The maximum daily evapotranspiration data set for a project (ETproject) for Brazil. It has a spatial resolution of 30 seconds (~ 1 km²). The data set grid is in GeoTIFF format and corresponds perfectly to WorldClim. It uses the geographic coordinate reference system, with WGS84 projection (EPSG: 4326). The objective study is to estimate and provide evapotranspiration values of monthly reference and the maximum of twelve months, for dimensioning irrigation systems throughout the Brazilian territory. With the meteorological data of two hundred and fifty-nine conventional INMET stations, the daily reference evapotranspiration (ETo) for 15 years was calculated. For each weather station, the data was grouped by month and the ETo for the irrigation project (ETproject) was determined to meet the eighty percent probability of occurrence, following the recommendations of FAO24. In parallel, monthly images of 15 years of ETo were acquired for Brazil, and the climatic variables of WorldClim. Using the ET values of the stations design, it was modeled for the rest of Brazil, using machine learning algorithms and the covariates. After modeling, the following performances were achieved: mean square error of 0.306 mm / d, mean bias error of -0.004 mm / d, mean absolute error of 0.227 mm / d, determination coefficient of 0.938 and efficiency of Nash-Sutcliffe 0.937. ETo values for irrigation projects were similar to several others reported in the literature when compared at a given point. With this research it was possible to determine the monthly and annual ETo for irrigation projects throughout the Brazilian territory. The article has been submitted for publication.

{"references": ["Doorenbos, J., Pruitt, W.O., 1977. Guidelines for predicting crop water requirements, FAO-24, Rev. ed, FAO irrigation and drainage paper; 24. Food and Agriculture Organization of the United Nations, Rome.", "Fick, S.E., Hijmans, R.J., 2017. WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas: NEW CLIMATE SURFACES FOR GLOBAL LAND AREAS. International Journal of Climatology 37, 4302\u20134315. https://doi.org/10.1002/joc.5086", "R Core Team, 2020. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria."]}

We are grateful for the financial support (master's scholarship) granted by the National Council for Scientific and Technological Development - Brazil (CNPq) under the contract number: 131565 / 2016-8 and Coordination for the Improvement of Higher Education Personnel - Brazil (CAPES) Financial Code 001. We also thank the Department of Agricultural Engineering (DEA), the Study and Solutions Group for Irrigated Agriculture (GESAI) at the Federal University of Viçosa (UFV).