The purpose of this study is to compare with water quality prediction models for small and middle reservoir. These reservoirs has short rentention times, low depth, and weak stratification phenomenon.
Thus, this study was to develop program through ...
The purpose of this study is to compare with water quality prediction models for small and middle reservoir. These reservoirs has short rentention times, low depth, and weak stratification phenomenon.
Thus, this study was to develop program through a series of modeling procedure for water quality management and prediction of small and middle scale reservoir in Korea, the specific results are as follows.
First, The kind of state variables in model was focused on a series of water quality parameters connected with algae concentration which represent quantity of phytoplankton in water.
Spatial scales of Hoeong Reservoir was determined as morphological features. therfore Four models were applied to Hoedong Reservoir such as 'Modified Two-Box Model(MTB)', 'Completely Mixed Model' and 'Epilimnion-Hypolimnion Model', MINLAKE Model'. Prediction ability of developed models on this study was tried by means of regression coefficients, RMSPE(Root Mean Square Percent Error), and Absolute Mean Error with 'MINLAKE' model applicable to natural lakes.
Results showed that MTB(Modified Two-Box) model had higher prediction ability in comparison with 'Completely mixed' and 'Two-Box' model in chl-a, PO??-P, NH??-N, NO??-N concentration. Also, assuming that had correct input data, 'MINLAKE' model had high prediction ability comparatively.
But it was concluded that 'MTB' model had higher prediction ability in all aspects than complicated ecological model, 'Completely mixed' model and 'Two-Box' model in small and middle scale reservoir of Korea.