环境因子与玉米生长对地表温度监测土壤水分的影响

针对当前地表温度受太阳辐射、气象因素及作物生长状态影响，对早晨与傍晚土壤水分估算精度较差的问题，该研究在2020年夏玉米生长的拔节期与抽雄期，利用无人机搭载热红外传感器获取09:00、11:00、13:00、15:00以及17:00的地表温度数据，探究了太阳高度角、饱和水汽压差、植被覆盖度三者与地表-空气温差的相关性，提出了综合调整温度，构建了土壤含水率监测模型，分析模型在玉米吐丝期与水泡期的适用性并绘制了土壤含水率分布图。结果表明：1）同一时刻不同灌溉处理的地表温度与土壤含水率呈现负相关性，同一灌溉处理的地表温度日变化呈现上午升温较快下午降温较慢的负偏态分布趋势。2）太阳高度角正弦4次方根、饱和水汽压差、植被覆盖度与地表-空气温差的线性相关系数分别为0.509、0.948、-0.659。3）相比较基于地表温度构建的土壤含水率监测模型，基于综合调整温度的监测模型将决定系数由0.230提高到0.771，标准均方根误差由18.8%降低至10.3%。4）利用综合调整温度监测其他生育期的土壤含水率，决定系数由0.238提高到0.831，标准均方根误差由18.9%降低至9.5%，表明模型在玉米生长季的各个生育期的不同时段均有较强适用性。该研究可为无人机热红外遥感精准监测土壤水分亏缺状况提供参考。

Effects of environmental factors and maize growth on surface temperature to monitor soil water content

Abstract: Surface temperature can be one of the most important indicators to monitor the soil water deficit situation. However, the surface temperature depends mainly on solar radiation, meteorological factors, and crop growth state. The low accuracy of soil water content estimation cannot fully meet the requirements of precision irrigation, especially in the morning and evening. Fortunately, unmanned aerial vehicles (UAVs) with thermal infrared remote sensing can be expected to rapidly extract the surface temperature in recent years. In this study, an accurate and rapid detection of soil water content was proposed to clarify the effect of environmental factors and maize growth on the surface temperature using UAVs with thermal infrared remote sensing. Four gradients of irrigation treatment were set as severe water stress (T1), moderate water stress (T2), mild water stress (T3), and adequate irrigation (T4), whereas, the soil water content was controlled by the field capacity of 40%-50%, 50%-65%, 65%-80%, and 80%-100%. The layout of 12 plots was completely random, where each irrigation treatment was carried out with three replicates. Specifically, the jointing and sampling periods of summer corn growth were set on July 27, July 31, August 8, and August 10, 2020. Thermal infrared sensors were then used to obtain the field surface temperature at 09:00, 11:00, 13:00, 15:00, and 17:00 daily. The soil moisture content was collected simultaneously at a depth of 0-20 cm in the test area. Firstly, a systematic analysis was performed on the surface temperature at various time in the irrigation treatments. Secondly, three factors were selected as total solar radiation, meteorological parameters, and crop growth status. The influence of three factors on the surface-air temperature difference was then quantified using the solar height angle, vegetation coverage, as well as the pressure difference between saturated water and air. Thirdly, the comprehensive adjustment temperature was calculated using the surface temperature of 13:00. A monitoring model of soil water content was then constructed. Finally, the applicability of the monitoring model was verified at the silking and blistering stages. The results show that: 1) There was a negative correlation between the surface temperature in the different irrigation treatments and the soil water content at the same time. The daily change of the surface temperature in the same irrigation treatment presented a negative skewed distribution trend of fast warming in the morning and slow cooling in the afternoon. 2) The linear correlation coefficients were 0.509, 0.948, and -0.659, respectively, for the four power roots of the sine value in the solar height and surface-air temperature difference, the saturated water-gas pressure difference and surface-air temperature difference, as well as the vegetation coverage and surface-air temperature difference. 3) The coefficients of determination of the monitoring model using the integrated temperature increased from 0.230 to 0.771, and the normalized root mean square errors were reduced from 18.8% to 10.3%, respectively, compared with the model using surface temperature. 4) The comprehensive adjusted temperature was used to monitor the soil water content of other growth periods. The coefficient of determination increased from 0.238 to 0.831, and the normalized root mean square error was reduced from 18.9% to 9.5%, indicating the strong applicability in different growth periods of the maize growing season. This finding can provide a strong reference to accurately monitor the soil water deficit using the UAV thermal infrared remote sensing.