Study on Band Selection and Optimal Spectral Resolution for Prediction of Cu Contamination in Soils

Hyper-spectral data offers a powerful tool for predicting soil heavy metal contamination due to its high spectral resolution and many continuous bands.Band selection and spectral resolution,however,are the prerequisite of heavy metal inversion by hyper-spectral data.In this study,soil reflectance spectra and their Cu contents were measured for 181 soil samples in the laboratory.Based on these dataset,band selection was conducted to inverse Cu content using stepwise regression approach,and prediction accuracies of Cu based on partial least-squares regression(PLSR) model with different selected bands were analyzed.In addition,the influences of spectral resolution on prediction results of Cu were discussed by a Gaussian re-sampling function.It demonstrated that the optimal band number was 10 for Cu inversion and the corresponding model prediction accuracy was R2 0.7523 and RMSE of 0.4699.The optimal spectral resolution was 32 nm and the model on this basis had an accuracy of R2=0.7028 and RMSE=0.5147.Results of this paper may provide scientific verification for designing low-cost and practical hyper-spectral space-borne sensors and provide theoretical bases for simulating space-borne sensors to predict soil heavy metals content in the future.