Publications

Mahajan, G. R.; Sahoo, R. N.; Pandey, R. N.; Gupta, V. K.; Kumar, Dinesh (2014). Using hyperspectral remote sensing techniques to monitor nitrogen, phosphorus, sulphur and potassium in wheat (Triticum aestivum L.). PRECISION AGRICULTURE, 15(5), 499-522.

Abstract
In situ, non-destructive and real time mineral nutrient stress monitoring is an important aspect of precision farming for rational use of fertilizers. Studies have demonstrated the ability of remote sensing to monitor nitrogen (N) in many crops, phosphorus (P) and potassium (K) in very few crops and none so far to monitor sulphur (S). Specially designed (1) fertility gradient experiment and (2) test crop experiments were used to check the possibility of mineral N-P-S-K stress detection using airborne hyperspectral remote sensing. Leaf and canopy hyperspectral reflectance data and nutrient status at booting stage of the wheat crop were recorded. N-P-S-K sensitive wavelengths were identified using linear correlation analysis. Eight traditional vegetation indices (VIs) and three proposed (one for P and two for S) were evaluated for plant N-P-S-K predictability. A proposed VI (P_1080_1460) predicted P content with high and significant accuracy (correlation coefficient (r) 0.42 and root means square error (RMSE) 0.180 g m(-2)). Performance of the proposed S VI (S_660_1080) for S concentration and content retrieval was similar whereas prediction accuracies were higher than traditional VIs. Prediction accuracy of linear regressive models improved when biomass-based nutrient contents were considered rather than concentrations. Reflectance in the SWIR region was found to monitor N-P-S-K status in plants in combination with reflectance at either visible (VIS) or near infrared (NIR) region. Newly developed and validated spectral algorithms specific to N, P, S and K can further be used for monitoring in a wheat crop in order to undertake site-specific management.

DOI:
10.1007/s11119-014-9348-7

ISSN:
1385-2256