Studies On The Evaluation System For Surface Water Quality Models

Surface water quality models are cost-effective, time-saving, highly efficient and flexible, and nowadays they play an important role in a variety of studies concerning water environment monitoring, water quality simulation, management and recovery. A scientific evaluation system for water quality models is therefore needed in order to choose the right model and to achieve reliable results in engineering applications, or to facilitate decision making processes for water environment related issues. Such an evaluation system should include evaluation parameters and evaluation procedures which can quantify the reliability and operability of a surface water quality model, and determine its suitability for formulating a legally binding documentation or for application on a specific engineering project.An evaluation system for surface water quality models is established in this paper, based on the following research works:By employing the literature metrology, a comprehensive review of 54 most-widely-used water quality models is conducted to check if each of the model’s popularity is consistent with its capability. A model’s popularity is marked mainly by its citation rate, and its capability is characterized by its requirement for data and user experiences, technical support, application scope and simulated items, etc. A model’s availability is also considered in the review. It is found out that a model’s popularity is not necessarily consistent with its capability.An evaluation parameter system of surface water quality models is put forward based on the above-mentioned review of model capability. On the basis of AHP method, an evaluation procedure for surface water quality models was proposed. A standardized evaluation procedure is suggested and a standard evaluation index system is developed.A quantitative method to categorize the complexity of various lab flume experiments is presented based on flow structures and geometric conditions, and a complexity parameter ? is proposed to quantify the complexity of various lab experiments for flow in a bend or consecutive bends. Such a method not only evaluates the complexity for each experiment but also serves as a reference for the selection of typical test cases to validate numerical models, because accurate simulation of flow structure is vital to the reliability of water quality models. Meandering river is one of the most common channel patterns in natural rivers, which involves flow in consecutive bends. A surface water quality model’s capability to accurately simulate flow structures in consecutive bends is therefore directly related to its reliability in simulating sediment and contaminant transport in natural conditions. Chang’s experiment and He’s experiment are selected as typical test cases by using this complexity parameter. The methods of average 2-norm is chosen as the error analysis method in the proposed evaluation procedures.Based on the proposed evaluation parameter and evaluation method of surface water quality models, 5 representative models were evaluated for the 3D hydrodynamic and non-point source pollution simulations, respectively, by using the standardized evaluation procedures. The 5 models are SWAT, ELCIRC, MIKE 3FM, 3DMMC and HEC-6.These models are applied to the above-mentioned typical test cases for quantitative analysis based on the proposed evaluation parameters and error analysis methods.