Geoscience Reference
In-Depth Information
calibration has been made, future satellite imagery used to make estimates of
probable precipitation.
Estimates of rainfall rate are still made using knowledge of the extent and
nature of cloud but it has become usual to merge such knowledge with additional
meteorological information derived from satellite observations. The
TOVS
precipitation estimate
and the
AIRS precipitation product
are examples. Data from
the Television Infrared Operational Satellite (TIROS) Operational Vertical Sounder
(TOVS) instruments on board polar-orbiting platforms and AIRS instrument
aboard the Earth Observing System Aqua polar-orbiting satellite are both processed
to provide meteorological statistics. These two precipitation products then infer
precipitation from deep extensive clouds using a multiple regression relationship
between collocated rain gauge measurements and the several satellite data streams
(related to cloud volume, cloud-top pressure, fractional cloud cover, and relative
humidity profile). The relationship used is allowed to vary with season and latitude
and separate relationships are used for ocean and land.
Passive measurement of cloud properties
The essential basis of this approach is to calibrate and exploit empirical relation-
ships between precipitation at a particular location and physical characteristics of
the overlying cloud measured by remote sensing systems. The most commonly
used physical characteristic is the brightness temperature of the top of cloud as
diagnosed by the measured outward longwave radiation (OLR). Colder cloud top
temperature implies deeper clouds with more convective ascent and a greater like-
lihood of rain. Estimates based on cloud top temperature have greatest probability
of success in regions and at times where convective rainfall is dominant, i.e., in the
tropics, and in the summer season elsewhere.
One significant satellite precipitation estimate based on this approach is the
Global Precipitation Index
(
GPI
), which uses three-hourly infrared images provided
by three geostationary satellites. An empirical relationship based on data from the
Global Atmospheric Research Programme (GARP) Atlantic Tropical Experiment
(GATE) is used for this product. For a brightness temperature
235 K, a rain rate of
3 mm hr
−1
is assigned, while for a brightness temperature >235 K, a rain rate of
0 mm/hour is assigned. The GPI estimate works best over space and time averages
of at least 250 km and 6 hours, respectively, and in oceanic regions with deep
convection. Several other precipitation products based on OLR are also available.
Some, such as the
PERSIANN
precipitation product, have developed
interrelationships using neural networks rather than using fixed formulae or linear
regression.
At microwave frequencies, the atmosphere is almost transparent where no
precipitating clouds occur. Where there is high humidity and particularly where
there are precipitating clouds, microwave emissivity and absorption is greater.
Consequently, data from satellite sensors operating in the microwave waveband
can contribute information relevant to remote sensing-based precipitation
≤
