1. Recalibration of Spectral Models Using Spiking Techniques for Predicting Primary Nutrient Attributes

Soil spectral libraries are an important strategy for rapid prediction of soil fertility atributes in digital agriculture projects. However, their predictive performance may decline when models are applied outside the specific conditions for which they were calibrated. Even in regions with similar pedoclimatic characteristics, management practices can limit model accuracy. In this context, spiking-based recalibration has been proposed as a practical strategy to improve model performance, although... A. Ten Caten, V. Ormeño, J.A. Henriques, M.M. Reva, M.S. Silva