Multi-scale collaborative monitoring method and system for agricultural non-point source pollution

Abstract

This application provides a method and system for constructing multi-scale collaborative monitoring of agricultural non-point source pollution. The method includes: constructing field units, transport units, and convergence units; establishing cross-scale mapping relationships from field units to transport units and from transport units to convergence units based on hydrological connectivity and irrigation / drainage engineering topology; generating evidence objects associated with unit identifiers and writing these evidence objects into an evidence index; performing constrained fusion updates on a pollution state inference model based on event-level joint observation data to output pollution state estimates for each unit, convergence unit load estimates, and corresponding uncertainty intervals; generating a collaborative monitoring construction strategy based on the uncertainty intervals and cross-scale mapping relationships, and outputting the collaborative monitoring construction strategy for constructing or updating a multi-scale collaborative monitoring network for the target area. This application can improve cross-scale monitoring consistency, increase the accuracy of event peak capture, enhance the reliability of fusion results, and improve network update efficiency.

Images