Remote sensing for irrigation water management in the semi-arid Northeast of Brazil

Northeast of Brazil is a semi-arid region, where water is a key strategic resource in the development of all sectors of the economy. Irrigation agriculture is the main water consumer in this region. Therefore, policy directives are calling for tools to aid operational monitoring in planning, control and charging of irrigation water. Using Landsat imagery, this study evaluates the utility of a process that measures the level of water use in an irrigated area of the state of Ceara. The experiment, which models evapotranspiration (ET), was carried out within the Jaguaribe-Apodi irrigation scheme (DIJA) during two months of the agricultural season. The ET was estimated with the model Mapping Evapotranspiration at High Resolution and with Internalized Calibration (METRIC). The model uses the residual of the energy balance equation to estimate ET for each pixel in the image. The results of the estimates were validated using measurements of ET from a micrometeorological tower installed within a banana plantation located near the irrigation scheme. After evaluating the ET estimates, the average fraction of depleted water for a set of agricultural parcels combined with the monthly ET mapping estimates by METRIC provided a method for predicting the total water use in DIJA for the study period. The results were then compared against the monthly accumulated flow rates for all the pumping stations provided by the district manager. Finally, this work discusses the potential use of the model as an alternative method to calculate water consumption in irrigated agriculture and the implications for water resource management in irrigated perimeters. (C) 2009 Elsevier B.V. All rights reserved.