A hyperspectral detection model for permeability coefficient of debris flow fine-grained sediments, Southwestern China

Abstract

ABSTRACT Fine-grained sediments are Quaternary sediments with grain sizes of not more than 2 mm. They start first when meeting water, their stability is related to the initial water volume triggering debris flow, and thus plays an important role in debris flow hazards early warning. The permeability coefficient is the inter-controlled factor of fine-grained sediment stability. However, there is no hyperspectral model for detecting the fine-grained sediment permeability coefficient in large areas, which seriously affects the progress of debris flow hazards early warning. Therefore, it is of great significance to establish a hyperspectral detection model for the permeability coefficient of fine-grained sediments. Taking Beichuan County, Southwestern China as the case, a permeability coefficient hyperspectral detection model was established. The results show that eight bands are sensitive to the permeability coefficient with a correlation coefficient (R) of 0.6343. T-test on the model shows that P-values for sensitive bands are all less than 0.05, indicating the established model has a good prediction ability with a precision of 85.83%. These sensitive bands also indicate the spectral characteristics of the permeability coefficient. Therefore, it provides a scientific basis for fine-grained sediment stability detection in large areas and lays a theoretical foundation for debris flow hazards’ early warning.