Abstract: The habitat model is one of the important methods of fishery forecasting, which is highly applicable and accurate. According to the survey data of "Nanfeng" Fishery Research Vessel collected from Nansha Islands waters in 2013 in addition to data of acoustic navigation and environmental data from remote sensing, we applied biomass as adaptive index and established the sea surface temperature (SST) model based on concentration of Chlorophyll a (CHL), sea surface height (SSH) and sea surface salinity (SSS) by using maximum value method, minimum value method, geometric mean value method and arithmetic mean value method. The result shows the arithmetic mean value fits the best. The accuracy of the model was above 75%. The area of high biomass mainly distributed in the area of high HSI which was more than 0.5. The other areas of high HSI might be potential fishing grounds.