基于BEPS-Terrainlab v2.0模型鄂西犟河流域1999年—2016年蒸散发模拟分析

Simulation and analysis of evapotranspiration in Jiang River in western Hubei during 1999-2016 based on BEPS-Terrainlab v2.0 model

蒸散发在水循环中占据着重要地位,采用模型模拟法研究小流域尺度的蒸散发具有高时间、空间分辨率的优势.该文选取湖北省十堰市犟河流域为研究区,基于BEPS-Terrainlab v2.0模型,利用Landsat5-8遥感影像和气象资料等数据,以日为时间步长模拟了研究区内1999年-2016年的蒸散发.模拟结果与MODIS蒸散发产品MOD16A2进行对比,二者年内变化相关系数R^2=0.92,MODIS数据时空分辨率均低于模拟数据,但也在一定程度上验证了模拟的可靠性.模拟结果显示:时间变化上,流域内蒸散发在年内呈单峰状分布(春季156.9 mm、夏季302.5 mm、秋季133.5 mm、冬季18.5 mm);年际上呈现波动变化,有一定周期性,总体有上升趋势.空间分布上,流域内蒸散发分布植被区高于非植被区,不同植被类型蒸散发不同(阔叶>混交>针叶>农田),不同高程蒸散发也不同(500~700 m区间内最大).经过相关性分析,气温、太阳辐射对蒸散发年内变化影响较大,通过了0.01的显著性检验;叶面积指数、降水对年际变化影响较大,通过了0.05的显著性检验.

Evapotranspiration plays an important role in the water cycle. Model simulation method has the advantages of high temporal and spatial resolution for the study of small watershed scale evapotranspiration. This study selects the Jiang River Basin in Shiyan City, Hubei Province as the research area. Based on the BEPS-Terrainlab v2.0 model, using Landsat5-8 remote sensing image and meteorological data to simulate the evapotranspiration of the study area from 1999 to 2016 by daily time step. The simulation results were compared with MODIS evapotranspiration product MOD16 A2. The correlation coefficient R^2 was 0.92 within the year, which verified the reliability of the simulation. The simulation results show that in temporal, the evapotranspiration is monomodal in the year(156.9 mm in spring, 302.5 mm in summer, 133.5 mm in autumn and 18.5 mm in winter);the evapotranspiration presents fluctuating changes inter-annual, with a certain periodicity, and an overall upward trend. In the spatial distribution, the evapotranspiration distribution in the vegETtion area is higher, and differ in vegETtion types(broad>leaf>coniferous>farmland), and evapotranspiration is different in different elevations(maximum within 500-700 m). According to correlation analysis, the variation of evapotranspiration within the year is significantly affected by temperature and solar radiation(P<0.01), and the annual variation is significantly affected by leaf area index and precipitation(P<0.05).