Application of Generalized Additive Model for Identification of Potential Fishing Zones Using Aqua and Terra MODIS Imagery Data

Remote sensing applications can provide information on oceanographic conditions for identification of potential fishing zones by combining statistical approaches. Determination fish catch needs to be studied the relationship between parameters and catches improve efficiency effectiveness operations fishermen. Based this, examine using Generalized Additive Model (GAM) in Java Sea. GAM analysis was carried out data as response variables such sea surface temperature (SST) chlorophyll-a image processing results from MODIS, SSS CMES, Depth predictor variables. The selection best model is determined highest percentage CDE lowest AIC. modeling show that 60.3% Sea are influenced factors 39.7% other factors. parameter has most influence concentration chlorophyll-a. tend found have concentrations 0.2 mg/m3 – 0.5 mg/m3, SST 280C 310C, salinity 31.8 PSU 33 PSU, a depth 20 m. 50 meters. Potential were identified based analysis. March 2021 June varying spatial distributions. 3, 2021, which distributed around Pulau Laut, southern part island Borneo, north Madura.