Applicability evaluation of CMADS dataset in Hulan River basin

Abstract

【Objective】 The purpose of this paper is to explore the applicability of China Meteorological Assimilation Datasets(CMADS) in hydrological simulation of Hulan River basin. 【Method】 The accuracy and spatiotemporal distribution characteristics of precipitation and temperature data from CMADS and traditional hydrological stations was analyzed, SWAT models driven by two types of meteorological data: CMADS and traditional hydrological stations was constructed, and SUFI-2 algorithm was used to calibrate and validate model parameters based on monthly measured cross-sectional runoff data. The applicability of this dataset as meteorological driven data and its substitutability for traditional hydrological station data were evaluated. 【Result】 ① The two types of meteorological data had strong linear correlation and corresponding relationships, and their spatiotemporal distribution characteristics were similar, with consistent trends within the year. ② In the main stream of Hulan River without reservoir, the evaluation index under the CMADS data-driven model were the rate regularly period R2=0.96, NSE=0.93, PBIAS=20.63%, and the validation period R2=0.98, NSE=0.97, PBIAS=7.51%; In the Tongken River system with a reservoir, the model had a regular flow rate of R2=0.88, NSE=0.80, and PBIAS=12.86%. During the validation period, the models had R2=0.97, NSE=0.96, and PBIAS=12.05%, and the simulation results all meet the evaluation criteria for runoff simulation. 【Conclusion】 CMADS+SWAT model had better overall effect, simulation accuracy and applicability than traditional hydrological stations in runoff simulation of Hulan River basin, and was closer to the measured value, which can provide data support for establishing SWAT model in the study area lacking meteorological data.

Keywords

• swat distributed hydrological model
• cmads data set
• hydrological simulation
• spatial and temporal distribution characteristics
• applicability evaluation