Operational sea surface temperature (SST) monitoring was tested with the study of large-scale SST features in Indonesia from July 1981 to June 1985. Digital data were provided by digitizing 208 weekly SST charts of NOAA-NESS. The pixel size corresponded to 1 #xB0; 15' longitude and latitude. These data displayed a 0#xB7;5 deg K. r.m.s. error compared with 1985in situmeasurements. Iterative interactive factorial analyses combined with a parallelepiped classifier as a clustering technique enhanced the SST spatio-temporal features. The study area was divided into zones in which pixels had similar SST profiles and dates of occurrence of thermal anomalies were pointed out. Sea fronts and upwellings were mapped through spatio-temporal analyses of thermal gradients. This study stresses the possibility of operational SST monitoring for Indonesia, allowing simple data manipulation with hardware easily maintained locally.
展开▼
