Estimation of Rice Plant Nitrogen Content from Meris Data Using a Field Based Regressive Model

This study is part of ongoing research aimed to exploit Earth Observations (EO) to force physiological models in order to produce better estimation of rice production. The research wanted to analysed the feasibility of estimating Plant Nitrogen Content (PNC) from remotely sensed data. We developed a Normalised Vegetation Index (NDI) able to predict PNC in rice crops through a regressive model that was calibrated (2004 campaign) and validated (2006 campaign) with data of two field experiments carried out in Northern Italy. NDI exploits the availability of hyperspectral data in the visible (blue/green) region of the electromagnetic spectrum where nitrogen/chlorophyll compounds play a key role in radiation absorption. A method to scale up the model using MERIS data has been applied to produce regional maps of PNC. Estimations for 2004 and 2006 showed the expected behavior of a progressive dilution of PNC. The first attempt to use these results in WARM rice crop model gave promising results.