Avaliação da acurácia das temperaturas da superfície do mar obtidas por satélite para a região. sul-sudeste da costa brasileira / South Brazil satellite sea surface temperature accuracy assessment

AUTOR(ES)
FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

27/06/1997

RESUMO

The accuracies of the global Advanced Very High Resolution Radiometer (AVHRR) split-window Sea Surface Temperature (SST) algorithms, derived from NOAA 11 and 12 satellites, were determined from comparisons with a match-up data set of COROAS project drifting buoy measurements for the southeast coast of Brazil, in the western part of the South Atlantic Ocean. The match-ups were screened from the HRPT scenes recorded by the INPE station in Cachoeira Paulista -SP, from February 1993 to June 1994. The temporal and the spatial coincidences are within 12 hours and between 2 Km and 12 Km. The NOAA 11 split-window global algorithm had a RMSD of 0.87°C with a computed bias of 0.24°C, providing better results than the NOAA 12 algorithm, wich presented. a RMSD of 0.94°C and a bias of 0.32°C. Seasonal comparisons showed that the algorithms for both satellites had best accuracies in the summer. In order to eliminate the bias and reduce the RMSD s, new regional algorithms were derived for the southeast coast of Brazil, with two different approaches. In the first, simple regression lines are used to correct the SST values, generated by the global algorithm. In the second, new split-window equations were derived by multiple regression models. When compared with the global algorithms, the first method equations showed an improvement of 0.18°C for NOAA 11 (RMSD = 0.69°C) and 0.06°C for NOAA 12 (RMSD = 0,88°C). The second method leads to a 0.09°C improvement for NOAA 11 (RMSD = 0,78) and 0.05°C for NOAA 12 (RMSD = 0,89). Those news RMSD s values are slightly greater than the RMSD value claimed for the global algorithm (0,07°C). This is probably due to a data concentration in the Brazil Current west wall region, where eddies and meanders are present in variable spatial scales, and high SST gradients. For more accurate regional SST estimation, a match-up data set with a wide temporal distribution is recommended for the split-window functions.

ASSUNTO(S)

oceanografia temperatura da superfície do mar correção atmosférica oceanografia satelites noaa sensoriamento remoto processamento de imagem. key word: avhrr sea surface temperature atmospheric correction oceanography noaa satellites remote sensing image processing atlântico sul

ACESSO AO ARTIGO

Documentos Relacionados