Development of spectral band cloning techniques for soil nutrients estimation

Satellite‐based monitoring is an indispensable tool to guide soil‐specific crop management. However, it has attained little success in the estimation of soil nutrients due to the limitations incurred from inherent spectral characteristics. In this study, spectral band cloning (SBC) is developed and proposed to augment the soil nutrient predictive capabilities of broadband satellite data. Fine‐spectral channels of spectrometers were synchronized with coarse resolution of IRS satellite data to generate nutrient‐sensitive cloned IRS bands. Soil samples, collected at the time of satellite image acquisition in Lop Buri, Thailand, were analyzed both spectrally and chemically, viz., soil organic matter (OM), phosphorus, potassium and iron. The resulting SBC‐based models showed acceptable correlations, which otherwise were unattainable from raw IRS bands through prevailing models. Accuracy and validation measures showed good agreements between the measured and estimated nutrient surfaces. It is concluded that the SBC is a promising method of quantitative soil nutrient mapping, and could further be used for identification and mapping of other indiscernible biophysical parameters.