Retrieval of Paddy Crop Nutrient Content at Plot Scale Using Optimal Synthetic Bands of High Spectral and Spatial Resolution Satellite Imagery

Abstract

The main aim of the present study is to introduce Hyperion narrow band information into high spatial broad band imagery so that it will be helpful for the estimation of crop parameters at small field scale from multispectral data. Here, we are targeting to retrieve the nitrogen content of summer paddy at plot scale. Our findings concluded that paddy crop heterogeneity can be assessed by implementing narrow band index regression model designed for paddy crop from multispectral LISS IV data. Moreover, the proposed models to construct the synthetic narrow bands in green (GS), red (RS) and NIR (NS) regions of the spectrum for high-resolution satellite imagery (LISS IV) are also presented here {\(GS\rho_{LISS \, IV} = 0.032 \cdot G\rho_{IH} + 17.296\); \(RS\rho_{LISS \, IV} = 0.072 \cdot R\rho_{IH} + 16.025\); \(NS\rho_{LISS \, IV} = 0.087 \cdot N\rho_{IH} + 24.923\)}. It showed the spatial variability of nitrogen content ranged from 3.37–6.04% obtained by leaf nitrogen concentration (LNC: R705, R717, R491) index regression model and 3.82–5.96% by simple ratio (SR: R533, R565) index regression model over a rice agriculture system. Thus, the details of spatial variability at crop field level over a rice agriculture system can be assessed from multispectral LISS IV data with 5.6 m spatial resolution through the synthetic broad bands. The results conclude that narrow band-based index regression model derived from Hyperion bands by implementing band average method can effectively evaluate the nitrogen content distribution of croplands subjected to varying field heterogeneity. It confirmed that the proposed methodology can be applied for the variability mapping of crop parameters at plot-scale level within the fields from space platform, which can be beneficial in achieving operational precision farming more efficiently.