This algorithm is based on inorganic nitrogen, active phosphate and chemical oxygen demand data obtained from actual measurements in key ports in the Zhejiang Province. The remote sensing data used are the remote sensing of Sentinel-2 A and B satellites with a spatial resolution of 10 meters. According to the data, the return period is 5 days under the combined clear sky conditions of the double stars. Satellite-ground matching is performed based on the longitude, latitude and date information of the measured data. Machine learning algorithms (support vector machines and Gaussian regression models) are used, and satellite remote sensing band reflectivity is used as input to conduct inorganic nitrogen, active phosphate and chemical oxygen demand respectively. Modeling, and then produced a long-term series product for nearly 9 years from 2015 to 2023.

The time span of the inorganic nitrogen remote sensing dataset of key harbors in the Zhejiang Province is from 2015 to 2023, with a spatial resolution of 10 m and a temporal resolution of 5 days under clear weather conditions.