| Title |
Improvement of Mid-and Low-flow Estimation Using Variable Nonlinear Catchment Wetness Index |
| Authors |
현석훈(Hyun, Sukhoon) ; 강부식(Kang, Boosik) ; 김진겸(Kim, Jin-Gyeom) |
| DOI |
https://doi.org/10.12652/Ksce.2016.36.5.0779 |
| Keywords |
유역습윤지수;비선형성;장기 유출;IHACRES 모형 Catchment wetness index;Nonlinearity;Long-term runoff estimation;IHACRES model |
| Abstract |
The effective rainfall is calculated considering the soil moisture. It utilizes observed data directly in order to incorporate the soil moisture into the rainfall-runoff model, or it calculates indirectly within the model. The rainfall-runoff model, IHACRES, used in this study computes the catchment wetness index (CWI) first varying with temperature and utilize it for estimating precipitation loss. The nonlinear relationship between the CWI and the effective rainfall in the Hapcheondam watershed was derived and utilized for the long-term runoff calculation. The effects of variable and constant CWI during calibration and validation were suggested by flow regime. The results show the variable CWI is generally more effective than the constant CWI. The $R^2$ during high flow period shows relatively higher than the ones during normal or low flow period, but the difference between cases of the variable and constant CWI was insignificant. The results indicates that the high flow is relatively less sensitive to the evaporation and soil moisture associated with temperature. On the other hand, the variable CWI gives more desirable results during normal and low flow periods which means that it is crucial to incorporate evaporation and soil moisture depending on temperature into long-term continuous runoff simulation. The NSE tends to decrease during high flow period with high variability which could be natural because NSE index is largely influenced by outliers of underlying variable. Nevertheless overall NSE shows satisfactory range higher than 0.9. The utilization of variable CWI during normal and low flow period would improve the computation of long-term rainfall-runoff simulation. |