基于时滞效应的无人机热红外遥感土壤含水率监测模型

Time-lag-effect-based UAV Thermal Infrared Model for Soil Moisture Monitoring

摘要:

作为调控作物水分状况和保障农业高效用水的关键参数，土壤含水率（SWC）的精准监测至关重要。为提高SWC遥感反演精度，本研究以冬小麦为对象，基于连续2022、2023年4梯度（灌溉上限设定为田间持水率的95%（W1）、80%（W2）、65%（W3）和50%（W4））灌溉试验，提出融合冠-气温度时滞效应的无人机热红外遥感监测方法。通过固定式红外温度传感器（SI-411型）和无人机热红外影像获取冠层温度，结合气象数据，采用峰值时间差法量化冠-气温度时滞特征，并解析其对水分梯度的响应机制。在此基础上构建4类考虑时滞效应的SWC监测模型（包括理论模型、经验模型、混合模型和冠气温差模型）。结果表明：W1~W3处理时滞时间没有显著差异，W4处理时滞时间降低，当土壤含水率低于田间持水率（50%~65%）阈值时，时滞时间显著缩短，而水分充足条件下差异不显著。引入时滞效应后，作物水分胁迫指数（CWSI）均值降低0.05~0.13，其中CWSI混合模型分布最合理。除理论模型外，考虑时滞效应的SWC监测模型R2均有提升，其中CWSI经验模型表现最优，2022、2023年土壤含水率反演精度分别为R2=0.53、RMSE为1.99%和R2=0.51、RMSE为2.13%。本研究验证了时滞效应对无人机热红外遥感监测土壤含水率的显著影响，为大尺度农田水分监测提供了理论基础与技术支撑。

Abstract:

Accurate monitoring of soil water content (SWC) is essential for evaluating crop water status and optimizing agricultural water management. Conventional thermal infrared remote sensing often neglects the temporal lag between canopy and air temperature responses, potentially introducing estimation bias. The canopy-air temperature time-lag effect was integrated into UAV-based thermal infrared remote sensing for improved SWC estimation. A two-year winter wheat experiment with four irrigation levels was conducted, using fixed infrared sensors (SI-411) and UAV thermal imagery. The peak-time difference method quantified the lag and its response to soil moisture gradients. Four time-lag-adjusted SWC models were developed, i.e., theoretical, empirical, hybrid and canopy-air temperature difference models. Results showed that lag times in W1~W3 treatments did not differ significantly, while W4 (low moisture) had shorter lags. Lag time decreased markedly when SWC fell below 50%~65% of field capacity. Introducing the lag effect reduced the mean crop water stress index (CWSI) by 0.05~0.13, with hybrid model producing the most physiologically consistent spatial patterns. Except for the theoretical model, R2 were improved, with the empirical CWSI model performing best (2022: R2=0.53, RMSE was 1.99%;2023: R2=0.51, RMSE was 2.13%). The findings can underscore the importance of thermal response dynamics for enhancing SWC estimation and provide methodological advances for large-scale UAV-based agricultural water stress monitoring.