State Water Resources Research Institute Program

Project ID: 2012DC144B
Title: Water Pollution Modeling and Prediction Using Computational Intelligence Methods
Project Type: Research
Start Date: 3/01/2012
End Date: 2/28/2013
Congressional District: DC
Focus Categories: Hydrology, Non Point Pollution, Water Quality
Keywords: Nonpoint Source Pollution, Urban Runoff, Stormwater Management, Water Quality, Neural Networks, Prediction
Principal Investigators: Zhang, Nian (University of the District of Columbia); Behera, Pradeep K.
Federal Funds: $ 13,000
Non-Federal Matching Funds: $ 5,980
Abstract: To restore and maintain the physical chemical and biological integrity of water bodies in the United States, the Clean Water Act requires the development of Total Maximum Daily Loads (TMDL) for impaired waters [1]. A TMDL is the maximum amount of a pollutant that a waterbody can receive and still meet water quality standards. In the last decade, several models and methodologies have been proposed emphasizing system wide modeling of the bodies of water and the uncertainties associated with TMDL decisions. However, these methods are still evolving and their implementation is very limited. The Environmental Protection Agency (EPA) has emphasized that there is a need for the development of new and more flexible modeling systems, tools, Internet- based technologies, integrated modeling systems combining new solution techniques, source representation, new algorithms and decision-making tools to support the development of tens of thousands of TMDLs nationally, including a number of TMDLs for the Chesapeake Bay [2][3].

The broader goal of this research project is to assist in developing an innovative computational intelligence- based approach for optimizing TMDL by allocating pollutant loads from different sources. Our approach comprises two components: 1) computational intelligence- based algorithm for TMDL optimization via load allocation, and 2) Backward Propagation Recurrent Neural Network (BP-RNN) for missing data estimation, future flow prediction, and water quality prediction. This research will utilize the public domain Chesapeake Bay Watershed Model 5.3. Our proposed research will be applied to a specific TMDL segment of the Anacostia River for a specific pollutant (e.g., Phosphorus). Similar to other metropolitan cities, during wet weather events, when the capacity of combined sewer is exceeded, the excess flow, which is mixture of runoff and sewage, is discharged to the Anacostia and Potomac Rivers, Rock Creek and tributary waters through the sewer outfalls. The excess flow is called as Combined Sewer Overflow (CSOs). There are a total of 60 CSO outfalls in the combined sewer system listed in the National Pollutant Discharge Elimination System (NPDES) permit issued by the Environmental Protection Agency to WASA. The discharges from the separated storm sewer system generally directs to the river systems without any treatments. The CSOs and stomwater discharges known as urban wet weather flow or urban runoff can adversely impact the quality of the receiving waters.

The prediction of water quality (e.g., pollutant load) is dependent upon the reliable water quantity prediction (e.g., flow and runoff). To improve the water quantity prediction, BP-RNN will be used to fill in any missing input data and predict future data. After that, a watershed model will be used to simulate the pollution condition within the TMDL segment and associated loading allocation of point sources (WLAs) and allocation of non-point sources (LAs) from various point and non-point sources. Outputs of this watershed model are time series of daily flows, pollutant loads and concentrations and will be fed to the proposed computational intelligence-based algorithm as input. The algorithm will then optimize TMDL via load allocation. The proposed research is very useful for the District of Columbia because the District is located within the Chesapeake Bay watershed.

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Earl Greene
Page Last Modified: Tuesday, 15-Jan-2013 02:37:57 EST