Recast of the Outputs of a Deterministic Model to Get a Better Estimate of Water Quality for Decision Makings

摘要

The output of a deterministic water quality model are retreated to recast the model simulation. A multi-variance regression method is used for daily model outputs versus observed data to assess the systematic errors in model simulation of individual model cell. The model outputs are re-adjusted to better represent the actual values, and to yield a better model calibration. Such a recast is important to model application for water quality analysis, especially for TMDL (Total Maximum Daily Load) which requires accurate simulation to exam criteria attainment. This paper addresses the recast method, its pre-requisites, and how the results are extended for the scenarios with various load reductions.