Improved SST estimates for AVHRR from ATSR data through a neural network approach
- Univ. of Reading (United Kingdom)
Many applications such as weather forecasting, oceanography and to determine the possible climate change need frequent data such as Sea Surface Temperature (SST) with wider spatial coverage and reliable accurate estimates. Even though the sampling requirements for SST`s can be met from Advanced Very High Resolution Radiometer (AVHRR) data, the accuracy required from these SSTs remains a challenge. The Along Track Scanning Radiometer (ATSR) retrieved SST`s offer better estimates than the AVHRR Multichannel Sea Surface Temperature (MCSST) algorithm, but the swath width of the ATSR is 512 km and the repetition cycle is approximately 3 days. In this study an attempt has been made to generate a new SST retrieval model for AVHRR using the SST`s retrieved from ATSR data. A multilayer neural network approach is employed to generate the model parameters. The new approach for the retrieval of SST for AVHRR data using the neural network yields the residual error within {plus_minus} 0.24{degrees}C when compared with the ATSR derived SSTs.
- OSTI ID:
- 508186
- Report Number(s):
- CONF-960384--
- Country of Publication:
- United States
- Language:
- English
Similar Records
The validation of ATSR measurements with in situ sea temperatures
Heat flux exchange estimation by using ATSR SST data in TOGA area