WASP 7.2와 신경망모형을 이용한 낙동강의 수질모의 및 예측

Abstract

Mulgeum, a downstream area in the Nakdong River, is exceeding in its standard of eutrophication during almost period of 2003~2005. This study is to simulate and predict the eutrophication of Mulgeum through WASP 7.2 model. For the simulation of eutrophication, the observed number of zooplankton is inputted and the result is compared with the observed results. And, through the Artificial Neural Network model, the zooplankton number is forecasted, and then the short term change of water quality is forecasted by inputting the air temperature data from the meteorological administration. As a result of the statistical analysis, the PBIAS of simulation by inputting the death rate of phytoplankton was -39.95 % and the PBIAS of simulation by inputting the number of zooplankton was 18.87 %. The simulation by inputting the number of zooplankton turned out more similar to the observed value. While the observed value of 2005 was fixed as initial condition, the number of zooplankton was forecasted through the Artificial neural network model by inputting DO, NO₃-N, PO₄-P, and air temperature data. Then, the short term change of water quality after 1~3 days was forecasted by inputting the air temperature from the meteorological administration instead of water temperature. As a result of forecast, the eutrophic materials in the water quality items such as Chlorophyll-A, total Nitrogen, and total Phosphorus were always simulated as appropriate to the observed values during the forecast period. It is considered that the water quality simulation by inputting the observed data in the water quality items, the forecasted zooplankton number, and air temperature data from the meteorological administration into WASP 7.2 model will be meaningful water quality forecast in the Nakdong River.