Agrometeorological Indicators based on Satellite Imagery in Western Bahia, Brazil

Western Bahia stands out for its large-scale agricultural activity, which employs advanced technology and produces high yields of soybean, corn, and cotton. Geographical features, soil correction, and public policies have transformed the region'sagribusiness, making it one of the most productive in the country, and increasing its relevance for agricultural industry and for exportation, making it the largest grain cluster in the Northeast region of Brazil. Soybean is the region'smain crop, and occupies 69% of its agricultural areas. Although crops such as corn and cotton supplement the local productive matrix, soybeans drive the region'seconomy: 50% of fresh soybeans are sold to industries in the region, and 47% of the production is exported to countries like China (66%), Japan (11%) and the Netherlands (8%) (AIBA, 2016).

Western Bahia has well-defined seasons, flat topography and rainfall indices that contribute to the definition of territory limits, as well as an extensive watershed with perennial rivers over the Urucuia aquifer, which enhances the region'spotential for irrigated crops. With a well-defined rainy season, this stretch of large extensions allowed the development of various agricultural activities. Rainfall indexes up to 1,800 mm and other favorable climate and soil conditions contribute to the success of the region'sagribusiness.

The scarcity trend in water resources, a counterpoint to the increasing demand for them, has caused serious conflicts over water use. In recent years, the supervision of the Brazilian National Water Agency (ANA) has been more rigorous, and new plantations in irrigated areas have been interrupted due to water scarcity.

The major benefits of irrigated agriculture practices should be opposite to the huge consumption of water demanded by irrigation systems, which consume about 70% of the waters derived from rivers, lakes and aquifers, and require effective management to prevent environmental impacts and water-use conflicts. These conflicts are aggravated especially in years of severe drought like 2014, when water scarcity impacted human use, energy generation, agriculture, navigation and water transport. In Brazil, the situation becomes more complex due to the strong dependence on water resources for the generation of electricity, and the spatial distribution of irrigated areas clearly shows a concentration trend in regions with strong risk of conflict over water use for energy generation and human consumption.

In the last years, in Western Bahia, irrigated crops have quickly replaced natural vegetation. This land-use change highlights the importance of developing tools to quantify large-scale water productivity parameters, enabling dynamic analyses of mixed agroecosystems (Teixeira et al., 2015). Considering the effects of land-use changes on irrigation perimeters, quantifying biophysical parameters is important for the development and application of tools to evaluate the dynamics of agricultural systems in irrigated areas at Western Bahia, such as in the municipalities of Riachão das Neves and Barreiras.

Evapotranspiration estimation methods generally enable obtaining data on a local scale. However, regions featuring heterogeneous surfaces with different types of soil and vegetation show quite different evaporation rates, which cannot be perceived in traditional estimation methods. Remote sensing enables estimating evapotranspiration overlarge areas, as a function of the biophysical characteristics detected in each pixel. Another major advantage of the use of satellite imagery to estimate evapotranspiration on a regional scale is that the amount of water consumed in the evapotranspiration process may be detected without the need for quantifying other hydrological parameters, such assoil moisture.

Remote sensing using satellites is an efficient tool for estimating water parameters, and provides spatial information, location, and status on different agroecosystems (Teixeira et al., 2010).

Remote sensing techniques make it possible to obtain surface information without making direct measurements. Instead, they are obtained by capturing the converted energy as digital information. Thus, information on crops obtained from satellite images may be coupled to data obtained by meteorological stations, and used for estimating parameters, such as evapotranspiration and biomass, on a large scale. To obtain biophysical parameters, the agrometeorological-spectral model SAFER (Simple Algorithm for Retrieving Evapotranspiration) was applied to MODIS (Moderate Resolution Imaging Spectroradiometer) satellite images (Teixeira et al., 2013).

The SAFER algorithm was developed and validated using data from field experiments and Landsat images depicting natural vegetation and irrigated crops under Brazilian semi-arid conditions.

Monitoring actual evapotranspiration using remote sensing on irrigated crops is an important tool for applications such as agricultural management, water resources monitoring, analysis of water productivity, biomass estimates and agricultural production (Morris et al., 2013).

The model proposed for estimating biomass production (BIO) and the development of plant canopies based on global solar radiation (RG) has acceptable accuracy, and may be used remotely, with any satellite imagery, for different ecosystems (Bastiaanssen, 2003). Although several studies have already been conducted on large scales, research on the use of models for the combination of evapotranspiration (ET) and BIO are still needed, especially for applications on different surfaces under conditions of water scarcity and rational use of water resources.

The purpose of this study was to apply SAFER along with satellite images and meteorological data, made available by INMET (Brazilian National Institute of Meteorology), to quantify evapotranspiration and biomass on a large scale at irrigated areas in Western Bahia, to analyze the dynamics of natural vegetation and irrigated crops throughout the crop cycle.