青藏高原高寒荒漠区土壤湿度监测仪器的校正方法探讨

Calibration of the soil moisture sensor in cold desert areas on Tibetan Plateau

ECH_2O EC-5土壤水分传感器能有效地监测土壤含水量,并且能揭示不同降水事件下的土壤水分动态变化状况。然而青藏高原主要土壤类型砾石含量高、昼夜温差大、秋季和春季冻融现象显著,导致监测数据出现偏差。基于对羌塘高原日土、改则地区布设的野外观测站点采集的土壤样品,对其进行土壤含水量测定的室内标定试验。结果表明:(1)各个样点均获得了较好的拟合效果,基于拟合系数校正后的测量值可以有效地反映实际的土壤含水量情况;(2)不同土层深度及不同样点的综合校准曲线精度明显低于单点的校准结果,主要受土壤容重和有机碳含量的影响,为获取更精准的校正数据,分层的校准是很有必要的;(3)基于已有校准模型校正后的值可以反映含水量变化的整体态势,但会产生高估或低估于真实值的偏离。

ECH_2O EC-5 soil moisture sensor can effectively monitor the soil moisture content,and it can reveal the soil moisture dynamic change under different rainfall events. However the main soil type has high gravel content on Tibetan plateau, and there is a large temperature disparity between day and night, especially the freezing and thawing phenomenon which frequently occurs during spring and autumn. This paper discusses the property of the EC-5 soil moisture sensor＇s measuring data of soil water content, and the soil sample are taken from the field observation sites in the Qiangtang plateau western Tibetan Plateau of China.The results show that all samples have good imitative effects：（1） According to the fitted line, the measured data which had already been corrected could efficiently reflect the practical soil moisture content most R2 of the fitting were over 0.97, the fitting curve of measured data after calibration and the real value is very close to the standard curve;（2） Comprehensive calibration results of different soil depth, there are some differences of regression straight line and the ideal straight line（Y = X）, when the water content is quite low the water content value after calibration is higher than the real value, when the water content is quite high the water content value after calibration is lower than the real value. When the water content between 0.06~0.12 cm3/cm3, the calibration result is more reliable.（3）The existing calibration model can reflect the overall situation of the water content changes, but may cause overestimate or underestimate the true value of the deviation.