Accurate digital mapping of soil organic carbon (SOC) is important in understanding the\nglobal carbon cycle and its implications in mitigating climate change. Visible and near-infrared\nhyperspectral imaging technology provides an alternative for mapping SOC efficiently and accurately,\nespecially at regional and global scales. However, there is a lack of understanding of the impacts\nof spatial resolution of hyperspectral images and spatial autocorrelation of spectral information on\nthe accuracy of SOC retrievals. In this study, the hyperspectral images (380-1700 nm) with a spatial\nresolution of 1 m were acquired by Headwall Micro-Hyperspec airborne sensors. Then, hyperspectral\nimages were resampled into three dierent spatial resolutions of 10 m, 30 m, and 60mby near neighbor\n(NN), bilinear interpolation (BI), and cubic convolution (CC) resampling methods. The geographically\nweighted regression (GWR) model was used to explore the role of spatial autocorrelation in predicting\nSOC contrast with the partial least squares regression (PLSR) model. Results showed that (1) the\nhyperspectral images can be used to predict SOC and the spatial autocorrelation can improve the\nprediction accuracy, as the ratio of performance to interquartile range (RPIQ) values of PLSR and\nGWR were 1.957 and 2.003; (2) The SOC prediction accuracy decreased with the degradation of spatial\nresolution, and the RPIQ values of PLSR were from 1.957 to 1.134, and of GWR were from 2.003 to\n1.136; (3) Three resampling methods had a much weaker influence than spatial resolution on SOC\npredictions because the differences of RPIQ values of NN, BI, and CC resampling methods were 0.146,\n0.175, and 0.025 in the spatial resolutions of 10 m, 30 m, and 60 m, respectively; (4) Finally, the Global\nMoranâ??s I and the Anselin Local Moranâ??s I proved the existence of the spatial autocorrelation in SOC\nmaps. We hope that this study can offer valuable information for digital soil mapping by satellite\nhyperspectral images in the near future.
Loading....