Dissertation title: Analysis of techniques for estimating and forecasting evapotranspiration with
free software r
Artificial neural networks; R language; Penman-Monteith Equation
Water management is of great importance for the development sustainability, being even debated by the United Nations and included in the sustainable development goals. Monitoring water consumption can be accurately done using evapotranspiration estimation techniques, as well as advanced methods such as the Artificial Neural Network (ANN) that use climate data. This work aimed to use the language of R programming to implement methodologies and methods in computation, estimation and reference evapotranspiration (ETo) forecast. For the development of this work The R programming language was used as the main tool for: checking and data organization; implementation of methods and presentation of results. The data had as source the automated meteorological station called A315, located in Barbalha, Ceará. This work produced free software, available to the entire community academic, able to simplify the workflow by automating data preparation meteorological data and calculate evapotranspiration and estimate through meteorological data combined with neural networks, obtaining ANN models with R2 equal to 0.9962 using Atmospheric Pressure, Solar Radiation, Maximum Temperature, Minimum Temperature, Humidity Relative and Wind Speed as input data.