Detecting Biophysical Characteristics and Nitrogen Status of Finger Millet at Hyperspectral and Multispectral Resolutions

Finger millet ( Eleusine coracana Gaertn L.) is an important grain crop for small farmers in many countries. Reliable estimates of parameters, such as growth and nitrogen (N) content, through remote sensing techniques can improve in-season management finger millet. This study investigated the relationships hyperspectral reflectance with canopy height, green cover, leaf area index (LAI), N concentrations using optimal waveband selection procedure partial least square regression (PLSR). Predictive performance 13 vegetation indices (VIs) computed from original data well synthesized Landsat-8 Sentinel-2 were evaluated compared estimating various parameters simple linear (SLR) multilinear (MLR) models. The wavebands determined by PLSR mostly concentrated within 1,000–1,100 nm both LAI dry biomass but scattered other parameters. SLR statistics resulted ratio pigment SRPI ) red/green RGI performing best when predicting (R 2 v = 0.53–0.59) cover 0.72–0.76). blue/green BGI 1 was strongly related to height 0.65–0.78), 0.42–0.49), concentration 0.70–0.83) millet, regardless spectral resolutions. MLR approach, four maximum VIs input variables, improved prediction accuracy 14% methods. satellite similar or greater than indicating publicly accessible multispectral could serve alternative decisions via precision agriculture.