基于垂直足迹线性加权算法的宇宙射线中子法土壤水分测量校准与验证

(1.重庆市气象科学研究所, 重庆 401147; 2.重庆市农业气象与卫星遥感工程技术研究中心, 重庆 401147; 3.重庆市江津现代农业气象试验站, 重庆 402260; 4.合肥市气象局, 合肥 230041; 5.永川区气象局, 重庆 402181)

宇宙射线中子法; 垂直足迹线性加权; 测量足迹等权重加权; 土壤水分

Improving Calibration and Validation of Soil Moisture Measurements Using Cosmic-Ray Neutron Method Based on Linear Depth Weighting
WU Qiang1,2,3, XU Qianqian4, ZHU Yuhan1,2,3, YAN Mengling5, LUO Zizi1,2,3

(1.Chongqing Institute of Meteorological Sciences, Chongqing 401147, China; 2.Chongqing Engineering Research Center of Agrometeorology and Satellite Remote Sensing, Chongqing 401147, China; 3.Chongqing Jiangjin Modern Agrometeorology Test Station, Chongqing 402260, China; 4.Hefei Meteorological Bureau, Hefei 230041, China; 5.Yongchuan Meteorological Bureau, Chongqing 402181, China)

cosmic-ray neutron method; linear depth weighting; footprint equal weighting; soil moisture

备注

为了研究宇宙射线中子法的不同测量足迹模拟对土壤水分反演结果准确性的影响,以烘干称重法为标准对比,频域反射法为趋势稳定性对照,基于垂直足迹线性加权方法对宇宙射线中子法反演区域土壤水分进行参数校准与测量结果验证,比较了相对于测量足迹等权重加权方法的区域土壤水分反演准确性。结果表明:相较于测量足迹等权重加权方法,垂直足迹线性加权处理的宇宙射线中子法区域土壤水分反演结果与通过烘干称重法获得的标准值一致性更好,反演模型准确性提高,与烘干称重法多点平均所代表的区域土壤水分值进行拟合,决定系数R2增加0.036,均方根误差RMSE减小0.002 3 kg/kg,纳什效率系数NSE提高0.041; 与频域反射法连续性测量结果比较,研究宇宙射线中子法连续性测量结果稳定性,结果显示除降水时段外,利用等权重加权与垂直足迹线性加权算法的结果趋势具有相似的一致性; 垂直足迹线性权重配比更符合不同深度土壤水分在宇宙射线中子法反演计算区域土壤水分的贡献差异,基于垂直足迹线性加权的宇宙射线中子法区域土壤水分反演结果能够更加客观地反映测量足迹内土壤水分的实际情况。
In order to study the influence of different footprint simulation of cosmic-ray neutron method on the accuracy of soil moisture inversion results, and the optimization effect relative to the footprint equal weighting method, the parameter calibration and result validation of soil moisture inversion by the cosmic-ray neutron method based on the linear depth weighting method, as the oven-drying method being the standard comparison, while the frequency domain reflection method being the continuous comparison. The accuracy of measurement results was compared with the standard value represented by the oven-drying method, the determination coefficient of linear equation(R2)increased by 0.036, the root-mean-square error(RMSE)decreased by 0.002 3 kg/kg, the Nash-Sutcliffe efficiency coefficient(NSE)increased by 0.041, the consistency of measurement results was better and the accuracy of inversion model was improved. The results of continuous measurement were compared with the frequency domain reflection method, the results showed no significant difference in trend stability between the two methods except the precipitation period. This study indicated that the inversion result based on the linear depth weighting method in improved consistency and accuracy of inversion model as the standard value obtained by oven-drying method, and the stability was close to the results of frequency domain reflection method in continuous observation compared with the footprint equal weighting method. The linear depth weighting was more consistent with the contribution of soil moisture in different depth. The soil moisture inversion by the cosmic-ray neutron method based on the linear depth weighting method can reflect the actual situation of soil moisture in the footprint more objectively.