|
The next generation of passive microwave satellites (i.e. Windsat, AMSR on AQUA, and CMIS on NPOESS) provide an
opportunity to develop sea surface temperature (SST) products that are less affected by errors due to water vapor, aerosols and
clouds than IR-based SST. Although wind effects, undetected rain and large spatial resolution will provide challenges for
algorithm developers, the advantages of an all-weather SST product are numerous.
As a first step towards a microwave SST
product for pre-NPOESS (Windsat) and NPOESS era microwave satellites, a regression-based algorithm has been developed for
the TRMM Microwave Imager (TMI). Using brightness temperatures from the 10.7 GHz channel and 3 years of buoy matchups
(1998-2000), a preliminary algorithm has been developed for near real-time implementaion.
Results of this algorithm have been
compared to other data sources (IR-based SSTs as well as other in-situ data), and will be presented with sample SST products.
Additional improvements in the accuracy should be gained by the future development of algorithm coefficients based upon
radiative transfer models.
More information on the equations used to calculate sea surface temperature from TMI data and the validation program
obtained from Dr. Tim Mavor.
The Binary Data files (GeoTIFFs are scaled differently) for TMI-SST contain 8-bit data (0 - 255) using the following criteria:
| Value
|
Use
|
| 0
|
Land
|
| 1
|
No Data
|
| 2
|
Rain
|
| 3 - 255
|
Scaled SST values where Degrees_C = 0.15 * SSTscaled - 3.15 for sst byte values between 20 and 255
|
|
