高级检索+

基于时空融合算法的关中地区灌溉信息遥感识别

Remote sensing identification of irrigation information in Guanzhong region based on spatiotemporal fusion algorithm

  • 摘要: 卫星遥感技术凭借快速、大范围观测的特点,在灌溉面积和灌溉频次等信息识别中具有显著优势。然而,单一遥感数据源受影像时空分辨率限制,难以满足区域尺度灌溉信息高精度动态识别需求。该研究以关中地区为研究区,构建了基于干旱指数分析与时空融合算法的灌溉信息遥感识别方法。通过干旱指数与土壤湿度相关性分析,选取温度植被干旱指数(temperature vegetation dryness index,TVDI)作为识别指标,并结合高程校正与融合策略优化,提升复杂地形条件下TVDI对土壤水分的表征能力及其时空融合精度。随后,采用增强时空自适应反射率融合模型融合Landsat与MODIS影像,生成高时空分辨率TVDI时间序列,结合阈值判别法识别2024年关中地区春灌信息。结果表明,经高程校正后的TVDI在春季作物生育阶段与土壤湿度相关性较高(R最高达-0.77);TVDI融合过程中,“先融合NDVI与LST再计算TVDI”的策略融合精度更高(R2=0.76,RMSE=0.07)。灌区尺度验证结果表明,东雷二期抽黄灌区灌溉面积识别总体精度为90.8%,Kappa系数为0.80;3个验证灌区累积灌溉面积与实际灌溉面积平均识别精度分别为87.1%和87.5%。关中地区春灌信息识别结果表明,春灌活动主要集中于3—4月,灌溉次数以1~2次为主,灌溉区域主要分布于渭河平原。研究结果可为复杂地形条件下区域尺度灌溉信息识别提供方法参考。

     

    Abstract: Satellite remote sensing has significant advantages for identifying irrigation information, including irrigated area and irrigation frequency, because of its rapid and wide-area observation capability. However, the spatial and temporal resolution of a single remote sensing data source is often insufficient to meet the demand for high-accuracy dynamic identification of irrigation information at the regional scale. Taking the Guanzhong region as the study area, this study developed a remote sensing method for irrigation information identification based on drought index analysis and spatiotemporal fusion. The temperature vegetation dryness index (TVDI) was selected as the identification indicator through correlation analysis between drought indices and soil moisture, and elevation correction together with fusion-strategy optimization was introduced to improve the capability of TVDI to characterize soil moisture and to enhance its spatiotemporal fusion accuracy under complex terrain conditions. Subsequently, the enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) was used to fuse high-spatial-resolution Landsat imagery and high-temporal-resolution MODIS data to generate a high-spatiotemporal-resolution TVDI time series, and spring irrigation information in the Guanzhong region in 2024 was identified using a threshold-based method. The results showed that the elevation-corrected TVDI was highly correlated with soil moisture during the spring crop growth period, with the maximum correlation coefficient reaching -0.77. In the fusion of TVDI, the strategy of first fusing the normalized difference vegetation index (NDVI) and land surface temperature (LST) and then calculating TVDI achieved higher accuracy than the strategy of first calculating TVDI and then performing fusion, with R2 = 0.76 and RMSE = 0.07 for the former, and R2 = 0.44 and RMSE = 0.13 for the latter. Validation at the irrigation-district scale showed that the overall accuracy of irrigated area identification in the Donglei Phase II Irrigation District was 90.8%, with a Kappa coefficient of 0.80, while the mean accuracies for accumulated irrigated area and actual irrigated area in the three validation irrigation districts were 87.1% and 87.5%, respectively. Regional identification results indicated that spring irrigation in the Guanzhong region was mainly concentrated from March to April, with irrigation frequency predominantly ranging from one to two times, and irrigated areas mainly distributed in the Weihe Plain. The results can provide a methodological reference for regional-scale irrigation information identification under complex terrain conditions.

     

/

返回文章
返回