TRMM: The Tropical Rainfall Measuring Mission
563
are defined, normalized by surface precipitation rate. If the type of system can be defined by
its IR characteristics and the surface precipitation rate and fraction of stratiform precipitation
are determined from space by PR, TMI and VIRS, the heating profile can be inferred from a
look-up table.
20.6 TRMM Ground Validation Strategy
Since a rain radar has never flown in space, a concerted effort must be made to acquire surfacebased "ground truth" rain data to use as validation standards for the space-based instruments.
This is in itself a major undertaking, because there is no universally accepted standard for
ground-based large-scale rain estimation, given the highly variable nature of rain in space and
time.
To deal with this problem, TRMM is setting up a network of ground stations with a variety
of rain-measuring instruments in different tropical climate regimes to provide the best possible
long-term record of rainfall at selected locations for comparison to the satellite data. Figure
20.2 shows the distribution of currently operating and planned ground validation sites. The
network consists of four primary sites (Florida, Texas, Darwin, and Kwajalein) equipped with
various combinations of conventional rain gages, disdrometers, and rain radars (zenith pointing
or scanning). The Primary Rainfall Test Facility in Florida is intended as a testbed for surface
rain measurement techniques. It consists of a polarized multi-frequency microwave attenuation
link with a number of state-of-the-art rain gages underneath for calibration, as well as several
experimental rain gages (optical, Doppler). The four primary sites will be supplemented by
a number of secondary "Special Climatology" sites (Brazil, Taiwan, Thailand, Israel, Japan,
Hawaii, Guam, and perhaps India) whose mission is to provide several months of uninterrupted, high quality rainfall data per year to verify that the physics used by the TRMM satellite
algorithms properly translates to different geographical locations (Kummerow, personal communication).
The validation strategy has several components. Since the most accurate estimate of rainfall
on small scales comes from a dense network of gages, instantaneous space-based rainfall estimates at individual locations will be calibrated against gage data at the ground truth sites for
spacecraft overpasses of the sites. The gages in turn will be used to refine Z-R algorithms for
individual ground-based radars, and the larger-scale radar-derived rainfall patterns, averaged
over time to form a climatology, will be used to validate the 500 km x 500 km TRMM monthly
averages. The long time series obtained at the ground truth sites will also permit rainfall
statistics to be accumulated to provide a better understanding of the sampling issue and to
generate rainfall pdfs for use in some of the space-based instrument algorithms. Additional
data on vertical profiles of hydrometeors, drop size distributions, convective/stratiform separation, temperature and humidity soundings, and vertical velocities will supplement the basic
rain validation product. Finally, a number of airborne and shipborne Doppler radar data sets
will be acquired over the oceans adjacent to the ground truth sites, up to 200 km away from the
coast, to provide a comparison with the land-based estimates and an independent validation
standard for TRMM estimates of rainfall over ocean.
20.7 Conclusion
It must be emphasized that TRMM is a research mission to determine the optimum techniques
for measuring global rainfall on climatic time scales, rather than an operational rain mapping
563
are defined, normalized by surface precipitation rate. If the type of system can be defined by
its IR characteristics and the surface precipitation rate and fraction of stratiform precipitation
are determined from space by PR, TMI and VIRS, the heating profile can be inferred from a
look-up table.
20.6 TRMM Ground Validation Strategy
Since a rain radar has never flown in space, a concerted effort must be made to acquire surfacebased "ground truth" rain data to use as validation standards for the space-based instruments.
This is in itself a major undertaking, because there is no universally accepted standard for
ground-based large-scale rain estimation, given the highly variable nature of rain in space and
time.
To deal with this problem, TRMM is setting up a network of ground stations with a variety
of rain-measuring instruments in different tropical climate regimes to provide the best possible
long-term record of rainfall at selected locations for comparison to the satellite data. Figure
20.2 shows the distribution of currently operating and planned ground validation sites. The
network consists of four primary sites (Florida, Texas, Darwin, and Kwajalein) equipped with
various combinations of conventional rain gages, disdrometers, and rain radars (zenith pointing
or scanning). The Primary Rainfall Test Facility in Florida is intended as a testbed for surface
rain measurement techniques. It consists of a polarized multi-frequency microwave attenuation
link with a number of state-of-the-art rain gages underneath for calibration, as well as several
experimental rain gages (optical, Doppler). The four primary sites will be supplemented by
a number of secondary "Special Climatology" sites (Brazil, Taiwan, Thailand, Israel, Japan,
Hawaii, Guam, and perhaps India) whose mission is to provide several months of uninterrupted, high quality rainfall data per year to verify that the physics used by the TRMM satellite
algorithms properly translates to different geographical locations (Kummerow, personal communication).
The validation strategy has several components. Since the most accurate estimate of rainfall
on small scales comes from a dense network of gages, instantaneous space-based rainfall estimates at individual locations will be calibrated against gage data at the ground truth sites for
spacecraft overpasses of the sites. The gages in turn will be used to refine Z-R algorithms for
individual ground-based radars, and the larger-scale radar-derived rainfall patterns, averaged
over time to form a climatology, will be used to validate the 500 km x 500 km TRMM monthly
averages. The long time series obtained at the ground truth sites will also permit rainfall
statistics to be accumulated to provide a better understanding of the sampling issue and to
generate rainfall pdfs for use in some of the space-based instrument algorithms. Additional
data on vertical profiles of hydrometeors, drop size distributions, convective/stratiform separation, temperature and humidity soundings, and vertical velocities will supplement the basic
rain validation product. Finally, a number of airborne and shipborne Doppler radar data sets
will be acquired over the oceans adjacent to the ground truth sites, up to 200 km away from the
coast, to provide a comparison with the land-based estimates and an independent validation
standard for TRMM estimates of rainfall over ocean.
20.7 Conclusion
It must be emphasized that TRMM is a research mission to determine the optimum techniques
for measuring global rainfall on climatic time scales, rather than an operational rain mapping