Article on Soil Organic Carbon Prediction is published in IEEE JSTARS

Congratulations to Chirag for this study which has compared nine nearest neighbor (NN) models with partial least squares regression (PLSR) to improve global soil organic carbon (SOC) estimation using spectral libraries. The optimized NN model achieved the highest accuracy, while PLSR showed better generalizability across datasets. The study also highlights that careful choice of performance metrics and calibration strategies is key to developing faster, more reliable SOC prediction tools.