting is a common and general 
one, so one must concentrate on the more difficult second ques¬ 
tion to forecast icing. In general, with a given type of air¬ 
craft, the degree of icing(light, moderate, heavy) will vary 
directyly with the quantity of condensed moisture per unit vol¬ 
ume; the type (rime, clear) will depend on the predominant 
droplet si z e. The larger the droplets, the more likely they 
are to break on impact with the aircraft and form clear icing. 

*o avoid non-essentials, the problems of usually light frost 
and rare melting-snow ice will be omitted; also it is assumed 
that the reader realizes rime and clear icing can exist in 
all degrees of combination. 

The drop sixe in a given cloud depends on the strength 
of vertical currents available to support larger drops and 
on the length of time the cloud has existed. The quantity of con 
densed moisture deponds, among other complex factors, on the 
temperature ana the length of tLme a source of fresh moisture 
or added condensation has been able to feed additional or lar¬ 
ger drops Into the cloud. Thus It Is seen that old and active¬ 
ly convective clouds are theoretically the source of most dan¬ 
s' rous aircraft icing. 


Such Is the case in Northwest Europe, and two general 
synoptics situations often occur in which such bad icing may 
be expected: 

a. During the winter half-year (October through March) 
with a west-northwest to northerly flow of cold unstable mari¬ 
time air across the North Sea, a mass of active cumulus and cu¬ 
mulonimbus clouds, extending with strong convection to heights 


10 





of 15,000-25,000 feet. Is built up. Such clouds appear to reach 
500 miles inland during the day before losing most of their 
intensity. Often such an air mass has already had a long tra¬ 
jectory across the North Atlantic Drift before reaching the ad¬ 
ded convective impulse of the relatively warm North Sea. These 
clouds are hard to avoid, particularly at night, and are usu¬ 
ally a source of heavy rime and/or dangerous clear icing above 
their verv low freezing levels. 

b* During the late spring and summer months (April 
through August), a protracted southwest to south-southeasterly 
flow floods Northwest Europe with a mass of very warm, fairly 
stable air, always trooicnl In origin and subject to varying 
degrees or maritime Influence. With the approach of a low from 
the west or southwest such air undergoes a decrease of stab¬ 
ility, which progresses from high level down to the surface 
where it is greatly enhanced by afternoon heating from the 
strong seasonal insolation. With the approach of a fast-moving 
c'lri front (or cold front type occlusion) from the northwest, 
the convective instability of the tropical air is violently 
released. In either case, that of the low or that of the cold 
front, the result is the formation first of alto cumulus cas- 
tellatus, then of strong cumulonimbus cloud 3 and thunder¬ 
storms. Such clouds do not have the age vut they definitely 
have the c onvetion to produce very dangerous icing. The air 
mass cumulonimbus are fairly easily avoided in daylight; the 
cold-frontal cloud wass Is almost impossible to avoid in flight, 
and severe turbulence renders such a situation dangerous even 
below the freezing level. 

The one important source of dangerous icing which does 
not depend on convection and to produce it. The fact is that 
strong air mass discontinuities and sharp frontal inversions 
are very rare in Europe. However, freeaing rain or glaze situ¬ 
ations sometimes occur over 3mall areas. The conditions are a 
southwesterly or southerly flow of maritime of continental po¬ 
lar air overrun by maritime tropical air from a warm front or 
occlusion approaching from a westerly quadrant or from waves 
forming on a quasi-stationary in the area. The worst icing is 
moderate or heavy clear and is nearly always met within the 
overrunning raT rather than below the frontal inversion. Often 
the release of convective instability in the mT is a contri¬ 
buting factor. In all the above forecasting it must be borne 
in mind that orographic influences always increase 'icing dan- 
ger. 

Before continuing a discussion of synoptic situations, 
some remarks are in order about the temperature effect on 
icing intensity As we pointed out in Part I, except in 
actively convective cloud3 hazardous icing is n rarity below 
-10° G. The most dangerous cases of such convective cloud for¬ 
mation have been mentioned above; anyforecaster can J*e6all 
a few others after a study of European weather. Excluding these 
situations, some "interesting conclusions can be drawn concern- 
in", lasers of icing hazard. Assuming a moist adiabatic lanse 


11 




rata in cloud the layer from 0° to -10° C. ranges around4500- 
5500 feet in thickness. This moans that in a large majority of 
cases the layer of dangerous icing is less than 5500 feet 
thick. In strong warm-frontal overrunning, such as that des¬ 
cribed above, the lanso rate is much stabler than mois t adia¬ 
batic, and the dangerous icing layer may be un to 1QD00 feet 
thick. In other situations where cloud occurs in air more sta¬ 
ble than moist adiabatic, observations such as those in part 
one demonstrate that little or no icing is usually the case. 

Thus an aim lane "ith a good rate of climb can very often as¬ 
cend rapidly through the icing layer without gathering a dan¬ 
gerously thick coating. This brings us to a very important orin- 
ciole: it is more fruitful in the long run to be able to fore¬ 
cast with assurance the absence of dangerous icing or the man¬ 
ner of avoiding it, than it is to be prepared with an ominous 
warning every time toer is the slightest possibility of such 
a ha sard. 

with that thought in mind, the following summary is t>re¬ 
sented of certain synootic a nd cloud situations when icing is 
not the hazard it is believed bl some to be: 

a. often during the winter half-year the same sort of 
northerly to westerly flow of cold air as described above oc¬ 
curs onto the European continent from the North Sea, except 
that the air is under anticyclonic influence and exgibits sub¬ 
sidence and a definite dry inversion somewhere between 4000 
and 10,000 feet. The normal situation here Ls an overcast 
stratus or stratocxiraulus cloud. It is often in two distinct 
layers, but the total vertical thickness is seldom more than 
5000 feet. If the trajectory has been westerly enough to 
pass over England or the Frenc oeninaulas the stratocumulus 
is broken. The usual icing condition is innocuous light or 
moderate rime. ''.'ith layers less than 1500 feet thick ther is 
rarely ant icing at all. It Is the writer's belief that cases 
of heavy rime reported in this cloud are caused by protracted 
flight in it, which permits considerable accumulation. Occa¬ 
sionally the problem is complicated by the presence of a few 
cumulus columns inside the str ’tocumulup, where the icing is 
locally nore severe. 

b. Throughout the year many cases occur in Northwest 
Europe of weak stationary fronts and the passage of dissi¬ 
pating ends of warm and occluded fronts. These are accompan¬ 
ied by little or no orecioitation. They consist of many rela¬ 
tively thin, broken layers of stratocumulus and altocumulus, 
often with altostratus or haze between layers. Except in very 
isolated patches of freezing rain, these fronts do not consti¬ 
tute a source of hazardous .Lcing. 'i'he maximum forecast should 

be for moderate rime, with no icing at temperatures below -15° C. 



12 




^hrou^hout the Tear scattered to broken cumulus 
huriilis and small’bulglng cumulus very frequently form bet¬ 
ween 1100 and 1900 hours In various maritime air masses. ine 
danger ms icing in these clouds is confined to their tops, 
especially with a fairly high freezing level. In these tops 
it is usual!v moderate clear, heavy rime or mixtures. The 
forecast should be to avoid the enmius topes, which is easily 
done. bV'-n with Isolated tower in- cumulus ahd cumulonimbus 
davligot flights can bo cleared perfectly safely with this 
■varninr,. 

Fins 11 , the following common cloud situations a re men¬ 
tioned, in 'hich most forecasters will agree the Icing hazard 
is. small to negligible: 

a. Throughout the veer end with almost any wind direc¬ 
tion, covergence and radiation effects combine to produce 
various thin scattered or brokende<* s of str bus, strato- 
''u*"u'>us, a! tostratus o ; - altocumulus. Individual layers are 
rarely more than 2000 feet thick. The normal forecast is for 
lh-ht rime icing above the free/,In - level. In well over 50 
por C'U'ii of such. CS363 there is no visible icing ut all. 

b. On fall and winter mornings with or without snow 
cover patches or layers of persistent fog or s tratus iorm as 
the result of nighttime radiation through clear air. Under 
freezing tempera cures such a situation is not productive of 
icln^. It is tie writer's belief that isolated reports cf 
take-off difficulties in such weather nr>'- explained by inade¬ 
quate measures for prevention and removal of frost, which 
forms during he nir.ht on wings and other exposed surfaces. 

It must be noted that many important cloud and synoptic 
situations have certainly been*omitted from the above. The 
forecasters are referred to the ,-general remarks made in the 
first four par-granhs. The primary sources of icing danger are 
active convection and extensive areas of freezing rain. 

To conclude, the suedesful icing forecaster is the man 
-•ho can tell a pilot with equal confidence where icing is and 
where it is not •> hazard. 


Notes: 


1. E.g.: B-26 aircraft flying a t much lower airspeeds 
have often encountered moderate or severe icing in clouds 
where P-51 pilots reported only light rime. By expansion on 
material in Air Ministry Meteorological OfficeJ4.0M. 395, 

Notes for the Guidance of Fore casters . Tge Accretion on 

Aire raft, and an o r o x im ate formula fo r the rate of Ice accretion 
is: 

1. A. “ KvEr Jj600 - t 1 

■where v is the true airspeed, E the quantity of liquid water 
in unit volume of air, r t e radius of water droplets and t 
the 0 # uit i <r rado tern 'era bum, all In c. ,# ->• units. The constant 


13 






K depends principally on the type of airfoil and little Is 
known theoretically or experimentally about its values. Iron 
qualitative results, K is much lover for the P-51 than for the 

13 m 

' 2 % G.H. Richardson, An Introduction to Statistical Ana - 
lvsis, T J3AFl, 1944 gives the following formula' for the stan- 
dard - deviation from th r ' mean of probabilities, assuming a 
uoir.t binomial curve is applied to the probabilities from 
a large number of series of n observations each: 


CT - v /npq 

wh*re n is the number of observations in each sample, and p 
and q are the respective probabilities that the event will or 
will not occur. For out observation series here: n 1340 
p 307/1340, q 1033/1340; 

V 1340x307x1033~ — 15.4 

1340x1540 


TT, 


.6745 


10.4 


The nrobable error E 10.4 cases of icing. The percentage pro¬ 
bable deviation from the assumed mean of <->0 1 is therefore E/M 
10.4/507 5.45. 

3, Air Ministry Meteorological Office M.OM. 393 and other 
texts state that ice accretion is proportional to the true air¬ 
speed (see formula in footnote 1.). P-51’s slying at cruising 
speeds of 290 to 310 mph did not gather ice in la rge enough 

rn antities to offer anv proof of this contention. Comparislons 
of P-51, B-26 and A-26 icing hint that the Icing may be much 
more Influenced by the character of the airfoil above 200 mph# 

4 , George F. Taylor, A eronautical Meteorology : New York, 

194'^ states: 

"...severe icing in air of continental extraction is very 
rare. . /.Any unstable layers in an air mass of recent mari¬ 
time extraction associated with temneratures below the freez¬ 
ing solnt is sufficient indication for moderate or severe 

icing conditions.” _ . w 

5, Nearly all the radiosonde instruments were Frlez 

or Washington institute of Technology variable audio-fre- 
quenc; type, with plastic temerature and chemical relative 

h.-nidify try Meteorological Officer M.OM. 393 states: 

"At temperatures below 15° F. serious ice accretion is 
mainly limited to convections! clouds." 

In contrast, Taylor, op. clt , warns: 

M It Iibs frequently been stnto.B t*v ‘ trlme icinf$ occurs at 

very low temneratures and that clear Icing is only found at 
temperatures above about 20° F. This as somption is -holly un^ 
founded and cagea of severe cle£r ~ Tclng have many times been 
resorted at 'tomoornTurns well beTow 


14 






7. 


Don McNeal, Ice Formation in the Atmosphere : Jour. 


Aer0 "B. jCi Th e'h (erraan training manual Verelsung, D(Luft)1209: 
Berlin, 1940, gives the following tab!? of effective rise In 
the freezing level due to dynamic heating: 


Airspeed 
200 km/hr 
300 
400 
500 
600 
700 
800 


Increase in F. L. 
150 m. 

300 

500 

800 

1200 

1600 

2100 


The relationshio is assumed to be of the theoretically cor¬ 
rect form 1 T - kv , where T is the dynamic temperature in- 

crease.and^v ^^^^^t^rologloal Office M.O.M. 393/tates 

"In the British Isles true rain ice is very rare, 
the European continent . rain ice is not so rare. 


HOLT ASHLEY 
1st Lt, AG, 
Weather Officer 







THIS DOCUMENT IS BEST 
QUALITY AVAILABLE. THE COPY 
FURNISHED TO DTIC CONTAINED 
A SIGNIFICANT NUMBER OF 
PAGES WHICH DO NOT 
REPRODUCE LEGIBLY.