基于TVDI的作物生育期农田表层土壤含水率时空动态演变

Temporal and Spatial Dynamics Evolution of Surface Soil Moisture Content in Agricultural Fields During Crop Growth Period Based on TVDI

摘要:

针对河套灌区农业用水效率偏低、水资源精细化水平不足的问题，本文基于多期Landsat-8/9数据构建作物生育期内的温度植被干旱指数（Temperature vegetation dryness index, TVDI），分析不同植被覆盖的地表温度（Land surface temperature，LST）-植被指数(Normalized difference vegetation index，NDVI)特征空间和干湿边变化特征。结合实测土壤含水率数据，评估TVDI模型在作物生育期农田土壤含水率反演精度与适用性，研究2022、2023年5—9月农田表层土壤含水率分布时空变化。结果表明：线性拟合干湿边时需排除偏离趋势的温度最值，NDVI为0.2~0.8区间内线性拟合得到干湿边方程，干边方程决定系数R2≥0.85，湿边方程R2为0.21~0.96。2022、2023年湿边斜率均呈一个周期余弦波规律变化。干边斜率均为负值，2022年干边斜率随TVDI均值由大到小再到大表现为坡度由陡变缓再变陡，2023年干边斜率变化不明显。干湿边截距先增后减，与地表温度变化趋势一致。土壤含水率变化显著，5月末—6月中旬下降，6月下旬开始上升，8月最高，9月末最低，且空间分布受灌溉和降水影响明显。TVDI模型能有效反映区域尺度土壤含水率时空变化，可为提高农业用水效率和制订精细化灌溉策略提供科学依据。

Abstract:

The issues of low agricultural water use efficiency and insufficient refined water resource management in the Hetao Irrigation District were addressed by constructing the temperature vegetation dryness index (TVDI) using multi-temporal Landsat-8/9 data. The characteristics of land surface temperature (LST)-vegetation index (normalized difference vegetation index (NDVI)) eigenspace and dry and wet side change characteristics of different vegetation covers were analyzed. Combined with the measured soil moisture content data, the accuracy and applicability of the TVDI model for the inversion of soil moisture content in farmland during the crop reproductive period were assessed, and the spatial and temporal variations of the surface soil moisture distribution in farmland from May to September in 2022 and 2023 were investigated, respectively. The results showed that the temperature maxima that deviated from the trend should be excluded when linearly fit the wet and dry edges, and the wet and dry edge equations were obtained by linear fitting in the interval of NDVI of 0.2~0.8, with the coefficients of determination of the dry edge equations of R2 not less than 0.85, and those of the wet edge equations of R2 ranged from 0.21 to 0.96. The slopes of the wet edges varied with a cyclic cosine wave law in both 2022 and 2023. The dry edge slopes were all negative, and the dry edge slopes varied from large to small to large with the mean TVDI value in 2022 showing that the slope changed from steep to gentle and then steep again, and the dry edge slopes changed insignificantly in 2023. The dry and wet edge intercepts increased and then decreased, which was consistent with the trend of surface temperature change. Soil moisture varied significantly, decreasing from late May to mid-June, increasing from late June, being the highest in August and the lowest in late September. The spatial distribution of soil moisture was significantly affected by irrigation and precipitation, and the TVDI model can effectively reflect the spatial and temporal changes of soil moisture at the regional scale, which can provide a scientific basis for improving the efficiency of agricultural water use and formulating refined irrigation strategies.