Spatio-temporal variability of biophysical parameters of irrigated maize using orbital remote sensing

Abstract

This study proposes to estimate the actual crop evapotranspiration, using the SAFER model, as well as calculate the crop coefficient (Kc) as a function of the normalized difference vegetation index (NDVI) and determine the biomass of an irrigated maize crop using images from the Operational Land Imager (OLI) and Thermal Infrared (TIRS) sensors of the Landsat-8 satellite. Pivots 21 to 26 of a commercial farm located in the municipalities of Bom Jesus da Lapa and Serra do Ramalho, west of Bahia State, Brazil, were selected. Sowing dates for each pivot were arranged as North and South or East and West, with cultivation starting firstly in one of the orientations and subsequently in the other. The relationship between NDVI and the Kc values obtained in the FAO-56 report (KcFAO) revealed a high coefficient of determination (R2= 0.7921), showing that the variance of KcFAOcan be explained by NDVI in the maize crop. Considering the center pivots with different planting dates, the crop evapotranspiration (ETc) pixel values ranged from 0.0 to 6.0 mm d-1during the phenological cycle. The highest values were found at 199 days of the year (DOY), corresponding to around 100 days after sowing (DAS). The lowest BIO values occur at 135 DOY, at around 20 DAS. There is a relationship between ETcand BIO, where the DOY with the highest BIO are equivalent to the days with the highest ETcvalues. In addition to this relationship, BIO is strongly influenced by soil water availability.