Remotely sensed vegetation indices for crop nutrition mapping.

Abstract

Spectral reflectance measurements and vegetation indices have been used widely to monitor plant nutritional status. Remote sensing methods are very fast, convenient, and non-destructive techniques for estimating crop nutrition indices. With the rapid development of remote sensing satellites, unmanned aerial vehicles, and the Internet of Things, remote sensing will increasingly be used in crop nutrition diagnosis. This can lead to lower operating costs for fertilization and can minimize loss of productivity. In this study, satellite images were used for 2 years to evaluate the use of vegetation indices for crop nutrition mapping, in two regions, located in the cities of Ray and Karaj. For each farm, eight reference points in each year were selected to determine the crop nutrition status for model development and accuracy assessment. Then, ten spectral vegetation indices were used to calculate the crop nutrition content at mid-season and at the end of the season. The results indicated that transformed chlorophyll absorption in reflectance index (TCARI) and modified chlorophyll absorption in reflectance index (MCARI) had the highest R2 and the lowest RMSE values, respectively (R2 TCARI = 0.83, RMSETCARI = 12.47 kg ha-1 and R2 MCARI = 0.75 and RMSEMCARI = 14.47 kg ha-1 ). The performance of TCARI and MCARI was strong across the two farms and allowed the creation of highly accurate crop nutrition maps. In conclusion, the use of the near infra-red and red-edge bands in mid-season led to better results than vegetation indices calculated at the end of the season or that do not included the red-edge band. © 2020 Society of Chemical Industry.