A Framework for Retrieving Soil Organic Matter by Coupling Multi-Temporal Remote Sensing Images and Variable Selection in the Sanjiang Plain, China

Soil organic matter (SOM) is an important soil property for agricultural production. Rising grain demand has increased the intensity of cultivated land development in Sanjiang Plain China, and there a strong SOM monitoring this region. Therefore, Baoqing County Plain, production area, was considered study area. In study, we proposed framework high-accuracy retrieval by coupling multi-temporal remote sensing (RS) images variable selection algorithms. A total 73 surface samples (0–20 cm) were collected 2010, Landsat 5 acquired during bare period (April, May, June) selected from 2008 to 2011. Three algorithms, namely, Genetic Algorithm, Random Frog Competitive Adaptive Reweighted Sampling (CARS), combined with Partial Least Squares Regression (PLSR) build models on spectral bands indices images. The results using single-date image showed that combination algorithms PLSR outperformed alone, CARS best performance (R2 = 0.34, RMSE 15.66 g/kg) among all only applied different year interval groups. To investigate effect acquisition time, divided into various groups, resulting then stacked. accuracy improved as lengthened. optimal result 0.59, 11.81 obtained 2008–2011 group, wherein difference derived 2009, 2011 dominated variables. Moreover, spatial prediction based model consistent distribution SOM. Our suggested couples stacked RS potential retrieval.