Sugarcane irrigation potential in Northwestern São Paulo, Brazil, by integrating Agrometeorological and GIS tools


Water deficit is one of the main limiting factors for sugarcane production around the world. Sugarcane yield is negatively affected by drought, and irrigation can be an alternative to improve yield rates. This study aimed to calculate the sugarcane irrigation requirement (SIR), the available surface water (ASW) for irrigation and create scenarios of the potential of sugarcane irrigation in Northwestern São Paulo, Brazil, by integrating agrometeorological and GIS tools. This region was chosen due to its continuous expansion of sugarcane area and the presence of a significant water deficit during the year. SIR was calculated on a daily scale by implementing the crop water balance, using rainfall and crop evapotranspiration (ETc) of 28 locations for a 33-year time series (1980–2013). ETc was calculated by multiplying reference evapotranspiration, estimated by Penman-Monteith equation, and an average weighted crop coefficient (Kc). The potential sugarcane irrigation was calculated dividing ASW, which was based on the ecological discharge (Q7,10) and multiannual average discharge (Q), by the SIR, considering four percentiles (60, 75, 80 and 90%). A multivariate linear approach was used to create maps of SIR, which varied between municipalities and had weighted average values for the entire region of 562, 750, 932 and 1062 mm year−1, respectively for the following percentiles 60, 75, 80 and 90%. The potential for sugarcane irrigation were higher for Q when compared to Q7,10 for all percentiles. This study pointed out that sugarcane should be irrigated in the Araçatuba region, Northwestern São Paulo, Brazil, nonetheless, in most of the scenarios considered it would only be possible when supplying part of the SIR.

In Agriculture Water Management
Vinicius Perin
Vinicius Perin
PhD student in Geospatial Analytics

My research interests include surface water, agriculture, irrigation, remote sensing and hydrological modeling.