Skip to main content

Full text of "Surface-Based Observations of Contrail Occurrence Over the US, Apr. 1993 to Apr. 1994"

See other formats


NASA Reference Publication 1404 




Surface-Based Observations of Contrail 
Occurrence Over the U.S., April 1993 - 
April 1994 



Patrick Minnis 

Langley Research Center, Hampton, Virginia 

J. Kirk Ayers 

Analytical Services and Materials, Inc., Hampton, Virginia 

Steven P. Weaver 

88th Weather Squadron, Wright-Patterson Air Force Base, Ohio 



National Aeronautics and 
Space Administration 

Langley Research Center 
Hampton, Virginia 23681-2199 



December 1997 



Acknowledgments 



The data from the Air Force Bases and Army Air Stations were 
supplied by Captain Carolyn Vadnais, United States Air Force. 
The data were transcribed by Theresa Hedgepeth of the Science 
Applications International Corporation, Hampton, Virginia. The 
fuel use data were obtained through Karen Sage of Science 
Applications International Corporation, Hampton, Virginia. This 
research is supported by the NASA Atmospheric Effects of 
Aviation Project Subsonic Assessment Program. 



Available from the following: 

NASA Center for AeroSpace Information (CASI) 
800 Elkridge Landing Road 
Linthicum Heights, MD 21090-2934 
(301) 621-0390 



National Technical Information Service (NTIS) 
5285 Port Royal Road 
Springfield, VA 22161-2171 
(703) 487-4650 



Summary 

Surface observers stationed at 19 U.S. Air Force Bases 
and Army Air Stations recorded the daytime occurrence of 
contrails and cloud fraction on an hourly basis for the period 
April 1993 through April 1994. Each observation uses one 
of four main categories to report contrails as unobserved, 
non-persistent, persistent, and indeterminate. Additional 
classification includes the co-occurrenceof cirrus with each 
report. The data cover much of the continental U.S. 
including locations near major commercial air routes. The 
mean annual frequency of occurrence in unobstructed 
viewing conditions is 13 percent for these sites. Contrail 
occurrence varied substantially with location and season. 
Most contrails occurred during the winter months and least 
during the summer with a pronounced minimum during 
July. Although nocturnal observations are not available, it 
appears that the contrails have a diurnal variation that peaks 
during mid morning over most areas. Contrails were most 
often observed in areas near major commercial air corridors 
and least often over areas far removed from the heaviest air 
traffic. A significant correlation exists between mean 
contrail frequency and aircraft fuel usage above 7 km 
suggesting predictive potential for assessing future contrail 
effects on climate. Additional surface observations and a 
concerted satellite observation effort are needed to accurately 
assess the climatic effect of aircraft condensation trails. 

Introduction 

Condensation trails or contrails have become a 
common feature in the Northern Hemisphere since World 
War n. These anthropogenic clouds represent the most 
visible byproduct of jet fuel combustion at high altitudes. 
The mechanism for contrail formation is complex, 
depending on a variety of parameters including the type of 
jet engine, the sort of fuel, and the ambient temperature and 
humidity (Karcher, 1994; Schumann, 1996). The exhaust 
may produce numerous sulfate aerosols that act as cloud 
condensation nuclei which initiate tiny droplets that 
subsequently freeze. The resulting cloud usually contains 
large concentrations of small ice crystals (e.g., Murcray, 
1970). They generally form at temperatures less than -30°C 
at high relative humidities or below -50°C at moderate to 
low moisture levels (e.g., Appleman, 1953). If formed in 
clear air, contrails can spread, developing into cirrus 
indistinguishable from natural clouds. Their persistence and 
growth depend on the available moisture and ambient 
temperature. When aircraft fly through existing clouds they 
can produce contrails ordistrails (cloud-free trails) depending 
on the conditions (e.g.. Scorer, 1972). In either case, they 
produce an immediate effect by altering the microphysical 
properties of the existing cloud. 



Increases in cloud cover or cloud particle concentrations 
due to contrails can alter the local radiative balance by 
reflecting more solar radiation and absorbing and emitting 
longwave infrared radiation (e.g., Kuhn, 1970). The overall 
effect of contrails on climate depends on a number of factors 
including frequency and timing of occurrence, areal 
coverage, lifetime, altitude, location, and microphysical 
properties. The upper troposphere is a relatively clean 
(aerosol-free) environment so that the addition of high 
concentrations of cloud condensation nuclei have the 
potential for making a larger impact than they would in the 
lower troposphere. With commercial air traffic expected to 
increase by more than 200 percent by 2015 (Baughcum, 
1996), the effects of aircraft exhaust on the atmosphere have 
become a subject of considerable interest leading to the 
NASA Atmospheric Effects of Aircraft Program (AEAP) 
which has sponsored the Subsonic Assessment (SASS) 
Project (Thompson, et al., 1996). One of the goals of the 
AEAP SASS Project is to evaluate the effect of contrails 
on climate. This paper presents the results of a study of 
contrail occurrence frequencies over the U.S based on recent 
surface observations . 

Background 

Evaluations of contrail coverage or occurrence have 
been made either directly or indirectly from surface and 
satellite observations since the 1980's. These efforts have 
been sporadic and generally confined to a few particular 
areas. Examples of inferred contrail coverage include the 
conclusions of Chagnon (1981) and Angell et al. (1984) 
thatdecreasedsunshine and increased cloudiness since 1960 
and between 1950 and 1972, respectively, are attributable to 
contrails. Seaverand Lee (1987) found that the number of 
cloudless days over the continental United States (US) 
decreased significantly for the period 1 950- 1 982 compared 
to 1900-1936 possibly due to the appearance of contrails 
during the latter period. In a follow-up study, Angell 
(1990) found that US cloudiness continued to increase 
through 1988 while sunshine duration decreased. The 
relative magnitude of the change in sunshine was not as 
great as the cloudiness increase. This finding suggests an 
increase in thin cirrus due, most likely, to contrail-generated 
cirrus. Significant decreases in insolation were also 
observed in Germany during the past 20 to 40 years. Weber 
(1990) suggested that increased cyclonic activity increased 
cloud cover and decreased sunshine over Germany. Liepert 
et al. (1994) estimated that contrail coverage, based on a 
surface study of contrails over a single site, was too small 
to account for the diminished sunshine. Discrepancies 
between the conclusions of these various studies highlight 
the uncertainties in the current assessment of the climate 
impact of contrails. 



Satellite data have been used in a variety of ways to 
study contrails over larger areas and longer time periods. 
Joseph et al. (1975) used two photographs from an Earth 
resources satellite to demonstrate the detection of contrails 
from space over the Mediterranean. From preliminary 
studies of Defense Meteorological Satellite Program 
(DMSP) imagery, Carieton and Lamb (1986) found that 
infrared (IR) data were more valuable for detecting contrails 
than visible data. Lee (1989) showed that brightness 
temperature differences between the split window channels 
on the NOAA Advanced Very High Resolution Radiometer 
(AVHRR) could be used to detect contrails more easily than 
simply examining infrared window channel imagery. 
DeGrandet al. (1990) used the single-channel IR imagery 
from the Sun-synchronous DMSP satellites to develop a 
climatology of contrail occurrence over the US for the 
period 1977-1979. Although they provided estimates of the 
relative magnitudes of the mean seasonal and diumal cycles 
over US, the actual frequencies of occurrence were not 
reported. Engelstad et al. (1992) added image processing 
techniques to the brightness temperature differenceimagery 
to automatically detect contrails without human 
intervention. Their method, however, has not yet been 
applied to significant amounts of satellite data. Schumann 
and Wendling (1993) also developed an automated technique 
but they have reported only preliminary results from 99 
AVHRR images over central Europe. Bakan et al. (1994) 
used visual analysis of thousands of quicklook AVHRR IR 
images taken over the northeast Atlantic and Europe to 
estimate contrail cloudiness for 1979-1981 and 1989-1992. 
They found a distinct seasonal cycle with a southward 
displacement of the contrail maximum during winter. 
Maximum contrail coverage in their analysis occurred 
during summer centered along the North Atlantic air routes. 
The coverage increased in that area during the 10-year 
interim. Similar analyses over the air corridors of the US 
have not yet been performed. The surface observations 
reported here represent the first step to better defining the 
contrail-based cirrus coverage over the US . 



Nomenclature 




AEAP 


Atmospheric Effects of Aircraft 
Program 


AFB 


Air Force Base 


AVHRR 


Advanced Very High Resolution 
Radiometer 


c 


mean annual contrail frequency 


DMSP 


Defense Meteorological Satellite 
Program 



IR 


infrared 


LAFB 


Langley Air Force Base 


LT 


local time 


M 


mean contrail frequency 


Max, 


primary maximum hourly 




contrail frequency 


Max, 


secondary maximum houriy 




contrail frequency 


Min 


minimum hourly contrail 




frequency 


NASA 


National Aeronautics and Sp 



NOAA 



Administration 

National Oceanographic and 
Atmospheric Administration 



R 


linear correlation coefficient 


RH 


relative humidity 


SASS 


Subsonic Assessment 


7/ 


monthly average temperature at 
mean flight level 


U.S. 


United States 


US 


Continental United States 


UTC 


Universal Coordinated Time 


WPAFB 


Wright-Patterson Air Force Base 


Zp 


monthly mean tropopause 
altitude 



/ 



mean annual fuel usage above 7 
km, 10^ lbs 



Data 

The U.S. Air Force established a unique network of 
surface observers to develop a climatology of contrail 
occurrences over the US for studying contrail formation and 
forecast models. This dataset was made available to NASA 
in its raw form: log sheets of hourly meteorological 
observations with distinct contrail codes or special log 
sheets used only for recording the contrail codes. The 
contrail observations were listed as one of four 
classifications: no contrails, non-persistent contrails, 
persistent contrails, and indeterminate contrails. In 
addition, each classifier was qualified as being with or 
without cirrus. Finally, the contrail observations were not 
always taken even though weather data were recorded. 
Thus, there is a no-observation category yielding a total of 
nine possible contrail classifications. These contrail codes 
as well as the sky cover in tenths were transposed to a 
computer spreadsheet format for additional analysis. 



The observation network, comprising 19 U.S. Air 
Force Bases and Army Air Facilities, was spread over the 
US (Figure 1) to determine the spatial variability of contrail 
formation. The nominal period of observation for this 
special effort was April 1, 1993 through March 31, 1994. 
The actual period varied with reporting station. The earliest 
month is January 1993, while the latest is May 1994. As 
many as 15 and as few as 7 months of data were taken at a 
given location. Most sites recorded data for 13 months. 

Contrail observations were taken every hour at the 
same time as the standard meteorological data, but only 
during the daytime. Nocturnal data were not recorded 
because of the ambiguities associated with taking high- 
cloud observations in the dark (e.g., Hahn et al., 1995). 
Sky cover was recorded at most locations for much of the 
period, although the special contrail logs were the only 
available data for a few stations. The indeterminate 
category was selected if the sky was overcast below the 
level of cirrus clouds. Persistent contrails were defined for 
this study as those extending at least several miles behind 
the aircraft with no tendency for dissipation. A 
nonpersistent contrail is defined as one that tends to 
evaporate and only extends a short distance behind the 
aircraft. 

Figure 2 shows an example of the raw data in the form 
of a coded summary of the hourly contrail observations for 
Langley Air Force Base (LAFB) taken during December 
1993. At night, no contrail data were taken although cirrus 
occurrences were recorded. Indeterminate contrail conditions 
dominated during at least 10 days, especially December 4, 
5, 15, 16, 28, and 29. A period of persistent contrails with 
cirrus on the 2nd was followed the next day with 
sporadically persistent and nonpersistent contrails and no 
cirrus. A few temporary or nonpersistent contrails were 
seen early during December 6 followed by several hours 
without contrails or cirras. After 1200 UTC, cirrus 
occurred every hour during the next day with a few 
temporary contrails and 2 hours of indeterminate contrails. 



I Contrails, Persistant a No Contrails + Cimis present 

B Contrails, Not Persistant t Indeterminant Contrails 



•Fairchild 



h-1 V, 



_lfMt.Hpme — 

J •Hill - 

Jeale 

5. McOellan ^ 

\ \'' i " f 

^ * Edwards 
^; i •Luke 



. ,^^ — u^ Lorine • 

) y — ^•Griffis 
i >, nWrieht-Pat. / 



•offutr I 

^ •Whiteman 

1- -- ,/ ^ S Langley 

^ker ^^ ., . ^ ^^y^' 

»Barksdale ' ^X^ 



•Kelly 

i 



^!_ii>« Cairns 
-""^^ Egltn • 






1 


+ 


1 












T 
«-■ 

a 


▼ 

+a 

B 
T 


▼ 


» 




+ ▼ 


N 
V 


» 

T 
D 
+ E 

d 

■ 


+ + 


+ 


2 
3 


- 












D 


m a 




-!-■ 






4 
5 




i 


^ 1- 








T 
T 


T 


T T 
T T 


T I 


6 

7 








- - 






.° 


: 

+• 
▼ 

▼ 

B 

T 
+ C 
+ ■( 


B 

+a 
a 
■ 
+1 

T 
° 

▼ 
D 

+! 

T 
+ C 

: 

+ 

▼ 
D 


D 
+ T 

a 

m 


Q a 

fOtT 

■ ■ 


D 

D 

a 

w 


D 
¥a 
a 

T 
▼ 


+d' + 

+B* + 


f 


8 
9 


: -- 


+ + ' 






: 


+ D + 

■ : 




10 






^ 


+ 


+ 


^m 


+ T 

T 

+a 
n 
▼ 
▼ 
B 
▼ 
□ 

±-" 
T 
D 

: 

T 

+- 
▼ 
▼ 
D 
D 


t-TJ+T 








11 














▼ 






12 








\ 


+ 


+ 
+ 


T 
▼ 






D 


D 




13 


+ 


+ +\ 


+ >^+ 


l-B 

n 

T 
T 
fB 
T 
D 
t? 

a 

a 


□ 

T 
T 

to 

ta 
tr 

a 
m 


□ 
▼ 


+■ 


14 


' 


;+ + 


▼ 
T 
T 
tB 
▼ 


T 
+B 
T 








, 


^ 


- 


▼ 


+ 
+ 


17 
IB 


+ ' -(■ ^ + 


+ , + 

, 


L ^ 








▼ 
+ 
+ 

+ 


a 

l-T 
D 

D 


+■; + 


19 

30 


;.. _^+ 


+ +^ 


+ + 


+ + + 


+ 


+ D. _ 


21 
22 


i i 


'" 


. .1 


. - 


T ▼ 


n D 




23 
24 


+ + + 


+ + 


■ 1 




p4-. 

D ▼ 


T 

■_ 
▼ 

T 

D 


T 


T 

a 

T 


T 
D 

▼ 
▼ 

a 

b' 


;g 


..u 


25 




' 


+ +• 


a 

T 


D 1 


26 




; 


.-^. 


+ 


+ 

+ 


,t. 
T 
T 
D 

n 






T' 


27 








T 

a 


T 
T 

a 
a 


T T ' 


28 












T 


T 




29 




■ 








T^ 




+ 


30 


+ :+ + + +1 

: !+ + 






+E + ! + 

lO D 


31 


+ + 


^ |n 




i 2 3 


i 5 


6 7 


8 9 


10 


11 12 13 

UTC 


14 


15 


16 


17 


18 


19 


20 


21 


22 23 


24 



Figure 1. Contrail observation network, comprised of 19 U.S. 
Air Force Bases and Army Air Facilities. 



Figure 2. Coded summary of Langley AFB December 1993 
contrail and cirrus observations. 

The other significant periods of persistent contrails include 
December 9, 10, 14, 20, and 24. Contrails were observed 
for a total of 52 hours. They were not observed during 85 
hours. Persistent contrails account for 65 percent of all of 
the contrail observations. Of these, only 34 percent were 
unaccompanied by cirrus . 

Figure 3 summarizes the 24-hour, daily observations 
over LAFB for the period from March 1 993 through April 
1994. The colors at the top of graph correspond to 
indeterminate (light gray) and contrail-free cases, while the 
black area in the middle registers the percentage of missing 
and nighttime data. Along the bottom of the graph, the 
colors refer to various contrail conditions including 
indeterminate contrails with cirrus (gray at bottom). No 
data were taken during November 1993 and January, 
February, and March 1994. Persistent contrails occurred 
most often during a given day in the spring during both 
1993 (days 110-150) and 1994 (days 465-480) and least 
during the late summer of 1993 (days 190-220). This latter 
period also has the most observations of no contrails with 
cirrus. 

Temperatures and humidities at the standard 
meteorological levels were taken from 1 2-hourly National 
Meteorological Center analyses. These data are available on 
a 2.5° latitude-longitude grid. They were bi-linearly 



>. PeiMslCilt 



Cunlmifs. I'cisisicnl, w'Ciit 
f'<inlr;iils, Noi iViMsicni. u. 
liHli.-li.'milnaic Conliails 
No C'oiilrails 





CMC\jcjc\jc\jc\jcocr)cococr5 
Day Number (3/93 - 4/94) 



Figure 3. Summar>' of daily observations over LAFB for the period from March 1993 through April 1994. 



interpolated to match the location of each surface site. 
Only data from July 1993 and January and April 1994 are 
considered here. 

Results 

The data for each for the contrail categories and the 
cloud observations were averaged for each month and hour 
in Coordinated Universal Time (UTC). The seasonal, 
regional, and diurnal variation of these averages is given 
below. 

Seasonal Variations 

Examples of the monthly means are plotted in Figure 4 
for LAFB. Contrails occurred most frequently during April 
1993 as ~32 percent of the total observations (Figure 4a). 
A similar number of contrail observations was reported 



during the following April. Nonpersistent contrails were 
most frequent during June 1993 and April 1994. The 
fewest contrails (~5 percent) were seen during July 1993 
when cirrus clouds were most abundant. The worst viewing 
conditions were found during March and December 1993 
when the indeterminate levels were greatest. Figure 4b 
combines the categories into four classes that do not 
consider the occurrence of cirrus. The indeterminate 
classifications, more easily discerned in this figure, 
generally correspond to the occurrence frequencies of 90 and 
100 percent cloud cover (Figure 5a). If indeterminate data 
are removed and the classifications are normalized to the 
number of remaining observations, the relative temporal 
pattern of contrail occurrence remains much the same except 
for the substantial increase in contrails during December 
(Figure 4c). These normalized percentages may be a more 
accurate accounting of the contrails because the 
indeterminate data are almost equivalent to missing data. 




J FMAMJJASOND JFMAM 
Month 




J FMAMJ J ASOND JFMAM 
Month 



1 
0.8 

|0.6 

gO.4 

0.2 

c) 




Vy 



^ 



M U . ,m 



% 



H 


Indeterminate Contrails, w/Ci 


(3 


Indeterminate Contrails 





No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


m 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



JFMAMJJASOND JFMAM 
Month 



Figure 4. Summary of monthly observations for Langley AFB, Virginia from January 1993 to May 1994: (a) relative frequency of 
contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus with 
indeterminate data removed. 



The months having the most clear days are October and 
December 1993 and April 1994 (Figure 5a). Results 
corresponding to Figure 4 are provided in Appendix A for 
each observing site. The missing data evident in Figure 5 
for November 1993 and for January, February, and March, 
1994 are not typical for most of the sites. Better sampling 
was obtained for 14 other sites. 

Table 1 lists the mean contrail occurrences for each site 
and the corresponding period of observation. Figure 6 
shows the means and standard deviations based on monthly 
averages of the combined persistent and nonpersistent 
contrails from data like those in Figure 4a which includes 
indeterminate data. The fewest number of persistent 
contrails occurred over Eglin AFB (3.6 percent) and Minot 
AFB (3.8 percent), while the greatest number were seen 
over Wright-Patterson AFB (WPAFB; 15.1 percent) and 
Edwards AFB (14.9 percent). If indeterminate cases are 
omitted, then contrails were most frequent over WPAFB 
(28.5 percent) and LAFB (19.9 percent). Witiiout the 
indeterminate observations, the sites with the fewest 
persistent contrails are Eglin (5.1 percent) and Luke (5.4 
percent) AFB's. Nonpersistent contrails were most often 
observed over Luke AFB (9.2 percent) and LAFB (~5 
percent), while the fewest were seen over Kelly AFB . Both 
persistent and nonpersistent contrails are most likely to 
occur when cirrus clouds are present. The mean probability 



0.35 




^ 90% to 100% 
H 70% to 80% 
50% to 60% 
^ 30% to 40% 
010% to 20% 

□ o% 



= 20 
u 

0° 



FfWaMj'jIaIs'oInIdIj'fIm'a'm' 

Month 



Figure 5. Summary of cloud cover for Langley AFB, Virginia 
from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency 
centered at local noon, and (c) monthly mean cloud cover. 




Figure 6. Means and standard deviations of combined 
persistent and non-persistent monthly contrail frequencies, 
sites ordered north to south according to latitude. 



for seeing a contrail when cirrus are present is 24.8 percent, 
compared to only 7.3 percent when cirrus are not observed. 
Contrails are also more likely to persist vvhen cirrus are 
present. When cirrus and contrails occurred together, 75 
percent of the contrails were classified as persistent 
compared to 55 percent when cirrus were absent. The 
greatest monthly variability (Figure 6) occurs over 
McClellan AFB, while contrail occurrence over Eglin is 
consistently low from month to month. 

Figure 7, which summarizes the results in Appendix 
A, reveals that the maximum contrail occurrence occurs 
between January (Figure 7a) and April (Figure 7b) for most 
sites. A notable exception is Minot AFB where October 
(Figure 7d) is the most favored month for contrails. 
Minimum contrail occurrence is generally found during July 
(Figure 7c). The seasonal variations in contrail events 
averaged for all sites are shown in Figure 8. If the events 
are referenced to the total number of observations (Figure 
8a), then there is a distinct maximum during February and 
an apparent secondary peak during October. Although the 
October value is statistically diiferentfrom the September 
mean, it does not differ significantly from the November 
result. Overall, contrails are scarcest during July. If 
indeterminate observations are excluded from the total 
(Figure 8b), the seasonal curve becomes smoother. The 
maximum occurs during March or between February and 
March with no secondary peak during October. Regional 
variability is considerable. In absolute terms, it is greatest 
during February. If computed relative to the mean 
occurrence values, the greatest geographical variability 
occurs during July and November. 




5 10 15 20 25 30 35 40% 



Figure 7. Comparison of persistent contrail frequency with 
indeterminate data excluded for several months, a) January, 
b) April, c) July, d) October. 

Diurnal Variability 

Figure 9 depicts the diurnal variations in contrail 
frequency over LAFB. The greatest number of contrails 
was observed (Figure 9a) during midmoming at 1400 UTC 
(0900 LT). A broad secondary maximum covers the period 
from 1800-2100 UTC. Normalization to the total number 




Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 
Month 




Jan Feb Mar Apr Vfay Jun Jul .Aug Sep Oct Nov Dec 
Month 



Figure 8. Monthly mean persistent contrail frequency for 19 
sites. Bars indicate standard deviations: a) All daytime data, 
b) Indeterminate data excluded. 

of observations (Figure 9b) reveals a substantial peak at 
2400 UTC. This maximum may not be representative, 
however, because it is based on a limited sampling of ~1 10 
observations (Figure 9a) that are confined to particular 
months. This number of observations is about half of the 
total between 1200 and 2200 UTC. The differences in 
hourly sampling are primarily due to changes in the day 
length with season. It is important to consider the 
sampling when examining these diurnal results. If only 
those hours having more than 150 samples are considered, 
the peak contrail frequency remains at 1400 UTC and the 
fewest contrails occur at 1200 and 1700 UTC. When 
indeterminate data are removed (Figure 9c), the relative 
diurnal cycle is the same although the morning maximum 
is enhanced slightly. The primary minima occur at 1200 
(0700 LT) and 1700 UTC (local noon). Plots of the mean 
hourly contrail statistics are provided in Appendix B for 
each observation site. 

By excluding the indeterminate data and using only 
those hours with more than half the maximum number of 
hourly samples for each site, it is possible to determine the 
primary and secondary diurnal maxima in contrail 
occurrence. Here, the primary maximum Max, is defined 
as that hour having the greatest contrail frequency. The 
secondary maximum MaX; is the hour with the next 
highest frequency that is at least 3 hours removed from the 
primary maximum. The amplitude of these maxima is half 
of the difference between a given maximum and the 
minimum Min divided by the mean total contrail 
occurrence M. In Figure 9c, Max, at 1400 UTC is 39 
percent and Min is 20 percent at 1800 UTC. The 




101112131415161718192021222324 1 
UTC 




101112131415161718192021222324 1 
UTC 



Q Indeterminate, w/Ci 

KJ Indeterminate 

0None,w/Ci 

□ None 

Q Not Persistent, w/Ci 

Not Persistent 
Q Persistent, w/Ci 

Persistent 




c) 



101112131415161718192021222324 1 
UTC 



Figure 9. Summary of hourly contrail and cirrus observations 
from Langley AFB, Virginia, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, 
and (c) relative frequency of occurrence with indeterminate 
data removed. 

amplitude of this maximum is 40 percent because M is 24 
percent (Table 1). The secondary peak at 1800 UTC with a 
magnitude of 19 percent may not be particularly significant. 
Table 2 summarizes the diurnal characteristics of contrail 
frequency for each site. Loring AFB has the greatest 
amplitude with Max, in the evening. The smallest 
amplitude is 23 percent at Barksdale AFB, also with an 
early evening maximum. The average primary and 
secondary amplitudes are 58 and 46 percent, respectively. 
Amplitude does not appear to have a longitudinal 
dependence. Plots of the primary and secondary maxima 
times in Figure 10 also show some interesting features. In 
Figure 10a, the primary maxima are concentrated between 
1500 and 1800 UTC and between 2300 and 0100 UTC, 
while many of the secondary maxima occur between 1800 
and 2300 UTC. When converted to local time (Figure 10b), 
the primary maxima cluster around 0900 and 17(X) LT, 
while the secondary peaks primarily occur during the 
afternoon. Figure 1 1 , a plot of the maxima as a function of 
longitude, shows that, except for Loring AFB, the times of 
primary maximum are found near 0830 LT over the eastern 
US, during the late morning or late afternoon over the 
central US, and during the early morning or late afternoon 
in the west. The secondary maxima generally fall in the 
afternoon except for some of the westernmost sites. 




7 

6 

^5 

14 

|3 

£2 

1 

(b) 



12 13 14 15 16 17 18 19 20 21 22 23 24 25 
UTC(hr) 



Primary |[[] Secondary 



si 



iiliMlk 



M 



Li 



OiJlfi, 



6 7 8 9 10 11 12 13 14 15 16 17 18 
Local Time (hr) 



Figure 10. Times of maximum contrail occurrence for the 19 
study sites. 



20 





18 


T- 


16 


e 




u 
E 


14 






H 






12 


^ 


10 




8 





* 


Primary A 


Secondary 






. 








* 


- 




A 






-■» 


■Sk 4- 








* 


A A 




A 




* A 




A 




A 


- 


A 








A A 


A 




A 




— 


* * 


A 






A 


Jt 








-A * 


* 








A ■{!■ 




«}■ 


« 




-A.* 




* 


* 




- , , , 


1 , 1 


1 1 1 


1 


J 



125 115 105 95 85 

West Longitude (°) 



75 



65 



Figure 1 1 . Longitudinal dependence of times of maximum 
contrail occurence 

Cloud Cover 

The fractional cloudiness frequencies and monthly mean 
area! cloud coverage are given in Figtire 5 for L AFB. The 
monthly frequency of a given cloud fraction (Figure 5a) was 
discussed earlier. The mean annual diurnal variation of each 
fractional cloud amount is shown in Figure 5b. The times 
are given in UTC, but the abscissa scale is shifted such that 
the hour closest to local noon is at the center of the graph. 



In this particular case, cloud observations were taken for the 
complete diurnal cycle. Clear skies are most frequent near 
local midnight (5 - 8 UTC) and seen least often around local 
noon (17 - 21 UTC). Both fractional and overcast 
cloudiness follow a diurnal cycle complementary to the 
clear skies. The monthly mean cloudiness in Figure 5c 
varies from a low in June 1993 to a peak during March 
1993. The missing months prevent a more complete 
determination of the annual cycle. Data corresponding to 
Figure 5 have been plotted in Appendix C for each site. No 
cloud observations were available at five sites: Griffls, Hill, 
Kelly, McClellan, and Tinker Air Force Bases. Complete 
24-hour sampling is available at the remaining sites. 

The monthly mean cloudiness for the 14 sites with 
observations is summarized in Figure 1 2 with the mean US 
surface-observed cloud amounts from Hahn et al. (1986) for 
the period 1971-1981. The average for both datasetsis -55 
percent. In both cases, cloud amount peaks during the 
winter with a minimum during late summer and autumn 
although the annual range for the contrail dataset is ~26 
percent compared to 12 percent from the US average. 
Figure 13 shows the monthly averages of mean cirrus 
frequencies. These results show that cirrus was observed 
over the surface sites least often during the winter and most 
frequently during the late autumn months. Thus, the cirrus 
frequencies are slightly out of phase with the corresponding 
cloud amounts. Overall, cirrus was observed in ~55 percent 
of the observations. Figure 13b also plots seasonal mean 
US cirrus frequencies based on surface observations during 
1971-1980 (Hahn et al., 1984). The seasonal means 
correspond well to the present results for winter and spring. 
Cirrus clouds were observed 4 and 10 percent more often in 
the current dataset than in 10-year average for summer and 
autumn, respectively. The general agreement between the 
cloud amounts and cirrus frequencies suggest that the 
selected sites are representative of the US as a whole. 

The mean monthly variations of persistent contrails 
with and without cirrus are shown in Figure 14. In general, 
both curves follow the total contrail seasonal trends seen in 
Figure 8b. The ratio of seasonal contrail frequencies with 
cirrus to those without cirrus are 3.1, 4.3, 3.6, and 5.1 for 
winter, spring, summer, and fall, respectively. These co- 
occurrence ratios are consistent with the minimum and 
maximum frequencies of cirrus in Figure 13b. Figure 15 
shows the contrail frequencies as a function of coincident 
cloud amount averaged for the sites reporting hourly cloud 
amounts. Few persistent contrails were seen in otherwise 
clear skies (Figure 15a), while cloud amounts near 75 
percent correspond to the most frequent occurrence of 
persistent contrails. This result is consistent with the 



70 



1971-81 




20 



6 
Month 



10 



12 



Fig. 12. Comparison of U.S. climatological average and 
contrail-data monthly mean cloud cover. 




on 

ON 

I 



[J- S < 5 ' 



3 
< 






2? 


^ts 


S^^S^ 


I 







70 



^60 

|50 



Month 



tt. 



U 



30 



AFB sites 




40 L US seasonal mean (1971-80) 



b) Monthly averages 

J I 1_J I L_J I I 1 L. 



6 

Month 



10 



12 



Figure 13. Frequency of cirrus occurrence over the contrail-data 
sites and the U.S. climatological mean. 



o 

c 



0.2 



0.15 



t 0.1 

o 

O 

I 0.05 

■(0 

Q. 




- PC w/o Cirrus 



~i — I — i — r 

JFMAMJJASOND 
Month 



Figure 14. Monthly mean persistent contrail frequency with 
and without cirrus for all 19 surface sites. 

frequent occurrence of contrails with cirrus. Conversely, 
non-persistent contrails were observed most often in almost 
clear skies (Figure 15b) suggesting that the nonpersistent 
contrails form primarily in drier conditions. When all 
contrails are considered, the maxima arise in mostly cloudy 
and partly cloudy conditions with minima in almost clear or 
overcast conditions and when the cloud fraction is around 50 
percent (Figure 15c). 

Discussion 

The occurrence of contrails is primarily determined by 
two factors: the presence of aircraft exhaust and the 
ambient conditions at flight level. An observation of a 
contrail requires both proper timing and a sufficient line of 
sight from the observer to the contrails. These factors and 
their relationships to the observations are discussed here. 

Aircraft Fuel Use 

A preliminary dataset of fuel usage was developed from 
the estimates of commercial scheduled and other non- 
scheduled and military air traffic by Baughcum et al. (1993) 
for May 1990. The data were compiled on a 1° x 1° 
latitude-longitude grid with a vertical resolution of 1 km. It 
was assumed that the May data are representative of the 
annual mean. A later analysis by Baughcum (1996) 
confirms that assumption. Figure 16 shows the mean fuel 
usage as a function of altitude for the nine 1 ° boxes centered 
over the 19 contrail sites. 

The maximum fuel use above the boundary layer 
occurs at flight levels between 10 and 12 kilometers. 
(33,000 - 38,000 ft) for all of the sites. The low-altitude 
fuel is primarily expended on the runway and during 




1 I \ I I i i I I r 

10 20 30 40 50 60 70 80 90 100 
Cloud Amount (%) 




1 I I I I I I r 

10 20 30 40 50 60 70 80 90 100 
Cloud Amount (%) 




1 — I — I — I — i — 1 — r 

10 20 30 40 50 60 70 80 90 100 
Cloud Amount (%) 

Figure 15. Contrail frequency as a function of cloud amount: 
a) Persistent contrails only, b) Non-persistent contrails, 
c) All contrails. 



o 
•o 

3 

■a 




Fuel (10^ kg) 

Fig. 16. Vertical distribution of annual mean fuel use for 3° 
latitude-longitude boxes centered over the 19 U.S. contrail 
observation sites based on May 1990 data from Baughcum et 
al. (1993). 

takeoffs. To account for all flight levels in which most 
contrails are likely to occur, the data were summed for all 
altitudes above 7 km. The distribution of the sums in 



10 



Figure 17 reveals the main flight corridors over the US 
with a primary maximum over the Midwest between 
Chicago and New York. Other routes to Florida, Atlanta, 
Dallas, and southern California from the northeast and 
Midwest are also evident. The geographical variation of 
contrail occurrence on an annual basis (Figure 18) roughly 
coincides with the fuel data in Figure 17. For example, 
maximum contrail frequency occurs over WPAFB in the 
heart of the Chicago-New York corridor. Offut and Edwards 
AFBs, which have relatively high frequencies, are under the 
Chicago-Los Angeles jetways and Eglin, Loring, and Minot 
AFBs, where contrails are not often observed, are off the 
main air thoroughfares. Presumably, the discrepancy 
between the BealeandMcClellan AFB contrail frequencies 
occurs because the latter is closer to the edge of the San 
Francisco- East Coast airway. 

Although the quantity of consumed jet fuel probably 
increased between 1990 and 1993, the relative pattern of air 
traffic likely changed little during the interim. Therefore, it 
should be possible to correlate the 1990 fuel usage with the 
data in Figure 18 to determine the relationship between the 
fuel use and contrail occurrence. A surface observer can see 
high-altitude clouds that are a considerable distance from the 



surface position. Furthermore, contrails can advect rapidly 
from their formation location. For example, the 
climatological mean zonal wind velocities at 300 mb range 
from -40 km/hr in July to neariy 100 km/hr in January 
between 30°N and 45''N (Sadler, 1977). To determine how 
fuel usage relates to contrail frequency, the effective 
viewing area for the surface observer's houriy reading must 
be determined. This area was estimated by correlating the 
mean annual contrail frequencies to fuel-use averages 
computed from arrays of 1° boxes. It was assumed that the 
optimal area corresponds to the array size yielding the 
greatest correlation coefficient. A 3° box centered at each 
site produced a linear correlation coefficient, R = 0.73, the 
maximum correlation between mean fuel use for a square 
anay and persistent contrail frequency. The values of R 
for 1°, 5°. and T boxes are 0.64, 0.62, and 0.41, 
respectively. Total contrail frequency including both 
persistent and non-persistent contrails shows a stronger 
relationship to fuel usage with R = 0.78. Scatterplots and 
linear regression fits forced through the origin are shown in 
Figure 19 for the 3°-box fuel averages and total and 
persistent contrail frequencies determined without 
indeterminate data. 




O.OxloOO S.OxIqO^ l.OxloO^ 1.5x10^8 2.0xloO^ 2.5x10^ 
Fuel Use (lbs) 
Figure 17. Sum of aircraft fuel usage for altitudes greater than 7 kilometere. 



11 






5 10 15 20 25 30 35 40% 
Figure 18. Annual mean persistent contrail frequency excluding indeterminate data for 19 surface observation sites. 



According to these fits, the mean annual total and persistent 
contrail frequencies are 



c, = 0.00176/, 



cp = 0.00127/, 



and 



(1) 
(2) 



respectively, where / is the mean fuel use above 7 km in 
the nine 1° boxes centei^ed over a given location. Fuel 
consumption is given in 10^ lbs yr"'. While these 
correlations demonstinte the obvious, that the likelihood of 
observing a contrail increases as the number of planes at 
altitude increases, it also quantifies, for the first time, the 
relationship between aircraft fuel usage and contrail 
frequency, Moreover, it shows that over the US, fuel 
expenditure can account for 61 percent (/?^) of the variance 
in mean annual persistent contrail occurrence and that 
contrail occurrence will increase as air traffic volume 
expands. 

Using the mean 3° regional US fuel usage above 7 km, 
4.8 X 10^ lbs yr"', in Eq. (1) yields a mean occurrence 
frequency of 0.085 for the country as a whole. This result 
suggests that, on average, an observer situated at a random 
location and time in the US will have an 8.5 percent chance 
of seeing a contrail if the sky is not totally obscured. 
Before the commercial jet age began in earnest during the 



I960's, contrails were a rare sight. In some regions like the 
midwestem US, especially during winter, the likelihood of 
observing a contrail is on the order of 40 percent, an almost 
every-other-day occurrence. 

Meteorological Conditions 

Fuel use cannot account for all of the variability in 
contrail occurrence. Most of the remaining variance is 
probably due to the diverse temperature and iielative 
humidity RH conditions at flight level, although engine 
and fuel type as well as the operating conditions also 
influence contrail formation. While a detailed examination 
of the meteorological conditions affecting contrail 
occurrence is beyond the scope of this study, monthly 
averages of certain parameters are examined to demonstrate 
how atmospheric profiles may affect the contrail frequencies 
in this dataset. 

Because of the typically low relative humidities in the 
stratosphere, a plane is unlikely to produce a significant 
contrail if it flies above the tropopause. Figure 20 shows 
the variation of mean tropopause altitude Zp with 
observing site for 3 months. During January 1994, Zp 
varies from 9.6 km in Maine to over 12 km in Texas. 
Most of the heights are between 10.5 and 11.5 km. The 
tropopause height generally increases during April 1994 to 
between 10.8 and 14 km. During July 1993, the range is 
10.8 to 16 km. If 10.5 km is assumed to be the average 
flight level, then most of the air traffic over the US takes 



12 



place in the troposphere, even during much of the winter. 
The two exceptions are Loring AFB and Minot AFB where 
Zp = 9.6 and 10.1 km, respectively, during January 1994. 

As seen in Appendix A, the maximum contrail 
frequencies primarily occur during the winter and early 
spring months except over these two sites. The maximum 
for Loring occurs during May and June, while the peak 
contrail frequency over Minot is seen during October. The 
lowered tropopause during winter in northern latitudes is 
also the likely source for the southward displacement of the 
contrail maximum during winter over the North Atlantic air 
traffic routes reported by Bakan et al . ( 1 994). 

The mean temperatures Tf at the average flight level 
provide further explanation of the seasonal variability. In 
Figure 21, Tf increases from January through July at all 



18 
17 
16 
15 

£ 13 

60 

11 

10 

9 



: 




'- 


July 93 ^^ _*.__-^ 


1 1 1 1 


Jan. 94/ 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 



.11 



I^li 



a 



AFB 






PQ 



0.5 



0O.4 

c 

3 



0.3 



1 
gO.2 

U 

^0.1 



a) 



- 










~ 


^i - 


= 0.00176/ 




- 


t 

r = 


0.78 




• / 


'- 






•^ 


• 


_ 


• 




\y^ 


• 


- 


t 


9 


r^» 




~ 


> 


^.- 






-^ 


I 1 


1 I 1 


1 1 1 J 1 1 1 1 1 


III 1 1 1 1 1 



50 100 150 200 250 

Fuel Use (lO^lbs/yr) 



0.5 



c =0.00127/ 
P 




I I I I — I i I I I I I I I I I I I I I I I I I ■ 



50 100 150 200 250 
Fuel Use (lO^lbs/yr) 

Figure 19. Contrail frequency as a function of fuel use: a) Total 
contrail frequency, b) Persistent contrail frequency. 



Figure 20. Variation of mean tropopause altitude for July 
1993, January 1994, and April 1994. Sites ordered North to 
South according to latitude. 

locations. During July 1993, Tf is greater than 225 K 
over all sites but Fairchild AFB in Washington. The mean 
flight level temperatures are 225 K or less during January 
1994. According to Hanson and Hanson (1995), contrail 
formation at 10.5 km or near 250 mb requires temperatures 
lower than 226 K (-47°C)for RU less than 100 percent. 
As the temperature decreases, the relative humidity required 
for contrail formation also decreases. Thus, the probability 
for contrail occurrence increases as the temperature drops. 
The winter maximum in contrail frequency, therefore, is 
primarily due to the colder temperatures at flight level. 

Contrails were observed over all of the sites during 
July 1993 when Tf generally exceeded 225 K. The non- 
zero contrail occurrence may be attributed to variations in 
Tf over the month or to contrails occurring at higher 
levels. The variability in Tf can probably account for 
July contrails over the northern sites but not over the 
southern locations. The Hanson and Hanson (1995) 
calculations indicate that contrails can form at temperatures 
as warm as 244 K but only in very moist conditions at 
much lower altitudes. For example, they found that the 
critical temperature for contrail formation at 5(X) mb for a 
low bypass engine is -40°C at JUl = 70 percent with 
respect to liquid water. Thus, contrails may be formed 
when jet aircraft fly through moist layers at lower altitudes. 
However, this phenomenon is not likely to occur during the 
summer. As seen in Figure 16, there is still considerable 
air traffic at 12 km. Thus, these flights at the higher 



13 



250 




210 - 



r - 1 I I I I I I I I I I I I I I I I 






^ I ^ a = 



•c 






1^ 

EQ 



AFB 




I I I I I I I I I I I I I I I 



s :g 5P 



•S 2 -5 



I I I I I I I I I I I I I I I 



^3 OU ,^' ?! 3 



^^s 

I 



•c 



■a a 



CO 



AFB 



Figure 21. Comparison of temperatures at 10.5 kilometer 
altitude for July 1993, January 1994, and April 1994. Sites 
ordered North to South according to latitude. 

altitudes (colder temperatures) are probably the source of 
contrail development over the southern sites during July. 

Figure 22 shows the mean relative humidities for each 
site during the 3 months considered earlier. These values 
generally range between 33 and 42 percent. July appears to 
be the driest month overall. The increased RH during 
January and April is not much greater than the July values, 
however. Thus, these humidity values are not likely to 
explain much of the variation in contrail occurrence. 
Despite this apparent lack of humidity dependence, the 
frequent co-occurrence of contrails and cirrus is a clear 
indication that contrails form more often when water vapor 
is more abundant. The absence of an association between 
RH and contrail occurrence may reflect the oft-expressed 
need for better measurements of humidity in the upper 
troposphere (e.g., Schumann, 1995). 

Earlier Contrail Observations 

As noted earlier, contrail observations over the US 
have been limited to either a small area or time period 
except for the satellite study by DeGrand et al. (1990). 
That brief report discussed the occurrence frequencies and 
found the greatest density over southern California, central 
Arizona and New Mexico, and over the Midwest. The 
current results are consistent with that finding except for the 
southwestern US relative maximum. This discrepancy may 
be due to a change in flight patterns, the location of Luke 
AFB (the only southwestern site) south of the primary 
flight corridors (Figures 1 and 17), or to differences in the 
interpretation of persistent and non-persistent contrails. 



Figure 22. Comparison of mean relative humidities at 10.5 
kilometer altitude for July 1993, January 1994, and April 
1994. Sites ordered North to South according to latitude. 

It is unlikely that flight patterns have changed considerably 
since 1978 and, although this site is not immediately 
underneath the flight corridors, a considerable amount of 
fuel is expended within 150 km of Luke AFB (see Figures 
1 and 17) resulting in the potential for many advected 
contrails. If non-persistent contrails are included for Luke 
AFB, the overall frequency of occurrence increases from 5.4 
percent to 14.6 percent, a value more consistent with the 
fuel usage around Luke and close to values obtained for the 
nearest contrail site, Edwards AFB. Given that the 
frequency of non-persistent contrails over Luke AFB is 
more than twice that of any other base and that satellite 
observers generally only see persistent contrails, it is 
possible that the observers at Luke AFB used a different 
criterion for determining contrail persistence than those at 
the other sites. Nevertheless, if the non-persistent contrails 
are included in the average, then the current results are 
qualitatively consistent with the earlier satellite study. 
DeGrand et al. (1990) also found a seasonal variation that 
differs from the present results. Their maximum frequency 
occurs during October with a minimum during July. In 
Figure 8, the maximum contrail frequency occurs during 
February or March regardless of consideration of 
indeterminate data. A minor, statistically insignificant, 
secondary peak is evident during October (Figure 8a) in the 
raw data but it is less than both of the January and April 
values. If indeterminate data are not considered, October 
ranks seventh for contrail occurrence. This difference from 
the DeGrand et al . ( 1 990) results is difficult to reconcile. It 
is possible, but unlikely, that the difference is due to 
sampling. Although only an average of 13 months of data 



14 



was used for the present study, the seasonal results are 
consistent for almost all sites over the US and with the 
meteorological trends. Whether this seasonal cycle is 
typical on a climatological scale is a question that can only 
be addressed with further observations. 

The discrepancy between the DeGrand et al. (1990) and 
the current results, however, may arise from the differing 
viewing perspectives of the satellite and surface observers. 
The surface observer can easily recognize a thin contrail 
shortly after its formation. If it undergoes significant 
growth, however, it may be interpreted as a cirrus by a 
surface observer but could still be recognized as a contrail in 
infrared satellite imagery, especially against a warm 
backdrop, because of its linear characteristics. Thus, 
advecting contrails that have had time to grow may often be 
identified as cirrus clouds by a surface observer while 
counted as contrails in infrared image analysis. It is 
possible that much of the excess cirrus cloudiness detected 
in the autumn surface observations relative to climatology 
(Figure 13b) may be contrail cirrus rather than natural 
cirrus. Such an interpretation would be compatible with 
the results of Angell (1990) who found that cloudiness 
increases over the US were greatest during autumn and were 
most likely due to thin cirrus. 

Conversely, contrails may be easily distinguished 
against a blue sky from the surface but may be difficult to 
detect in morning and evening satellite imagery when the 
background is cold. Because contrails are optically thin, 
they provide minimal contrast in infrared imagery unless 
the background is significantly warmer than the cloud. 
B ispectral brightness temperature difference techniques (e .g . , 
Lee, 1989) rely more on the contrail's small particle size 
and should be more effective for detecting contrails in low 
contrast conditions. Because the Degrandet al. (1990) data 
were limited to a single infrared channel, it is probable that 
many contrails were not detected during winter and early 
spring when the background is significantly colder than it is 
during summer and autumn. This contrast problem would 
be exacerbated by the lack of DMSP satellite data during the 
afternoon when the contrast between clouds and the surface 
is greatest during all seasons. Furthermore, because the 
satellite resolution is between 1 and 8 km, only the largest 
persistent contrails can be detected. It is possible that the 
differences in the seasonal cycle of contrails are due to a 
combination of viewing perspective and contrail growth. 
Coordinated surface and satellite analyses would be needed 
to better reconcile the differences . 

Diurnal Variability 

Because the observations were limited to sunlit hours, 
the diurnal variations are incomplete and the mean 
frequencies may be in error. Commercial aircraft frequently 



operate at night over the US, especially before local 
midnight and after 0600 LT. In particular, contrails from 
early morning flights will be missed in the observations 
during the winter months. Inclusion of nocturnal 
observations could change the results for the diurnal 
statistics and would affect the overall mean frequencies. 
Because air traffic is generally heavier during the daytime, 
the times of maximum contrail occurrence found here are 
probably accurate for most of the sites. The minimum 
hourly frequencies, however, would probably be lower if 
24-hour observations were used. For the same reasons, the 
mean contrail frequencies would be smaller than the current 
daytime values. 

The diurnal maxima seen in Figure 1 1 may reflect, to 
some extent, the timing of flights over the US. Contrails 
observed over a particular site are probably due to flights 
that originated or terminated at least one half hour from the 
site because of the time needed to reach or descend from 
cruising altitude. Primary morning maxima over the east 
(75-85°W) suggest that most flights commence early in the 
day before 0800 - 0900 LT. The secondary afternoon 
maxima are probably the result of later originating flights 
and the arrival of eastbound flights. A similar breakdown 
of flights occurs over the west coast with a mixture of 
afternoon and morning primary maxima. The number of 
long distance flights in either direction plus the north-south 
traffic would shift both the primary and secondary peaks to 
the late morning and late afternoon over the center of the 
country. The exception to this general pattern is Loring 
AFB which is primarily affected by US-European air traffic. 
Much of the eastbound traffic originates during the late 
afternoon and early evening for morning arrival while the 
westbound flights arrive earlier in the day. While the 
connection between flight times and contrail occurrence is 
complex and cannot be fully explained here, the 
observations are consistent with the general constraints 
imposed by commercial air traffic. A complementary 
analysis of satellite data covering the entire day would help 
complete the depiction of contrail diurnal variability. 

Other Considerations 

Relying on the interpretation of surface observers, 
these data are subject to some errors based on the judgment 
of a particular observer. Distinguishing a contrail from a 
natural cloud can be difficult a short time after the contrail's 
formation. As a consequence, some bias toward 
underestimation of contrail frequency is probable because 
only those contrails that can be confidently identified will 
be included in the observations. It is unlikely that any 
older contrails missed in the statistics will be offset by 
cirrus clouds mistaken as contrails. The threshold between 
the persistent and non-persistent contrails is also subjective 
to some degree and will result in uncertainty in the actual 



15 



ratio of persistent to non-persistent contrails. It was noted 
earlier that the high incidence of non-persistent contrails 
over Luke AFB compared to other sites may be the result of 
different criteria used for determining persistence . 

The viewing conditions can also influence the detection 
of contrails. Observations represent the conditions at one 
instant. Therefore, the line of sight to contrails may be 
blocked at a given time by scattered or broken cloud 
conditions. In overcast situations, this effect is recognized 
in the indeterminate category, but it is possible that 
contrails can remain unobserved in other circumstances. 
Although Figure 15 indicates that persistent contrails were 
most often seen during mostly cloudy conditions, it is 
possible that the actual frequency is even greater due to line- 
of-sight obstruction. This effect would also cause 
additional underestimation of the true contrail frequency . In 
this analysis, it was assumed that the frequency in the 
indeterminate cases is the same as in the determinate 
viewing conditions. This assumption has not been tested 
yet. Proper accounting for the cases in which the contrails 
are potentially obscured will require analyses of satellite 
data coincident with the surface observations. 

Concluding Remarks 

This paper provides the most complete inventory of 
contrail frequencies over the US to date. It is just the first 
step, however, in assessing the impact of aircraft 
condensation trails on climate. Because only 1 year of 
observations was available, it is not possible to 
unequivocally conclude that this dataset is a reliable 
climatology of US contrails. Data from other years are 
needed to develop such a climatological picture. Much 
additional information is also required to confidently 
estimate the radiative effects of contrails. Statistics 
regarding the lifetimes, areal coverage, optical properties, 
and advection of contrails and contrail cirrus are essential to 
properly characterize changes in the mean radiation fields. 
Assessment of climate change due to increasing air traffic 
appears feasible, however, because of the strong 
relationship found between fuel usage and contrail 
occurrence and the consistency between seasonal 
meteorology and contrails. The contrail dataset presented 
here can be exploited to refine the relationships between 
contrails and fuel consumption and meteorology. 
Correlations between temperature and humidity from 
soundings coincident with hourly contrail observations will 
be critical to empirically quantify the trends found here. A 
more detailed seasonal analysis of fuel use and contrail 
frequency could be useful for simulating contrail occurrence 
over a given location in the US. With supplemental 
nocturnal data, it may be possible to realistically simulate 
the diurnal cycle as well. The hourly observations can also 



serve as validation data for coincident satellite retrievals of 
contrail occurrence. Combination of satellite retrievals and 
these surface observations will be required to completely 
depict the entire diurnal cycle. 

Contrails have become a prevalent feature of American 
skies. The relationships established here indicate that they 
will become even more common in the future as airline 
service expands. Because contrails add directly to cloud 
cover, they will affect the radiation budget at some 
magnitude. Even if the impact is determined to be small on 
a global scale, the local effects may still be substantial. 
Thus, it is important to determine the relationships between 
contrail frequency and changes in cloud cover. This surface 
analysis of contrails should also be repeated in a few years 
when air traffic has increased significantly to detect any 
changes in contrail occurrence and to test any 
prognostication schemes developed from this dataset. 
Commercial air traffic is growing worldwide with the 
potential for an increase in contrails over many areas 
outside of the US. To fully assess contrails on a global 
scale, their detection and reporting should be made a routine 
part of standard meteorological observations . Because they 
are a distinct type of cloud, they could easily be included as 
part of the cloud type codes currently used in the global 
meteorological observing system. An accurate evaluation 
of the climatic impact of contrails will require an effort that 
combines surface and satellite observing systems. 



NASA Langley Research Center 
Hampton, VA 23681-0001 
December 2, 1997 



16 



Appendix A 
Monthly Mean Contrail Frequencies for Each Site 



17 



1 J- 




F M A M J 



JASONDJFMAM 
Month 



0.8 - 



1^0.6 h 

3 



fo.4 




I I I I I I I \ \ 1 \ \ \ \ \ \ r 

b) JFMAMJJASONDJFMAM 

Month 
1 




E 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


E3 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


E2 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~l Indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



J FMAMJ JASONDJ FMAM 

Month 

Figure Al . Summary of monthly observations for Barksdale AFB, Louisiana from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



18 



0.8 



o0.6 
fo.4 



0.2 




a) 







^- 



J FMAMJ J ASOND J FMAM 

Month 



1 



0.8 



J'o.e 

(D 

|0.4 



0.2 




■ . . 

b) JFMAMJJASONDJFMAM 

Month 



1 



0.8 T 



o0.6 
|0.4 



0.2 




c) 




1- 

J FMAMJ JASOND J FMAM 

Month 



□ 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


D 


No Contrails, w/CI 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


n 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



n Indeterminate Contrails 

I I No Contrails 

r~| Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


n 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/CI 


■ 


Contrails, Persistent 



Figure A2. Summary of monthly observations for Beale AFB, California from January 93 to May 94: (a) relative frequency 
of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus 
with indeterminate data removed. 



19 



1 


- 


^— 




\ 














0.8 






S^O.6 

ST 


- 










. 


/ 




^ ^, 




£0.4 


— 








- 












0.2 


- 






















^/ 


// 






== 






77 




^^ 


2^ 


2^ 







' — 1 1 1 


















1 — i — 1 — 1 — i — 1 



a) JFMAMJJASONDJFMAM 

Month 



0.8 



JO.6 h 
|0.4h 



0.2 



b) 




-1 — I \ — r 



— I i n 

JFMAMJJASONDJFMAM 

Month 



0.8 



o0.6 

0) 

,?0.4 



0.2 




c) 



S 






2 



T r 



JFMAMJ JASONDJFMAM 

Month 



□ 


Indeterminate Contrails, w/Ci 


El 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


n 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



[~| Indeterminate Contrails 

[~| No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 



Q No Contrails, w/Ci 

I I No Contrails 

H Contrails, Not Persistent, w/Ci 

H Contrails, Not Persistent 

p^ Contrails, Persistent, w/CI 

H Contrails, Persistent 



Figure A3. Summary of monthly observations for Cairns AAF, Alabama from January 93 to May 94: (a) relative frequency 
of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus 
with indeterminate data removed. 



20 



1 

0.8 



"0.6 

CD 

3 



g" 



a) 



0.4 
0.2 h 




%A 




i I 

JFMAMJJASONDJFMAM 



Month 



1 



0.8 



o0.6 
o 

,20.4 



0.2 



b) 





1 
0.8 



o0.6 

3 






J FMAMJ JASONDJ FMAM 

Month 



S" 



c) 



0.4 

0.2 





/ / / / / 




/ / 



/ / / -.' 



JFMAMJ JASONDJFMAM 

Month 



□ 


Indeterminate Contrails, w/Ci 


H 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



[~~1 Indeterminate Contrails 

I I No Contrails 

[~1 Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


n 


No Contrails 


El 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure A4. Summary of monthly observations for Edwards AFB, California from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



21 



1 

0.8 



|0.6 

3 



g" 



a) 



0.4 - 
0.2 - 




L^ 



-\ 1 r 



■zz 



z 



J FMAMJ JASONDJ FMAM 

Month 



1 



0.8 



o0.6 

03 

|0.4 



0.2 



. ^ 

b) JFMAMJJASONDJFMAM 

Month 

1 



0.8 - 



1^0.6 
fo.4 




H 


Indeterminate Contrails, w/CI 


n 


Indeterminate Contrails 


u 


No Contrails, w/Ci 


u 


No Contrails 


s 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~| Indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

■ Contrails, Persistent 



□ 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


a 


Contrails, Persistent, w/Ci 


B 


Contrails, Persistent 



Month 
Figure A5. Summary of monthly observations for Eglin AFB, Florida from January 93 to May 94: (a) relative frequency of 
contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus with 
indeterminate data removed. 



22 



1 

0.8 



o0.6 
fo.4 



0.2 




a) 























-^ T"T" 






- 


."^ 






-, < 


> 


• 




■. . 






^ ■ , 




- 


■ >; 


\^ 










-'■ 








" _ '^ 










' _ "^ 










>. 


. N. 




w '^ 




- 


















. V 








- 




















— 






: 
















C>0 










- 
















y ■ 










— 












^ 




■,A,-> 


ISI 








- 


i?^ 






■7~? 


a 


^«r 


^^?5 




i i i 


MA 


B 


H 


■ 


■ 


■■"" 


M 


= 


f^ 




■ 





J FMAMJ JASONDJ FMAM 

Month 



1 



0.8 



"0.6 
fo.4 



0.2 - 








b) JFMAMJJASONDJFMAM 

Month 



0.8 



o0.6 
|0.4 



0.2 




c) 




— i — \ — I— 

JFMAMJ JASONDJFMAM 

Month 



s 


Indeterminate Contrails, w/Ci 


m 


Indetenninate Contrails 


n 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~| Indeterminate Contrails 

I I No Contrails 

r~1 Contrails, Not Persistent 

HI Contrails, Persistent 



[3 No Contrails, w/Ci 

I I No Contrails 

[\] Contrails, Not Persistent, w/Ci 

H Contrails, Not Persistent 

F^ Contrails, Persistent, w/Ci 

H Contrails, Persistent 



Figure A6. Summary of monthly observations for Fairchild AFB, Washington from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



23 



0.8 



fO.6 



u. 



a) 



0.2 


1 




J FMAMJ JASONDJ FMAM 

Month 



0.8 - 



o0.6 
|0.4 




□ 


Indeterminate Contrails, w/Ci 


E3 


Indeterminate Contrails 


E3 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



[~~| Indeterminate Contrails 

[~| No Contrails 

r~l Contrails, Not Persistent 

H Contrails, Persistent 



E3 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure A7. Summary of monthly observations for Griffis AFB, New York from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



24 



1 

0.8 



"0.6 
a> 

o 



a) 



0.4 - 

0.2 - 






JFMAMJJASONDJ FMAM 

Month 



1 



0.8 



^0.6 

Q 

|0.4 



0.2 




I I I I i I i I I Ill 

t>) JFMAMJJASONDJFMAM 

Month 

1 



0.8 



o0.6 
f 0.4 h 



c) 



0.2 





JFMAMJ JASONDJFMAM 

Month 



Q 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


D 


No Contrails 


El 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~| Indeterminate Contrails 

I I No Contrails 

r~l Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


n 


No Contrails 


El 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure A8. Summary of monthly observations for Hill AFB, Utah from January 93 to May 94: (a) relative frequency of 
contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus with 
indeterminate data removed. 



25 



0.8 



1^0.6 
|0.4 



0.2 




a) 



^ 



^ 



2=2 



ZZZ2 



2 



77 



2:2 



za 



2 



JFMAMJ JASONDJFMAM 

Month 



1 



0.8 



0O.6 
I0.4 



0.2 





b) JFMAMJJASONDJFMAM 

Month 
1 



0.8 - 



00.6 

CD 

3 

20.4 



c) 



0.2 




z 



33 



^^ 









JFMAMJ JASONDJFMAM 

Month 



□ 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


□ 


No Contrails, w/CI 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


m 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



I I Indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 






No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure A9. Summary of monthly observations for Kelly AFB , Texas from January 93 to May 94: (a) relative frequency of 
contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus with 
indeterminate data removed. 



26 



1 

0.8 



o0.6 

3 



S" 



0.4 



0.2 - 



a) 







4^ 



'5- 



•y- 







)>:A 



X X 
XX, 



:<^ 






LB 



_U 



FMAMJ JASONDJ FMAM 
Month 




J FMAMJ JASONDJ FMAM 

Month 



1 
0.8 



1^0.6 h 

a> 

|0.4 



c) 



0.2 




-■■/ 










JFMAMJ JASONDJFMAM 

Month 



Q 


Indeterminate Contrails, w/Ci 





Indeterminate Contrails 


□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


m 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~] indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure AlO. Summary of monthly observations for Langley AFB, Virginia from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



27 



0.8 



o0.6 
|0.4 



0.2 




a) 



: ; 


*, 




,. ',. ' 




- 




. 




^ 


.;. .-. 


> 


^ J 




:> 


' 


, " 


-. : 




. s 










. 




\ - 


■ ■ ■ K 

> . ■ y 

".' ■ ^ 

> < ^ .. 






"i 












- 
















— 
















'.^ 


. ■ / 


.^y ■ 




























1 


wm 




L^ 


^ 


i. 






i ~ 


T^^ 1 





J FMAMJ JASONDJ FMAM 

Month 



1 



0.8 



o0.6 
2 0.4 



0.2 



b) 







I I — I — I — 

J FMAMJ JASONDJ FMAM 

Month 



0.8 



o0.6 
^0.4 




E 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~| Indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure All. Summary of monthly observations for Loring AFB, Maine from January 93 to May 94: (a) relative frequency 
of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus 
with indeterminate data removed. 



28 



1 

0.8 



"0.6 

CD 

|o.4 



0.2 




a) 




J FMAMJ JASONDJ FMAM 

Month 



1 



0.8 



o0.6 
fo.4 



0.2 



b) 



"1 r 



J FMAMJ JASONDJ FMAM 

Month 



1 
0.8 



o0.6 

<D 

1 0.4 



0.2 




c) 



- 




'/'/ 


--■ / 


,,■ / 


/ / 


/ / 


'// 


/ ■ ,/ 




- 


,-■" / 


// 


,-■' / 


/ / 


/ / 




■'■ / 


/ / 




- 


/ / 


// 


--'/' 


/ / 


/ / 










— 


/ / 


■' /' 


/-''/ 




/ / 


■' / 


, ■'' . '■' 






" 


/ / 




/ / 




// 


/ / 












// 


/ / 




/ / 


/ / 








- 




■' / 


// 




/ / 


/ /' 








_ 










y / 


/ / 








- 




■" / 






/ / 


/ / 








- 










// 


/ / 








■ 










/ / 

■■■ / 










_ 










/ / 


/' _/' 








- 










/ / 










— ; — 


g 










^ 




\\ 


1 1 1 1 1 1 


1 




^ 


^ 



JFMAMJ JASONDJFMAM 

Month 



□ 


Indeterminate Contrails, w/Ci 


H 


indeterminate Contrails 


D 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


n 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



|~1 Indeterminate Contrails 

I I No Contrails 

r~] Contrails, Not Persistent 

H Contrails, Persistent 



[3 No Contrails, w/Ci 

I I No Contrails 

H Contrails, Not Persistent, w/Ci 

H Contrails, Not Persistent 

Contrails, Persistent, w/Ci 

H Contrails, Persistent 



Figure AI2. Summary of monthly observations for Luke AFB, Arizona from January 93 to May 94: (a) relative frequency 
of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus 
with indeterminate data removed. 



29 




J FMAMJ JASONDJ FMAM 

Month 



0.8 - 



0O.6 



2 0.4 



0.2 




b) JFMAMJJASONDJFMAM 

Month 




Q 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


Q 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


n 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~| Indeterminate Contrails 

r~| No Contrails 

n Contrails, Not Persistent 

I Contrails, Persistent 



El 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



JFMAMJ JASONDJFMAM 

Month 

Figure A13. Summary of monthly observations for McClellan AFB , California from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



30 



0.8 



o0.6 

3 



8" 



a) 



0.4 
0.2 


1 
0.8 



o0.6 

0) 

10.4 



S 



77 



J FMAMJ JASONDJ FMAM 

Month 




0.2 



I \ 1 i f i i \ \ \ 1 1 \ \ ; i r 

b) JFMAMJJASONDJFMAM 

Month 
1 

0.8 



o0.6 
|0.4h 



0.2 h 




c) 






-'■'A/A 



//■ 



/ /A 

^// 
// 
'\/ 




^-'z- 






/ / . 



JFMAMJ JASONDJFMAM 

Month 



u 


Indeterminate Contrails, w/Ci 


u 


Indeterminate Contrails 


u 


No Contrails, w/Ci 


u 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


n 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



n Indeterminate Contrails 

[~| No Contrails 

n Contrails, Not Persistent 

I Contrails, Persistent 



[3 No Contrails, w/Ci 

r~l No Contrails 

Fx] Contrails, Not Persistent, w/Ci 

H Contrails, Not Persistent 

Contrails, Persistent, w/Ci 

H Contrails, Persistent 



Figure A14. Summary of monthly observations for Minot AFB, North Dakota from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



31 



0.8 



"0.6 
fo.4 



0.2 




a) 




J FMAMJ JASONDJ FMAM 

Month 



1 



0.8 



^0.6 
20.4 



LL 




0.8 - 



JO.6 
|0.4 



0.2 




c) 





JFMAMJ JASONDJFMAM 

Month 



B 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



n Indeterminate Contrails 

I I No Contrails 

I I Contrails, Not Persistent 

H Contrails, Persistent 



Q No Contrails, w/Ci 

I I No Contrails 

[\] Contrails, Not Persistent, w/Ci 

Q Contrails, Not Persistent 

p^ Contrails, Persistent, w/Ci 

■ Contrails, Persistent 



Figure A15. Summary of monthly observations for Mountain Home AFB, Idaho from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



32 




0.8 - 



"0.6 
fo.4 




J FMAMJ JASONDJ FMAM 

Month 



1 
0.8 



"0.6 

3 



S" 



lOA p 




B 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


n 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


H 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



r~l Indeterminate Contrails 

r~| No Contrails 

I I Contrails, Not Persistent 

H Contrails, Persistent 



[3] No Contrails, w/Ci 

I I No Contrails 

[\] Contrails, Not Persistent, w/Ci 

H Contrails, Not Persistent 

Contrails, Persistent, w/Ci 

H Contrails, Persistent 



0.2 

, 

CJ JFMAMJJASONDJFMAM 

Month 

Figure A16. Summary of monthly observations for Offut AFB, Nebraska from January 93 to May 94: (a) relative frequency 
of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails and cirrus 
with indeterminate data removed. 



33 



i 
0.8 


- 






■'-.■■ 


'.A 


, ■ / 


K • 






:•> 








.' . 




S^O.6 

o 

3 


— 




^ 




' 


' 








' 


— 










£0.4 










0.2 


1 1 


^ 


zs 




7=? 








gi^i 


^ 


1 







2 


iZ 


"^Z 



a) 



J FMAMJ JASONDJ FMAM 

Month 



0.8 



|0.4 



0.2 




r— r-^ 

b) JFMAMJJASONDJFMAM 

Month 



1 



0.8 



o0.6 
fo.4 



0.2 




c) 




JFMAMJ JASONDJFMAM 

Month 



B 


Indeterminate Contrails, w/Ci 


m 


Indeterminate Contrails 


Q 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



I I Indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

■ Contrails, Persistent 



o 


No Contrails, w/Ci 


D 


No Contrails 


El 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure A17. Summary of monthly observations for Tinker AFB, Oklahoma from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



34 



0.8 



o0.6 
Jo.4 



0.2 




a) 



IT" 


. ^ 










/ 




^ 


- > 


Y: 

^ 

^ 


\ 

* \ 

■ > 








77, 




s 


k 


^ 




1 ^ 

V 


1 


',- 



□ 


Indeterminate Contrails, w/Ci 


E3 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 





Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



JFMAIVIJ JASONDJFMAM 

Month 



1 



0.8 



o0.6 
fo.4 



0.2 - 




[~| Indeterminate Contrails 

I I No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 



D 


No Contrails, w/Ci 


D 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



JFMAMJ JASONDJFMAM 

Month 

Figure Al 8. Summary of monthly observations for Whiteman AFB, Missouri from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



35 




J FMAMJ JASONDJ FMAM 

Month 



0.8 



I" 0.6 
|o.4 



0.2 - 




b) JFMAMJJASONDJFMAM 

Month 

1 



0.8 



o0.6 
|0.4|- 



0.2 I- 




c) 



m 




JFMAMJ JASONDJFMAM 

Month 



□ 


Indeterminate Contrails, w/Ci 


E3 


Indeterminate Contrails 


□ 


No Contrails, w/Ci 


n 


No Contrails 


H 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


E2 


Contrails, Persistent, w/Ci 


m 


Contrails, Persistent 



|~1 Indeterminate Contrails 

r~| No Contrails 

n Contrails, Not Persistent 

H Contrails, Persistent 



□ 


No Contrails, w/Ci 


n 


No Contrails 


Q 


Contrails, Not Persistent, w/Ci 


B 


Contrails, Not Persistent 


Q 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure A19. Summary of monthly observations for Wright Patterson AFB, Ohio from January 93 to May 94: (a) relative 
frequency of contrails and cirrus, (b) relative frequency and persistence of contrails, and (c) relative frequency of contrails 
and cirrus with indeterminate data removed. 



36 



Appendix B 
Hourly Mean Contrail Frequencies for Each Site 



37 



450 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



0.8 - 



^0.6 

3 

£0.4 



b) 



0.2 





r— 1 — r 
7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 - 



o 

Z3 
CT 

E 0. 



0.2 



C) 







1 — I — r 




1 — i — r 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





■ •.. \ 


Indeterminate Contrails, w/Ci 




X.>: 


Indeterminate Contrails 


''' / 


No Contrails, w/CI 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




^ 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B 1 . Summary of hourly contrail and cirrus observations from Barksdale AFB, Louisiana, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



38 




n — i — I — i — r 
a) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

1 



0.8 



|o.6 
o 

2 0.4 



b) 



0.2 I- 





1 I i r 
7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 

0.8 



gO.6 
2 0.4 

U- 



0.2 




c) 



-1 — i — I — i — r 






/ 



^ll^lillii 



'Y / ^ / 

/ / 



^Z.7X 



T 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





■, '\^ 


Indeterminate Contrails, w/Ci 




il 


Indeterminate Contrails 




No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




^ 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B2. Summary of hourly contrail and cirrus observations from Beale AFB, California, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



39 



250 




I i 1 r 
7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 

0.8 



8 0.6 

a- 
2 0.4 

U- 



0.2 




b) 



1 — I — r 



zzz 



Z2 



ZZiid 



^Z 



fesS^ 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 - 



8 0.6 

Q 

2 0.4 

u. 



0.2 




c) 



T—i — r 



Z 



y 



21 



'^'azs:^ 



Zc'j.Z 



^ 



3^ 



z 



a 



^ZL 



\ I I r 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



Indeterminate Contrails, w/Ci 



Indeterminate Contrails 



Y~\ No Contrails, w/Ci 

I I No Contrails 

Contrails, Not Persistent, w/Ci 
Contrails, Not Persistent 



A Contrails, Persistent, w/Ci 
^1 Contrails, Persistent 



Figure B3. Summary of hourly contrail and cirrus observations from Cairns AAF, Alabama, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



40 



g 
'•S5 

Si 
O 

o 

O 

X 



a) 



350 
300 
250 
200 
150 
100 
50 


1 



1 — I — i — I — r 



Z 



y, 



;^^^^9^PP 






^^'Aa 



</ 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



0.8 - 



^0.6 
2 0.4 



0.2 - 



1 — I — r 




b) 7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 



' 


- 




■'' / 




'/ 




''' / 


'' / 


/ 


'' / 


/' y 


/' , 


/ .-■ 


-■'/ 










^ 


'/ 


./' ^ 


/ y 






'' / 




'/ 


■' / 






/ 


/ 


/ 


' / 
















'' / 






■/ 


'' 


' / 


' y' 


'' / 


' / 


' / 






0.8 


— 




■ / 


'/[ 


/ / 


' / 
/ 


■' / 


;,/ 


•■'' / 




/ 


/ 






■ / 


/ 






- 




■ / 




■' / 




/ 


/ 






' / 


' ,./ 


^ / 




■ / 






>- 


- 


"' / 


— 




' / 


" / 










-^ y 




/ 


/' 








gO.6 














3 




































C7 




































£0.4 

u_ 


— 


































0.2 


- 












^ 


5 


^ 


? 


7, 


^ 


I 


?■ 


■■ 








¥ 


r. 


i 





I r -r — i — i — 


. 


E 


E 


t 


i 


^ 


1 


2 


1 


^ 


i 


1 


t 


■■-. 


1 1 





-N 


Indeterminate Contrails, w/Ci 




•« 


Indeterminate Contrails 


''X 


No Contrails, w/Ci 




No Contrails 


\5 


Contrails, Not Persistent, w/Ci 




Contrails, Not Persistent 


% 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure B4. Summary of hourly contrail and cirrus observations from Edwards AFB, California, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence , and (c) relative frequency of occurrence with indeterminate 
data removed. 



41 



250 



I 200 (- 
I 150 h 

o 

"5 100 - 

CO 

I 50 h 







fl 



T I I r 



a) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



0.8 - 



gO.6 
o 

£0.4 



b) 



0.2 




^^irii 



z 



H2W 



1 — i — I — r 



7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 



1 



0.8 



gO.6 
2 0.4 

U- 



c) 



0.2 h 




T — r 



^222 



1 — i — I — r 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





- \^ 


Indeterminate Contrails, w/Ci 






Indeterminate Contrails 


''/,. 


No Contrails, w/Ci 


u 


No Contrails 


^ 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




y 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B5. Summary of hourly contrail and cirrus observations from Eglin AFB, Florida, centered at local noon: (a) number 
of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate data 
removed. 



42 




a) 7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 
1 



0.8 



|o.6 

o 

XT 

2 0.4 



0.2 




b) 



I I i r 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 




'' / ■'' / 








■■ "''- 


Indeterminate Contrails, w/Ci 




V ■'"'■■■ 


Indeterminate Contrails 




No Contrails, w/Ci 




No Contrails 


\3 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 






Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure B6. Summary of hourly contrail and cirrus observations from Fairchild AFB, Washington, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



43 




a) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 
1 



0.8 



gO.6 

C7 

£0.4 
u. 



0.2 




b) 



^Z 



SSi 



Z^^Z'>^ 



2;a 



^^ 



z 



z 



T^ 



— I — i — 1 I I I I i i I i i i I i I i i I 1 1 I 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 



gO.6 

0) 

a- 
2 0.4 



0.2 




c) 



-I — r 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





■■■ \ 


Indeterminate Contrails, w/Ci 




C'-y 


Indeterminate Contrails 


, ' ,. 


No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




^ 


Contrails, Persistent, w/CI 


■ 


Contrails, Persistent 



Figure B7. Summary of hourly contrail and cirrus observations from Griffis AFB, New York, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



44 



400 

350 

c 

iaoo 

CO 
§250 

g200 

° 150 

CO 

iioo 

^ 50 



a) 







I I i r 



22iii^^i^^2Z2z 



T — r 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 




I I i r 
7 8 9 101112131415161718192021222324 1 2 3 4 5 6 



UTC 



0.8 



gO.6 

o 

2 0.4 



0.2 




c) 



T — I — r 



= Z2^ 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 







Indeterminate Contrails, w/Ci 




X J<. 


Indeterminate Contrails 




No Contrails, w/Ci 




No Contrails 


N^ 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




^ 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B8. Summary of hourly contrail and cirras observations from Hill AFB, Utah, centered at local noon: (a) number of 
observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate data 
removed. 



45 



400 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



0.8 



^0.6 

(D 

2 0.4 



0.2 



b) 







z£3Z 



^^ 



-1 — r 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



0.8 



^0.6 

cr 
2 0.4 



c) 



0.2 F- 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





■ ^ X 


Indeterminate Contrails, w/Ci 






Indeterminate Contrails 


-••' / 


No Contrails, w/Ci 




No Contrails 


s 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




^ 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B9. Summary of hourly contrail and cirrus observations from Kelly AFB, Texas, centered at local noon: (a) number 
of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate data 
removed. 



46 



300 




a) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

1 




b) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 
1 



0.8 



^0.6 

=3 

2 0.4 



0.2 



A A / 




/ '/ '■ A A 



"T— i — i — r 



Indeterminate Contrails, w/Ci 






Indeterminate Contrails 

No Contrails, w/Ci 

No Contrails 

Contrails, Not Persistent, w/Ci 

Contrails, Not Persistent 



'/y Contrails, Persistent, w/Ci 



Contrails, Persistent 



C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure BIO. Summary of hourly contrail and cirrus observations from Langley AFB, Virginia, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



47 




a) 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 

0.8 

1 0.6 

o 

cr 
£0.4 

u_ 

0.2 




b) 




T 1 1 i — 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 - 



8 0.6 

cr 
£0.4 

U- 



0.2 








I I I r 



Indeterminate Contrails, w/Ci 






Indeterminate Contrails 
[/^ No Contrails. w/Ci 
I I No Contrails 

Contrails, Not Persistent, w/Ci 



^B Contrails, Not Persistent 
y^ Contrails, Persistent, w/Ci 



Contrails, Persistent 



C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure Bl 1 . Summary of hourly contrail and cirrus observations from Loring AFB, Maine, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



48 



250 




a) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 
1 



0.8 



^0.6 

o 

2? 0.4 



0.2 




/ , 





b) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 



^0.6 


cr 
2 0.4 



0.2 




^ 












., ''X^ 


Indeterminate Contrails, w/Ci 




"X 


Indeterminate Contrails 


■'/'] 


No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




'A 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure B 1 2. Summary of hourly contrail and cirrus observations from Luke AFB , Arizona, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



49 



300 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

1 I 



0.8 



^0.6 

CT 

£0.4 



0.2 




b) 




— I — I — r — I — r 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 



^0.6 

0) 
CT 

£ 0.4 



0.2 




Mi 



'//y//<' 



c) 



— I — I — I — I — r 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





"^ 


Indeterminate Contrails, w/CI 




"'' "^x 


Indeterminate Contrails 


,-■" / 


No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




\A 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B13. Summary of hourly contrail and cirrus observations from McClellan AFB, California, centered at local noon: 
(a) number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with 
indeterminate data removed. 



50 



400 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



0.8 



1 0.6 

a- 
2 0.4 



0.2 



b) 



n — i — r 



7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 



0.8 



|o.6 

(U 

2 0.4 



0.2 




c) 



tlZ^ 



:z 



ZEl 



I i r 
7 8 9 101112131415161718192021222324 1 2 3 4 5 6 



UTC 







Indeterminate Contrails, w/CI 






Indeterminate Contrails 




No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 






Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B 14. Summary of hourly contrail and cirrus observations from Minot AFB , North Dakota, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



51 



c 
o 

'IS 

(D 

<n 
O 
o 

CO 

3 

o 

X 



a) 



400 
350 r 
300 
250 
200 
150 h 
100 
50 





— I — I — i — I — r 

7 8 9 101112131415161718192021222324 12 3 4 5 6 



UTC 



0.8 - 



gO.6 

0} 

=} 
a- 

£0.4 



0.2 





T — i — r 
b) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

1 



0.8 



gO.6 
£0.4 



0.2 




Kge«'*HK™»H 



till 





' ■ ..^ N 


Indeterminate Contrails, w/Ci 






Indeterminate Contrails 


/ / 


No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




E 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



-r — i — r~r 
C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure B 15. Summary of hourly contrail and cirrus observations from Mountain Home AFB, Idaho, centered at local noon: 
(a) number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with 
indeterminate data removed. 



52 




b) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

1 






, \ 


Indeterminate Contrails, w/Ci 






Indeterminate Contrails 


,•'/ 


No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


N 


Contrails, Not Persistent 




// 


Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

Figure B 16. Summary of hourly contrail and cirrus observations from Offut AFB, Nebraska, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



53 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 
0.8 

1 0.6 

o 

3 
D" 

£0.4 
0.2 



b) 







-1 — 1 — r 



-1 — I — r 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 
1 I 



0.8 - 



gO.6 

0) 

£0.4 h 

U- 



0.2 



C) 








T I I I ) I I I I i I I I I I i I t < I I I 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





■-.. \ 


Indeterminate Contrails, w/Ci 




.■■■■f>< 


Indeterminate Contrails 


/< 


No Contrails, w/Ci 




No Contrails 




Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 




^ 


Contrails, Persistent, w/CI 


■ 


Contrails, Persistent 



Figure B 17. Summary of hourly contrail and cirrus observations from Tinker AFB, Oklahoma, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



54 





500 




450 


c 
o 


400 


CO 

o 
tn 

o 


350 
300 
250 


o 


200 


o 

I 


150 
100 




a) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 

1 



0.8 



^0.6 

C3- 

2 0.4 



X 
>: 



K^>x 



.$<>>< 






x^: 



X 



'xX- 

X nX- 







X 



\X 



:X 



X 



// 



XI 

'' X --■ 



0.2 

1 I i i — i — i — i — i — i — i i — i — I — i — I — i — I — I — I — I — 1 — \ — r 

b) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 
1 



0.8 



^0.6 
2 0.4 



0.2 h 


C) 7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



- 




-''/ 






■■■'' / 


•-''/ 






/ ^ 










■ / 


' / 


/ 




- 




'' 




,' 


' A 


/' 


/ ^, 


■■'' ^ 


^ ' 




,/■■ / 






/ ^ 








- 






' / 


/\, 






/ 


- X 


/ 

' / 


'■■' / 


'' / 


"" / 


'V 


'\/ 


/ 








/ 


-' '' 


^ / 


■' -- 


■■ / 




■' / 




/ ^ 


' ,. 


.■ 


-' / 


.-'■ , 


' _, 


■' ■' 






, 


' / 




' / 






^' / 


■'' / 


''X 


/ y 


,x 


'' / 


'■' / 


./- . 


'''/ 


■ / 






- 


■ / 


'■ / 


■" / 


■■ / 


.-' / 


' / 


' / 


' X 


"/ 


■'' ./ 


■' / 


■■'/ 




/ 


' / 


/... 




- 


X 

/ 

''' / 


V 




/ 




-^ 


■' / 


/^ 


/ 




' / 


■■;/ 


■'''/ 




'x 
'/ 


/ > 
/ 

/ 








- 


■ / 






























-' / 




- 




' •' 
































— 


'' / 


,- . 
































- 
































/ 




































■-■ 




^ 




































- 


^ 




2 






2 


s 


S 










^ 


s 


s^ 






\^ 


s 




^ 


- 


/. 


Y 


z 




^ 


2 


2 


Z. 




£ 




z 


?. 


2 


f 


y. 




1 1 1 




















^n 




\ 










1 1 1 







Indeterminate Contrails, w/Ci 




a 


Indeterminate Contrails 


/ / 


No Contrails, w/Ci 




No Contrails 


\k 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 






Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B 1 8. Summary of hourly contrail and cirrus observations from Whiteman AFB, Missouri, centered at local noon: (a) 
number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with indeterminate 
data removed. 



55 



350 
300 l 



c 
o 

■« 250 
S2OO h 
§150^ 

iioo^ 

o 

X 



a) 



50 


1 
0.8 



gO.6 

Q 

£0.4 



0.2 





1 — I I I { 11 I T I I I 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 



UTC 



b) 



T — i — r 



X>' 




7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 



1 



0.8 - 



^0.6 
2 0.4 

LL 



c) 



0.2 





1 — I — r 
7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 







Indeterminate Contrails, w/Ci 




x;": 


Indeterminate Contrails 


'\/[, 


No Contrails, w/CI 




No Contrails 


f^^ 


Contrails, Not Persistent, w/Ci 


■ 


Contrails, Not Persistent 






Contrails, Persistent, w/Ci 


■ 


Contrails, Persistent 



Figure B19. Summary of hourly contrail and cirrus observations from Wright Patterson AFB, Ohio, centered at local noon: 
(a) number of observations, (b) relative frequency of occurrence, and (c) relative frequency of occurrence with 
indeterminate data removed. 



56 



Appendix C 
Cloud Amount Statistics for Each Site 



57 




J FMAMJ JASONDJ FMAM 

Month 




7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 




□ 


90% to 100% 


m 


70% to 80% 


m 


50% to 60% 


□ 


30% to 40% 


Q 


10% to 20% 


D 


0% 



n — I — I — r 

JFMAMJJASONDJFMAM 

Month 

Figure CI . Summary of cloud cover for Barksdale AFB, Louisiana from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



58 




JFMAMJ JASONDJ FMAM 

Month 




b) 



7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 



1 \J\J 




90 


- 


^ 80 


_ 


t. 70 


- 


a5 60 


_ j^~~9 A fk 


> 


.^^/""^ / \ / \ 


o 50 


t ^y V • 


-o 40 


- \ Y^ 


o 30 


\ / 


O 20 


- \ / 


10 


V'-^^i^ 


c) 




1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 



D 


90% to 100% 


B 


70% to 80% 


E3 


50% to 60% 


EI 


30% to 40% 





10% to 20% 


D 


0% 



J FMAMJ JASONDJ FMAM 

Month 

Figure C2. Summary of cloud cover forBeale AFB, California from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



59 



0.8 - 



1 0.6 

o 

.2 0.4 



0.2 

a) 

1 

0.8 



1 — r 



p^ 



^ 



^^^ 






^_ 



f^ 



^ 



^^ 






^ 



Kte::^ 



1 — r 



J FMAMJ JASONDJ FMAM 

Month 




7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 




□ 


90% to 100% 


^^ 


70% to 80% 





50% to 60% 


El 


30% to 40% 


Q 


10% to 20% 


D 


0% 



1 — \ — \ — i — I — \ — \ — \ — I — i — 1 — \ \ \ \ r 

J FMAMJ JASONDJ FMAM 

Month 

Figure C3. Summary of cloud cover for Cairns AAF, Alabama from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



60 




0.2 
a) 
1 
0.8 



JFMAMJJASO 

Month 



N D J F M A M 




b) 



"I — I — i — i — \ — I — I — I — I — I — 1 — I — I — I — I — I — I — I — I — I — I — i — r 
7 8 9 101112131415161718192021222324 12 3 4 5 6 

UTC 




m 


90% to 100% 


m 


70% to 80% 


El 


50% to 60% 


□ 


30% to 40% 


Q 


10% to 20% 


D 


0% 



\ i I i i i i i I i i r 

J FMAMJ JASONDJ FMAM 

Month 

Figure C4. Summary of cloud cover for Edwards AFB, California from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



61 




7 8 9 1011 12131415161718192021222324 12 3 4 5 6 

UTC 




□ 


90% to 100% 


n 


70% to 80% 


Q 


50% to 60% 


□ 


30% to 40% 





10% to 20% 


n 


0% 



1 — i — i — I — \ — i I \ i I I I i r 

J FMAMJ JASONDJ FMAM 

Month 

Figure C5 . Summary of cloud cover for Eglin AFB , Florida from January 93 to May 94; (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



62 



0.8 



I" 0.6 

Q 

13 
C7 

2 0.4 



0.2 



a) 






/v 



f/;% 



z 



— I — I — l— 
J F M A M J J 






i^ 






v/. 



'J. 




A 



% 



^ 



/. 



■'::^ 



nart 






^?i 



^7 



/y 



I 



i 



A S O N D 

Month 



J F M A M 




b) 



7 8 9 1011 12131415161718192021222324 12 3 4 5 6 

UTC 





100 




90 


^,_^ 


80 


^ 


70 


CD 


60 


> 




o 


50 


T3 


40 


3 
O 


30 


O 


20 




10 


c) 







n 


90% to 100% 


B 


70% to 80% 


n 


50% to 60% 


□ 


30% to 40% 


Q 


10% to 20% 


D 


0% 



1 I I 1 I I 1 i I i I I I i I r 

J FMAMJ JASONDJ FMAM 

Month 

Figure C6. Summary of cloud cover for Fairchild APB, Washington from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



63 



No Data Available 



Figure C7. Summary of cloud cover for Griffis AFB, New York from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



64 



No Data Available 



Figure C8. Summary of cloud cover for Hill AFB, Utah from January 93 to May 94: (a) relative frequency of cloud cover, 
(b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



65 



No Data Available 



Figure C9. Summary of cloud cover for Kelly AFB, Texas from January 93 to May 94: (a) relative frequency of cloud cover, 
(b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



66 




0.2 
a) 
1 
0.8 



J FMAMJ JASONDJ FMAM 

Month 




b) 



7 8 9 ion 12131415161718192021222324 12 3 4 5 6 

UTC 



1 \J\J 

90 
^ 80 
^ 70 
1 60 


"^ ,,^ ' 


• 
1 1 1 


-D 40 

o 30 

O 20 

10 

c) 


•-• • • 
1 1 1 1 1 1 1 1 1 1 1 1 1 



□ 


90% to 100% 


■ 


70% to 80% 


E3 


50% to 60% 


E 


30% to 40% 


Q 


10% to 20% 


n 


0% 



JFMAMJJASONDJFMAM 

Month 

Figure CIO. Summary of cloud cover for Langley AFB, Virginia from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



67 




JFMAMJ JASONDJ FMAM 

Month 




7 8 9 1011 12131415161718192021222324 12 3 4 5 6 

UTC 




n 


90% to 100% 


^H 


70% to 80% 


o 


50% to 60% 


Q 


30% to 40% 


E2 


10% to 20% 


n 


0% 



1 — \ — r 

J FMAMJ JASONDJ FMAM 

Month 

Figure C 1 1 . Summary of cloud cover for Loring AFB , Maine from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



68 



0.8 



0.6 
a> 
cr 
2 0.4 



0.2 h 
a) 
1 
0.8 



-r 



~i — I — I — i i i — 

J FMAMJ JASONDJ FMAM 

Month 




b) 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 





100 




90 


^_, 


80 


£ 


70 


<D 


60 


> 




O 

o 


50 


T3 


40 


13 

o 


30 


O 


20 




10 


c) 







D 


90% to 100% 


m 


70% to 80% 


n 


50% to 60% 


u 


30% to 40% 


D 


10% to 20% 


n 


0% 



T I I i i i \ I I 1 — I — r 

J FMAMJ J ASONDJ FMAM 

Month 

Figure CI 2. Summary of cloud cover for Luke AFB, Arizona from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



69 



No Data Available 



Figure C13. Summary of cloud cover for McClellan AFB, California from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon , and (c) monthly mean cloud cover. 



70 



0.8 



I" 0.6 

3 

2 0.4 



0.2 

a) 

1 

0.8 






TZ 



I — r 



^^g. 



2k 




feg 



.x.^ 



J FMAMJ JASONDJ FMAM 

Month 




7 8 9 1011 12131415161718192021222324 12 3 4 5 6 

UTC 




1 1 1 I I i \ I I — I — r 

JFMAMJ JASONDJ FMAM 

Month 



D 


90% to 100% 


■ 


70% to 80% 


S 


50% to 60% 


S 


30% to 40% 


Q 


10% to 20% 


D 


0% 



Figure C14. Summary of cloud cover for Minot AFB, North Dakota from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



71 




0.2 

a) 

1 



J FMAMJ J ASOND J FMAM 

Month 




b) 



7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 




m 


90% to 100% 


m 


70% to 80% 


m 


50% to 60% 


□ 


30% to 40% 





10% to 20% 


D 


0% 



n — \ — \ — 1 — 1 — r 

JFMAMJJASONDJFMAM 

Month 

Figure C15. Summary of cloud cover for Mountain Home AFB, Idaho from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



72 




-I 1 1 i 1 1 i 1 [ 1 1 1 ! 1 ! 

J FMAMJ JASONDJ FMAM 

Month 




b) 



100 

90 

^ 80 

^ 70 

I 60 

o 50 

■D 40 

o 30 

O 20 

10 

c) 



-1 — I — r 

7 8 9 101112131415161718192021222324 1 2 3 4 5 6 

UTC 




I i i 1 i i i i i i i 1 \ 1 1 T 

J FMAMJ JASONDJ FMAM 

Month 



□ 


90% to 100% 


■ 


70% to 80% 


m 


50% to 60% 


□ 


30% to 40% 





10% to 20% 


n 


0% 



Figure CI 6. Summary of cloud cover for Offut AFB, Nebraska from January 93 to May 94: (a) relative frequency of cloud 
cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



73 



No Data Available 



Figure C 1 7 . Summary of cloud cover for Tinker AFB , Oklahoma from January 93 to May 94: (a) relative 
frequency of cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and 
(c) monthly mean cloud cover. 



74 



0.8 - 



"0.6 

o 

1 0.4 



0.2 



a) 



^ 






u«, 






z 



z 



^^ 



^ 



?; 



/. 



^ 






•^ 



J FMAMJ JASONDJ FMAM 

Month 




b) 



7 8 9 1011 12131415161718192021222324 12 3 4 5 6 

UTC 




"1 I I i I I r 

FMAMJ JASONDJ FMAM 

Month 



n 


90% to 100% 


B 


70% to 80% 


m 


50% to 60% 


E\ 


30% to 40% 


Q 


10% to 20% 


D 


0% 



Figure C18. Summary of cloud cover for Whiteman AFB, Missouri from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



75 



0.8 - 



"0.6 

IT 

.S^O.4 



0.2 
a) 



pmi 



>^ 



22^ 









/O' 






(A^ 



^ 



'/ 






/A 



JFMAMJJASONDJFMAM 

Month 




b) 



7 8 9 1011 12131415161718192021222324 12 3 4 5 6 

UTC 




m 


90% to 100% 


B 


70% to 80% 


E3 


50% to 60% 


□ 


30% to 40% 


Q 


10% to 20% 


D 


0% 



J FMAMJ JASONDJ FMAM 

Month 

Figure C19. Summary of cloud cover for Wright Patterson AFB, Ohio from January 93 to May 94: (a) relative frequency of 
cloud cover, (b) diurnal cycle of cloud cover relative frequency centered at local noon, and (c) monthly mean cloud cover. 



76 



References 

Angell, J. K. 1990: Variation in United States Cloudiness and 
Sunshine Duration Between 1 950 and the Drought Year of 
1988.7. Climate, 3, 296-308. 

Angell, J. K.; Korshover, K. J.; and Cotton, G. F. 1984: 
Variation in United States Cloudiness and Sunshine. J. 
Climate Appl. MeteoroL, 23, 752-761. 

Appleman, H., 1953: The Formation of Exhaust Condensation 
Trails by Jet Aircraft. Bull. Am. MeteoroL Sac, 34, pp. 
14-20. 

Bakan, S.; Betancour, M.; Gayler, V.; and Grassl, H. 1994: 
Contrail Frequency Over Europe From NOAA-Satellite 
Images. Annal. Ceophys., 12, 962-968. 

Baughcum, S. L. 1996: Subsonic Aircraft Emission 
Inventories. Atmospheric Effects of Aviation: First Report 
of the Subsonic Assessment Pro/'ecr, NASA RP- 1 385, pp. 
15-29. 

Baughcum, S. L.; Metwally, M.; Seals, R. K.; and Wuebbles, D. 
J., 1993: Emissions Scenarios Development: Completed 
Scenarios Database. Atmospheric Effects of Stratospheric 
Aircraft: A Third Program NASA RP-1313, pp. 185-208. 

Carleton, A. M.; and Lamb, P. J.: 1986: Jet Contrails and 
Cirrus Cloud: A Feasibility Study Employing High- 
Resolution Satellite Imagery. Bull. Am. MeteoroL Soc, 
67, 301-309. 

Changnon, S. A. 1981: Midwestern Sunshine and Temperature 
Trends Since 1901: Possible Evidence of Jet Contrail 
Effects. /. Appl. Meteorol.,20, pp. 496-508. 

DeGrand, J. Q.; Carleton, A. M.; and Lamb, P. J. 1990: A Mid- 
Season Climatology of Jet Condensation Trails From 
High-Resolution Satellite Data. Proc. of the AMS 7th 
Conf. on Atmos. Radiation, San Francisco, CA, July 23- 
27, 309-311. 

Engelstad, M.; Sengupta, S. K.; Lee, T.; and Welch, R. M. 
1992: Automated Detection of Jet Contrails Using the 
AVHRR Split Window. Intl. J. Remote Sensing, 13, 1391- 
1412. 

Hahn, C. J.; Warren, S. G.; London, J.; Chervin; and Jenne, R. 
L., 1984: Atlas of Simultaneous Occurrence of Different 
Cloud Types Over Land. NCAR Technical Note TN- 
241+STR, Boulder, Colorado, 212 pp. 

Hahn, C. J.; Warren, S. G.; London, J.; Chervin; and Jenne, R. 
L., 1986: Global Distribution of Total Cloud Cover and 
Cloud Type Amounts Over Land. NCAR Technical Note 77V- 
273+STR, Boulder, Colorado, 288 pp. 

Hahn, C. J.; Warren, S. G.; and London, J. 1995: The Effect of 
Moonlight on Observation of Cloud Cover at Night, and 
Application to Cloud Climatology. J. Climate, 8, pp. 
1429-1446. 



Hanson, H. M.; and Hanson, D. M. 1995: A Reexamination of 
the Formation of Exhaust Condensation Trails by Jet 
Aircraft. J. Appl. MeteoroL, 34, pp. 2400-2405. 

Joseph, J. H.; Levin, Z.; Mekler, Y.; Ohring, G.; and Otterman, 
J. 1975. Study of Contrails From the ERTS 1 Satellite 
Imagery. J. Geophys. /?«., 80, 366-372. 

Karcher, B., 1994: Transport of Exhaust Products in the Near 
Trail of a Jet Engine Under Atmospheric Conditions. J. 
Geophys. Res.,99, pp. 14,509-14,517. 

Kuhn, P. M. 1970: Airborne Observations of Contrail Effects 
on the Thermal Radiation Budget. J. Atmos. Sci., 27, 937- 
942. 

Lee,T.F. 1989: Jet Contrail Identification Using the AVHRR 
Infrared Split Window. J. Appl. MeteoroL, 2S, 993-995. 

Liepert, B.; Fabian, P.; and Grassl, H. 1994: Solar Radiation in 
Germany - Observed Trends and an Assessment of Their 
Causes Part I: Regional Approach. Beitr. Phys. Atmosph., 
67, 15-29. 

Murcray, W. B., 1970: On the Possibility of Weadier 
Modification by Aircraft Contrails. Mon. Wea. Rev., 98, 
pp. 745-748. 

Sadler, J. C, 1977: The Upper Tropospheric Circulation Over 
the Tropics, Part 11 - Statistics. University of Hawaii 
Technical Report UHMErr-77-02, Department of 
Meteorology, University of Hawaii, Honolulu, NSF Grant 
No. GA-36301, April, 24 pp. 

Seaver, W. L.; and Lee, J. E. 1987: A Statistical Examination 
of Sky Cover Changes in the Contiguous United States. /. 
Clim. Appl. MeteoroL, IS, pp. 88-95. 

Schumann, U. 1996: On Conditions for Contrail Formation 
From Aircraft Exhausts. MeteoroL Zeitschrift, 5, pp. 4- 
23. 

Schumann, U.; and Wendling, P. 1993: Determination of 
Contrails From Satellite Data and Observational Results. 
Air Traffic and the Environment, Proc. of DLR Intl. 
Colloquium, Bonn, Germany, Nov. 15-16, Springer- 
Verlag, NY, 138-153. 



Scorer, R. S., 1972: Clouds of the World. 
(Publishers) Ltd., 176 pp. 



David &. Charles 



Thompson, A. M., Friedl, R. R.; and H. L. Wesoky, 1996: 
Introduction and Overview. Atmospheric Effects of 
Aviation: First Report of the Subsonic Assessment Project, 
NASA RP-I385, pp. 1-14. 

Weber, G.-R. 1990: Spatial and Temporal Variation of 
Sunshine in the Federal Republic of Germany. Theor. Appl. 
ClimatoL, 41, 1-9. 



77 



^ 



c 
o 



> 

•a 



B 

s 

I 

o 

to 
O 

c 
u 
o 

51 



S 

e 
U 


o 

s 

00 


OS 

vb 


5^ 
5 


o 
m 




in 

o^ 

00 


P 
m 

(N 


00 

in 


en 

t 
fn 


en 

d 
en 

en 

evi 
es 


00 

in 

oo 

fs) 


en 
5 


00 

es 

en 

00 
00 

es 


so 

t 

in 

d 

en 


00 

en 
m 


en 

s 

00 

en 
es 


00 

en 
en 




en 












m 
cs 


si 


en 

CO 
ON 




en 
en 


in 

00 

en 


in 


3" 
t 


en 
m 


p 
en 
en 


00. 

od 
in 


in 

en 


00 


OS 
m 

00 


1 




00 


o, 

so 

o 


rn 


O 
en 
d 


en 


o 
S 


00 


"1 

•<* 

d 


p 
en 


d 

en 

d 


o 


es 

oq 




"? 


p 


P 

Ci 
"1 


o 
S 

00 

rn 


P-. 
en 

O 
es 


r-: 


>n 

d 


o 
o 


p 


d 




^ 
p 




o 
d 


00. 

en, 
00 


in 

s 

es 


r? 


■*. 

o 

en 

d 


00 

d 

p 


en 

OS 

d 




00 

e 

so 
P 


in 

ci- 


t-. 

en 

p 
es 




00 

p 


00 


in 


00 


00 


p- 


00 
00 


d 


in 

en 


P 
en 

OS 


OS 
P 




2; 

Csl 


P 

en 


in 

en 


OS 


Os 

en 


<» 
in 

en 


es 






o 


d 
d 


*5 




r- 

d. 

o 


B 

00 

d 


o 

P 


in 


s 

en 


d 


00 


en 

00 

d 


en 

in 
esi 


P 

SO 

d 


d 


00 

o 


'J 

en 


# Indet. 
withCi 


1 


s 


00 


OS 
CO 


m 


OS 


1 
in 

so 


Os 

s 

so 
m 


OS 
m 




s 

es 




en 


i 

en 
en 
2 


m 

00 


s 

S 


1 
m 

00 
00 




1 


!l 


in 


s 
^ 


00 


lO 


in 


00 

so 
tn 


OS 


00 
so 


m 
m 
so 
m 


1 

en 


OS 

5 


S 


m 


00 


oo 

fsl 


in 


5 


so 

m 

00 

m 


en 


! 


» 


1 

m 
en 


1 

s; 




in 


1 

5: 


rn 


in 
m 


en 
en 

CO 


en 
en 


en 

OS 

en 


en 


en 
en 

CO 


1 

en 

CO 


en 
en 
en 


in 

t 

en 

S 


en 


in 

en 
5 


en 

1 

en 


j 


z 

o 

in 


■!t: 
z 


r-. 

00 

z 

en 


z' 

OS 


5: 

>n 

Z 


? 

so 
5 


m 
p- 

Z 

9 


P 

Z 
5 


oo 

z 

OS 


z 

en 


bs 

z' 


ON 


R 

en 
en 


oo 
en 


o 
Z 

00 


a 

m 

Z 
en 


in 

Os 

Z 


z 

m 

en 


en 
OS 

Z 

e 
00 

en 


i 

z 

00 
OS 
en 


1 


B3 


o 

CO 


< 

E 

■a 
u 


< 

u 

1 


4 

c 


< 

i2 


S 
•c 
O 


5 

ir 




< 
> 

! 


1 


i 
3 


< 
U 

c 

1 

u 

u 


Q 

Z 

o 

c 


e 

E 
o 
X 

i 


<§ 


q 


O 

Mi 


X 

q 

i 

4 

1 



78 



Table 2. Diumal characteristics of total contrail occurrence without indeterminate data. 



Station 


Maximum 

Number of 

Samples 


Time of 

Primary 

Maximum 

UTC(LST) 


Time of 
Secondary 
Maximum 
UTC(LST) 

20 (12) 


Mean 
Occurrence 
Frequency 

of all 
contrails, 
M _ 

0.254 


(Max] - Min)llM (%) 
23 


{Max2 - Min)llM (%) 
19 


Barksdale, LA 


448 


23 (17) 


Beale.CA 


405 


1(17) 


17(9) 


0.115 


39 


38 


Cairns, AL 


235 


15(9) 


24 (18) 


0.076 


54 


44 


Edwards, CA 


348 
217 


23 (15) _ 
14(8) 


18 (10) 
18 (12) 


0.172 
0.063 


65 

72 


63 

48 


Eglin, FL 


Fairchild, WA 


320 


16(8) 


19(11) 


0.177 


51 


46 


Griffis, NY 


393 


J3(8)_ 
17 (10) 


21 (16) 


0.195 
0.165 


35 
71 


12 
67 


Hill, UT 


356 


23 (16) 


Kelly, TX 


369 


23 (17) 


17 (10) 


0.064 


72 


65 


Langley, VA 
Loring, ME 


276 
L_ 151 


14(9) 
24 (19) 
17 (10) 


_ J 8 (131 ._ 
20 (15) 


0.241 


38 


19 

85 


0.158 


98 


Luke.AZ 


236 


20 (13) 


0.134 


39 


25 


McClellan, CA 


274 
393 


1 (17) 
18(11) 


16(8) 
22 (15) 


0.175 
0.087 


50 
85 


46 
66 


Minot, ND 


Mt. Home, ID 


395 


23 (15) 


20 (13) 


0.107 


78 


51 


Offut,NE 
Tinker, OK 


454 

367 


18 (12) 


22 (16) 


0.241 


63 _ 
51 


49 
45 


17 (10) 


23 (16) 


0.129 


Whiteman. MO 


480 


16 (10) 


19 (13) 


0.143 


54 


33 


Wright-Pat., OH 


303 


15(9) 


21 (15) 


0.346 


61 


52 



79 



REPORT DOCUMENTATION PAGE 



FofTTi Approved 
OMB No^ 0704-0188 



Public reporting burden for this collectton of information Is estimated to average 1 hour per response, Including ttie time tor reviewing instructions, soafohing existing data sources, 
gathering and maintaining the dala needed, and completing and reviewing the collection of informatton. Send comments regarding this burden estimate or any other asject of ttiis 
collection of infomiation.Tnduding suggestions tor reducing this burden, to Washington Headquarters Senncos, Dirsctorate for Information Operations and Reports, 1215 Jefferson 
Davis Highway, Suite 1 204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Papenrork Reduclion Project (0704-0188), WasNngton, DC 20503. 



1. AGENCY USE ONLY (Leave blank) 



2. REPORT DATE 

December 1997 



3. REPORT TYPE AND DATES COVERED 

Reference Publication 



4. TITLE AND SUBTITLE 

Surface-Based Observations of Contrail Occurence Frequency Over the U.S., 
April 1993 -April 1994 



6. AUTHOR(S) 

Patrick Minnis, J. Kirk Ayers, and Steven P. Weaver 



5. FUNDING NUMBERS 



WU 538-08-12-01 



7. p£Af'6rMING 6rGANIZATI0N NAME(S) and AD0RESS(ES) 

NASA Langley Research Center 
Hampton, VA 23681-2199 



8. PERFORMING ORGANIZATION 
REPORT NUMBER 



L- 17633 



9. SPONSORING/MONrrORING AGENCY NAME(S) AND ADDRESS(ES) 

National Aeronautics and Space Administration 
Washington, DC 20546-0001 



10. SPONSORING/MONITORING 
AGENCY REPORT NUMBER 



NASA RP- 1404 



t1. SUPPLEMENTARY NOTES 

Minnis: Langley Research Center, Hampton VA; Ayers: Analytical Services and Materials, Inc., Hampton, VA; 
Weaver: 88th Weather Squadron, Wright-Patterson Air Force Base, OH. 



iSa. bi5^RIBUTl6N/AVAILABILrrY STATEMENT 
Unclassified-Unlimited 

Subject Category 47 Distribution: Standard 

Availability: NASA CAST (301) 621-0390 



12b. DISTRIBUTION CODE 



13. ABSTRACT (Maximum 200 words) 

Surface observers stationed at 19 U.S. Air Force Bases and Army Air Stations recorded the daytime occurrence of 
contrails and cloud firaction on an hourly basis for the period April 1993 through April 1994. Each observation 
uses one of four main categories to report contrails as unobserved, non-persistent, persistent, and indeterminate. 
Additional classification includes the co-occurrence of cirrus with each report. The data cover much of the 
continental U.S. including locations near major commercial air routes. The mean annual frequency of occurrence 
in unobstructed viewing conditions is 13 percent for these sites. Contrail occurrence varied substantially with 
location and season. Most contrails occurred during the winter months and least during the summer with a 
pronounced minimum during July. Although nocturnal observations are not available, it appears that the 
contrails have a diurnal variation that peaks during mid morning over most areas. Contrails were most often 
observed in areas near major commercial air corridors and least often over areas far removed from the heaviest air 
traffic. A significant correlation exists between mean contrail frequency and aircraft fuel usage above 7 km 
suggesting predictive potential for assessing future contrail effects on climate. 



14. SUBJECT TERMS 

Contrails, Cirrus Clouds, Surface Observation, Subsonic Assessment, 

Aircraft Effects, Climate Change 



17. iebul^iVY CUeSIFICATION 
OF REPORT 

Unclassified 



18. SECURITY CLASSIRCAT1CN 
OF THE PAGE 

Unclassified 



19. SECURrrY CLASSIFICATION 
OF ABSTRACT 

Unclassified 



15. NUMBER OF PAGES 

83 



16. PRICE CODE 

A05 



20. UMrTATION 
OF ABSTRACT 



NSN 7540-01-280-5500 



Standard Form 298 (Rev. 2-89) 

Prescribed by ANSI Std. Z-3g-18 
298-1 oe