Fish population spatial distribution data provide essential information for fleet monitoring and fishery spatial planning. Modern high resolution ocean color remote sensing sensors with daily temporal coverage can enable consistent monitoring of highly productive areas, giving insight in seasonal and yearly variations. Here is presented the methodology to monitor small pelagic fish spatial distribution by means of 500m resolution satellite data in a geographically and oceanographically complex area. Specifically, anchovy (Engraulis encrasicolus) and sardine (Sardina pilchardus) acoustic biomass data are modeled against environmental proxies obtained from the Sentinel-3 satellite mission. Three modeling techniques (Logistic Regression, Generalized Additive Models, Random Forest) were applied and validated against the in-situ measurements. The accuracy of anchovy presence detection peaked at 76% and for sardine at 78%. Additionally, the spatial distribution of the models’ output highlighted known fishing grounds. For anchovy, biomass modeling highlighted the importance of bathymetry, SST, and the distance from thermal fronts, whereas for sardine, bathymetry, CHL and chlorophyll fronts. The models are applied to a sample dataset to showcase a potential outcome of the proposed methodology and its spatial characteristics. Finally, the results are discussed and compared to other habitat studies and findings in the area.