Espacialização da evapotranspiração de referência diária no Espírito Santo

Abstract

The aim of this study was to compare interpolation methods aiming at spacialization of the evapotranspiration of reference (ETo) estimated by the PenmanMonteith method, in the state of Espírito Santo. The objective was also to identify the spatial variation of ETo follows the same behavior in periods with low radiation and high radiation The study area is located between latitudes 17° 52' and 21° 19' S and 39° 38' and 41° 50' longitudes O, in the Southeast region of Brazil, with a total area of 46,184.1 km ². 15 automatic weather stations of the National Institute of Meteorology (INMET) were selected as the basis for interpolation and ten to cross-validation, with eight stations belonging to the company Fibria and two stations to Weather Forecasting and Climate Studies center (CPTEC/INPE). The period analyzed in this study was a daily time series from 2006 to 2012 (Sep/12), prioritizing the use of data of the period between 2010-2012. Three interpolation methods were used: Kriging geostatistical method and the deterministic methods Inverse Square the Distance (IQD) and Spline. Three groups of data were interpolated, the first group with 75% of the stations with solar radiation of Ra ≤ 40% (solar radiation at the top of the atmosphere), being characterized as "Days with relative low radiation"; the second group with 75% of stations with precipitation <10 mm, characterized as "Days of Radiation on High"; and the third group, called "All Days" with interpolation every day without common to all stations flaws, totaling 78 days. The IQD interpolation performed better on the spatial distribution of ETo variable for the three conditions, thus representing a smaller deviation and variation between the ETo values estimated by the Penman-Monteith equation and ETo estimated by IQD interpolation method. The IQD method was the one that best adjusted for interpolation of ETo in the form of maps with the total average days used in each group, with this exhibiting less variation of the difference of percentage between the interpolations with all stations (INMET, Fibria, CPTEC/INPE) and only with the INMET stations. The IQD interpolation is therefore shown as a good method for estimating daily reference evapotranspiration variable in places where you do not have installed weather xi stations. The spatial variation of ETo follows the same behavior uniform distribution of interpolation periods with low radiation and high radiation on the surface. The estimated values of ETo obtained by IQD interpolation method can be used with confidence in irrigation management.