《Table 5 The modeling and predicted abilities of prediction models based on evaluation indexes》
本系列图表出处文件名:随高清版一同展现
《Combining Environmental Factors and Lab VNIR Spectral Data to Predict SOM by Geospatial Techniques》
Notes:PLSR:partial least squares regression model,MLR:multiple linear regression model,GWR:geographically weighted regression model,R2C:the coefficient of determination of calibration,R2P:the coefficient of determination of prediction,RMSEC:the root m
The distribution of SOM along with some of the predictors was described using classical descriptive statistics(Table 1).SOM values ranged from 17.34 to 50.50 g/kg,and the mean value was 30.12 g/kg and the median value was 28.96 g/kg.The standard variance was 6.69g/kg and the coefficient of variation(CV)was 22.21%,suggesting that SOM had a moderate spatial variation based on Wilding(1985).Six variables were selected as the predictors to construct the SOM prediction models.Table 1 shows the basic statistics of the other predictors.The mean values of elevation,ILUT,LSWI,EVI,RDLS and NDC were 27.47 m,3.06,0.08,0.12,9.28 and2390.67 m,respectively.The CVs of these auxiliary variables indicated that all of these predictors had a different degree of spatial variation across the study region.NDC had the greatest variation(75.71%)and ILUT had the least variation(13.35%).More detailed basic statistics of SOM and its environmental factors can be found in Table 1.
图表编号 | XD0046961800 严禁用于非法目的 |
---|---|
绘制时间 | 2019.04.01 |
作者 | GUO Long、ZHANG Haitao、CHEN Yiyun、QIAN Jing |
绘制单位 | College of Resources and Environment, Huazhong Agricultural University、College of Resources and Environment, Huazhong Agricultural University、School of Resource and Environment, Wuhan University、School of Resource and Environment, Wuhan University |
更多格式 | 高清、无水印(增值服务) |
查看“Table 5 The modeling and predicted abilities of prediction models based on evaluation indexes”的人还看了
- Table 1.The current status of coupled models and coupled assimilation in world modeling and prediction centers applied i
- Table 4 Evaluation of predictive performance of the model by using receiver operating characteristic method
- Table 3 Prediction performances of soil depth using the five different combinations and two prediction methods (Random F