PredictaMAR: A Species-Specific, Satellite-Based Weighted Scoring System for Artisanal Fishing Zone Prediction Using Copernicus Marine Service Data in the Humboldt Current System, Peru

Artisanal fishing fleets in Peru operate under conditions of high spatial uncertainty, with a significant proportion of fishing trips yielding low productivity in the absence of systematic oceanographic decision support. This paper presents PredictaMAR, a satellite-based fishing zone prediction system specifically designed for small-scale fishers operating within the Humboldt Current System (HCS). The system integrates nine oceanographic and environmental variables derived from the Copernicus Marine Service (CMEMS) — chlorophyll-a (Chl-a), sea surface temperature (SST), thermal gradient, Chl-a gradient, Chl-a stability, current velocity, salinity, bathymetry, and lunar phase — assigning differential species-specific weights to each variable according to the target species selected by the fisher. The scoring pipeline is fully deterministic: implemented in Python with automated data acquisition via the Copernicus Marine Toolbox API and output delivery via a Streamlit application, with no generative AI component in the scoring process. A field deployment with 24 artisanal fishing vessels across three coastal zones in Peru (Callao, Pucusana, and Ancón), comprising 120 registered fishing operations from January to December 2025, demonstrated an approximate 70% improvement in operational fishing effectiveness compared to traditional empirical methods, alongside reported monthly fuel savings of S/ 1,200–3,000 per vessel.