Pathfinder Contrail Studies

This page is devoted to research supported by the NASA Office of Earth Science Pathfinder Program under funding from the announcement NRA-99-OES-04.


  1. Objectives and Methods.
  2. Personnel.
  3. Examples of current research.
  4. Air Traffic Shutdown Study
    Contrail formation potential for September and October 2001.
    Air Traffic Denisty.
    Satellite Contrail Imagery.
    Contrail Analysis.
  5. Related References.

1. Objectives and Methods:

To better understand and model contrail formation and radiative effects, a comprehensive effort is underway to improve the measurement of contrail properties and the meteorological environment in which they form, develop climatologies of contrail conditions and contrail properties over the USA.

This study uses numerical weather analyses to diagnose the temperature and humidity levels at altitudes above 25,000 ft or pressures below 400 mb. Contrail formation conditions are determined with variants of the Appleman-Schmidt criteria from the observed temperature and humidity using assumptions about the propulsion efficiencies of modern commercial aircraft.

Flight paths including aircraft type and time are monitored, archived, and processed using near-real time input data from FlyteTrax (FlyteComm, Inc.). Flight paths from all commercial planes above 25,000 ft are monitored every 2 to 10 minutes depending on location. Data collection began September 1, 2000.

Contrails are being detected, measured, and tracked visually and automatically from multispectral satellite imagery. Automated techniques are applied to data from AVHRR on the NOAA satellites and to MODIS data on the Terra satellites. GOES data are used to track and monitor the spreading of contrails.

Microphysical properties are derived from the satellite imagery using a variety of techniques. Contrail coverage, temperature, altitude, optical depth, particle size, and radiative forcing are computed from the satellite data.

2. Personnel:

Patrick Minnis (NASA LaRC), PI: overall direction and management of the research.
David F. Young (NASA LaRC), Co-I: develop improved retrieval methods for cloud property determination.
David P. Duda (Hampton University), Co-I: develop contrail prediction and development programs.
Ulrich Schumann (DLR, Germany), Co-I: theoretical determination of contrail conditions .
Hermann Mannstein (DLR, Germany), Co-I: provide latest version of automated contrail detection algorithms .
Rabi Palikonda (AS&M, Inc.), Res. Scientist: adapt and implement automated techniques to derive a contrail climatology over USA.
Louis Nguyen (NASA LaRC), Res. Scientist: acquire and calibrate all satellite imager data .
Kay Costulis (NASA LaRC), Res. Scientist: acquire, reduce, and display aircraft flight data over USA .

3. Examples of current research:

3.1. Contrail formation potential for September and October 2001:
Figure 1. Potential persistent contrail frequency over CONUS for Sep 1 - Oct 1, 2001.
Figure 2. Potential persistent contrail frequency over CONUS for Oct 2001.

Contrails cannot form unless the conditions are right. These figures (fig. 1 and fig. 2 ) from Dr. David Duda show the distribution of the frequency of occurrence of conditions favorable for formation of persistent contrails over the USA. The results are based on hourly interpolations of 3-hourly RUC data smoothed to a 1-deg resolution. These numbers represent the integral over pressure (height) from 400 mb to 150 mb. All other things being equal, we should expect to see more contrails forming over the Pacific Northwest during both months (note the color scale change between months) than elsewhere. Other areas with high probabilities for contrail formation include the Great Lakes and Midwest. Favorable conditions over the Southeast during September gave way to more potential for clear skies during October. Note, the figures represent monthly means, therefore, on a particular day at a particular hour, the probabilities will be much different. Data are generated for each hour. Results like these will be used with satellite imagery and flight tracks to determine if the predictions are correct. If not, the prediction methods or input data will be adjusted to provide a more accurate assessment of the contrail probabilities.

3.2. Air Traffic Denisty:
Figure 3. Average daily frequncy of commercial airplanes over CUSA.

Contrails will not form unless an aircraft flies through the air mass that is suitable for contrail formation. The next figure ( fig. 3 ) from Kay Costulis shows the frequency of commercial airplanes that flew through a given 1-deg box in the USA at an altitude above 25,000 ft on September 10, 2001. The heaviest air traffic occurred over the eastern Midwest where more than 800 flights flew each day. These data are also available on an hourly basis also.

3.3. Satellite Contrail Imagery:
Figure 4. MODIS IR 11-12 um difference image over eastern USA at 1550 UTC, January 26, 2001
Figure 5. NOAA-16 AVHRR pseudocolor image over eastern USA at 1833 UTC, January 26, 2001
Figure 6. MODIS true color image showing contrails at 1610 UTC, November 13, 2001

When the air traffic and contrail conditions coincide, large outbreaks of contrails occur with subsequent development of extensive cirrus clouds, if cirrus clouds were not already present. This Terra MODIS infrared temperature difference image ( fig. 4 ) from Rabi Palikonda shows an outbreak of many contrails and cirrus clouds over Virginia, Ohio, West Virginia, and southern Pennsylvania at 1550 UTC, January 26, 2001. Three hours later, more contrails are forming and spreading as the air mass moves eastward. The pseudocolor NOAA-16 AVHRR image ( fig. 5 ) from Louis Nguyen shows these contrails and new ones with their associated cirrus clouds at 1833 UTC over eastern Virginia, Maryland, Delaware, New Jersey, and the Atlantic coast. The bright pink areas correspond to snow-covered regions. Another outbreak of contrails is seen over Ohio in this true color Terra MODIS image taken at 1610 UTC, November 13, 2001 ( fig. 6 ). Notice smoke plumes over Kentucky and Tennessee and contrails over North Carolina.

3.4. Contrail Analysis:
Figure 7. Contrail mask for the MODIS image over eastern USA at 1550 UTC, January 26, 2001
Figure 8. Frequency distribution of Optical Depth from the MODIS image at 1550 UTC, January 26, 2001

The following image ( fig. 7 ) shows an example of MODIS image at 1550 UTC, January 26, 2001 that is analyzed. The contrails covered 9308 sqkm, or 2.66% of the region shown in the image. The distribution of contrail optical depths derived is shown in Fig. 8. Most of the contrails have tau between 0.05 and 0.2, although contrails with optical depths as large as 1.0 are observed. On average, tau = 0.178. In this case, the unit radiative forcing is 7.3 Wm-2.

The results from these studies will be used to improve our estimation of the climatic effects of contrails.

  • Return to NASA Langley SASS Home Page
  • Return to our main NASA Langley Group Home Page

    This page last modified on March 26, 2008