多源遥感蒸散产品在小流域径流模拟中的适用性评价

Assessment of the Applicability of Multi-source Remote Sensing Evapotranspiration Products in Small Watershed Runoff Modeling

在小流域径流模拟中,从多源蒸散发产品中选择合适的产品有利于提高水文模型径流模拟精度。从遥感蒸散发原理出发,选取了基于机器学习模型的GLASS产品、基于陆面模式的PET_PU_3H025_002产品、基于遥感模型的MOD16、GLEAM、SSEBop产品作为研究对象。以Littlewashita流域为研究区,通过求取蒸散发产品的区域多年均值分析其时空变化特征,并探究多源蒸散发产品在CREST水文模型中的适用性。根据NSCE、Bias、CC三项指标,结合单指标评价法和多指标综合评价法,比较了率定期和验证期模拟结果。结果表明:①5种遥感蒸散发产品多年均值的时空分布格局差异较大。PET_PU、SSEBop产品在区域内的估算值相对偏高,MOD16产品相对偏低。GLASS和GLEAM产品区域内多年均值的变化范围相似,均维持在2.2~2.4mm/d。②模型对洪峰时间的模拟较准确,但在洪峰流量的模拟方面存在差异。对比Bias指标,在率定期内GLEAM产品的偏差最小,MOD16的偏差最大。对比NSCE、Bias、CC三项指标,验证期径流模拟效果较率定期明显降低。③不同遥感蒸散发产品在小流域径流模拟中的差异性较小。综合多指标评价法,GLEAM和GLASS产品表现更出色。

In small watershed runoff simulation,the selection of suitable evapotranspiration products from multiple sources can improve the accuracy of hydrological model runoff simulation.In this study,based on the principles of remote sensing evapotranspiration,we selected the GLASS product based on machine learning models,the PET_PU_3H025_002 product based on land surface models,and the MOD16,GLEAM,and SSEBop products based on remote sensing models as our research objects.This study focuses on the Little Washita watershed,analyzing the spatiotemporal variation characteristics of evapotranspiration products by calculating their multi-year regional averages.Additionally,it explores the applicability of multiple evapotranspiration products within the CREST hydrological model.Utilizing the NSCE,Bias,and CC indices,coupled with single-index evaluation methods and comprehensive multi-index evaluation methods,we compare the simulation results between the calibration and validation periods.The results show that①The spatial and temporal distribution patterns of five remote sensing evapotranspiration products show significant differences in multi-year averages.PET_PU and SSEBop products tend to have relatively higher estimated values within the region,while MOD16 products show relatively lower values.Conversely,GLASS and GLEAM products demonstrate similar ranges of variation in their regional multi-year averages,consistently maintained between 2.2 mm/d and 2.4 mm/d.②The model exhibits relatively accurate simulation of peak timing,yet disparities exist in peak flow simulation.Comparing using the Bias index during the calibration period,GLEAM products demonstrate the smallest deviation,while MOD16 exhibits the largest.In the validation period,the effectiveness of streamflow simulation noticeably decreases compared to the calibration period,as indicated by NSCE,Bias,and CC indices.③The differences among various remote sensing evapotranspiration products in small watershed streamflow simulation is minor.Employing a comprehensive multi-index evaluation method,GLEAM and GLASS products demonstrate superior performance.