The Use of General Inverse Problem Platform (GRIPP) as a Robust Backtracking Solution

This study addresses the challenge of identifying pollutant sources in aquatic coastal environments using inverse problem techniques hampered by particularities in hydrodynamic and Lagrangian models. An approach is presented employing the General Inverse Problem Platform (GRIPP) coupled with a General Simulated Annealing (GenSA) algorithm for robust backtracking. This methodology was applied to a hypothetical case study in Guanabara Bay, Brazil, using the MOHID Water platform for hydrodynamic and Lagrangian simulations. GRIPP significantly improved emission identification and pathway representativeness compared to traditional backtracking methods by exploring multiple potential particle origins and optimizing seeding parameters. The optimization yielded a solution with a mean error of 0.019 degrees between predicted and observed tracer locations. This demonstrates the potential of GRIPP and GenSA for solving practical environmental problems in coastal regions. The use of GRIPP can bypass eventual numerical errors in cases of long-term pollution source identification when compared to traditional methods. This approach contributes with blended models for the identification of potential sources of Lagrangian tracer emissions, as well as a potential solution to determine the pathways of pollutants.