Temporal Variation Of The Parameters In A Monthly Water Balance Model

Under the impacts of climate change and human activities,the runoff series have been found to exhibit the non-stationarity in many areas.Against the background,the time-invariant parameters in the traditional hydrological model may be inappropriate.This study investigates the impact of time-varying parameters on model performance for monthly time scales.Therefore,the model parameters are bulit as the functions of the vegegation indice(Normalized Difference Vegetation Index,NDVI)and human activity indices(population,gross domestic production,grain production,and irrigated area)on eight typical catchments in China.The main results of this study are summarized as follows:(1)Firstly,the non-stationarity of the environmental factor series in the study areas is analyzed.The Mann-Kendall method and Spearman's rho test are employed to detect the trends of the climatic-hydrological indices(runoff,precipitation and potential evapotranspiration)and vegegation indice(NDVI)series in the study areas.And the Pettitt's test method is applied to identify the change-point of the hydro-meteorological series.The results show that the annual runoff series of Zhangjiashan,Xianyang,Hengjiang and Danjiangkou stations present a significant decreasing trend.The seasonal and annual runoff series of the study areas have the downward change-points except the Zhimenda station.The annual NDVI series of Zhimenda,Hengjiang,Xiajiang and Hengshi show a downward trend,and the other stations have an increasing trend.(2)Considering the impacts of vegetation and human activies,a theoretical framework is proposed to study the temporal variation of the parameters of the twoparameter monthly water balance model.The model parameters are bulit as the functions of the vegegation indice(Normalized Difference Vegetation Index)and human activity indices(population,gross domestic production,grain production,and irrigated area).Four scenarios of the temporal variation of parameters are considered.In comparison with the model with time-invariant parameters,the model with timevarying parameters can improve the simulations and provide more reliable predictions.(3)Based on the theoretical framework of the temporal variation of the parameters in the two-parameter monthly water balance model,the applicability of the model with time-varying is analyzed in the areas without observed potential evapotranspiration data.The Penman-Monteith equation and three temperature-based equations(i.e.,BlaneyCriddle,Hargreaves and Hamon)are selected to calculate the potential evapotranspiration of the study areas.The potential evapotranspiration data calculated by these equations are used as the input of the two-parameter monthly water balance model.The performances of all models under the different scenarios of the temporal variation of the parameters are compared.The results show that the monthly potential evapotranspiration data calculated by Blaney-Criddle equation is more reliable for providing runoff predictions.