基于GF-1遥感数据预测区域森林土壤有机质含量

为探索国产卫星GF-1预测土壤有机质(SOM)的能力，本研究以广东省云浮市的罗定市为研究区，以GF-1多光谱遥感影像衍生的9个遥感变量和DEM提取的9个地形水文变量为预测因子，建立2种人工神经网络模型(A模型：地形水文；B模型：地形水文+遥感)，对5个土壤深度(L1:0～20 cm,L2:20～40 cm,L3:40～60 cm,L4:60～80 cm,L5:80～100 cm)的SOM进行预测。结果表明：5个深度的B模型全都比A模型的精度高，尤其是L1、L2土层，精度提升明显，其R2分别提高了13%和10%；而深层土壤(L3、L4、L5)的精度提升较小，仅提高了4%、5%和4%。另外，两个评价指标RMSE和ROA±10%也表现出相似的趋势。总体而言，GF-1遥感数据显著改善了上层(0～40 cm)森林土壤人工神经网络模型的预测精度，对下层(40～100 cm)森林土壤模型改善尺度较低，是预测森林土壤SOM含量可观的新遥感数据源。

Prediction of Soil Organic Matter Content Based on Artificial Neural Network Model and GF-1 Remote Sensing Data

To explore the capability of GF-1 satellite to predict soil organic matter (SOM). In this study, Luoding City of Yunfu City, Guangdong Province was taken as the study area, and 9 multi-spectral remote sensing variables retrieved from GF-1 and 9 terrain variables derived from DEM were used as predictors to establish two kinds of artificial neural network models (Model A: terrain; Model B: terrain remote sensing) was used to predict the soil organic matter (SOM) at five soil depths (L1: 0~20 cm, L2: 20~40 cm, L3: 40~60 cm, L4: 60~80 cm, and L5: 80 ~ 100 cm). The results show that the accuracy of SOM full-variable B model at five depths is higher than that of A model with topographic variable only. Especially for the L1 and L2 layers of soil, the accuracy is obviously improved. The R2 of the L1 and L2 layers of SOM are increased by 13% and 10% respectively, and the highest R2 occurs in the L1 layer of SOM, reaching 88%. However, the accuracy of deep soil (L3, L4, L5) was only improved by 4% , 5% and 4%, and RMSE and ROA±10% also showed a similar trend. The results show that GF-1 remote sensing image can be used as a new remote sensing data source to predict SOM of soil, and provide a reference for the study of prediction other soil components by using GF-1 remote sensing image.