Monitoring Change in the Arctic - Satellite-based Applications on Climate Change - page 128

Fig. 9.2 Time differences between successive overpasses of the Terra satellite ( a ) as a function of
latitude over the course of a 24-h period at the prime meridian and for both Terra and Aqua ( b ).
Only overpasses with sensor view angles less than 50 are considered
regions that cannot be filled by geostationary satellites because of poor viewing
geometry, polar-orbiting satellites are needed.
To estimate winds, clouds and water vapor features are tracked in sequential
images under the assumption that their movement represents the local airflow. It is
therefore necessary to track features over time in a sequence of images. Statistical
analyses of visible, infrared, and water vapor wind data sets from geostationary
satellites versus rawinsonde data have shown that the optimal processing intervals
are 5 min for visible imagery of 1 km resolution and 30 min for water vapor imagery
of 8 km resolution (Velden et al. 2000 ). How often can we obtain successive images
for wind vectors from a polar-orbiting satellite? The answer depends on the latitude
and the number of satellites. Figure 9.2a shows the frequency of time differences
between successive overpasses at a given latitude-longitude point during one 24-
h period with a single satellite (Terra). The points show only those overpasses
where the sensor, the Moderate Resolution Imaging Spectroradiometer (MODIS),
would view the Earth location at an angle of 50 or less. At larger scan angles, the
sensor would view the area near the pole on every overpass. At 60 latitude, there
are two overpasses separated by about 10 and 13 h. No useful wind information can
be obtained at this latitude with only one satellite. At 80 there are many views
separated by the orbital period of 100 min, but there is still a 13-h gap each day.
Although the 100-min temporal sampling is significantly longer than the optimal
processing intervals for geostationary satellites, in theory wind vectors can be
obtained during part of every day for the area poleward of approximately 70
latitude. Figure 9.2b shows the coverage with two satellites: Terra and Aqua.
Temporal gaps of a few hours still exist at the lower latitudes of the polar regions,
but at the higher latitudes, the temporal coverage is very good. With additional
satellites, e.g., the NOAA operational weather satellites with the Advanced Very
High Resolution Radiometer (AVHRR), it would be possible to obtain successive
views of a given location within minutes of each other.