Desenvolvimento de sistema meteorológico de previsão para aves através da aplicação do modelo WRF (Weather Research & Forecast Modeling System)

Abstract

The weather forecast is a science that has gained importance in the face of a climate change scenario. Technological evolutions allow to understand and predict events related to environmental physical variables in a precise way, adding a new tool to the productive system of broiler chickens. The weather forecast allows a future view of the system, preventing losses due to the climatic vulnerability of poultry facilities. The objective of the project is to develop a weather forecast system using the WRF (Weather Research & Forecast Modeling System) model, taking into account the Temperature and Humidity Index (ITU) for broiler chickens, aged between 35 to 42 days, through multiple regression analysis and Monte Carlo simulation, for the Maringá (PR), Dourados (MS), Piracicaba (SP), Xanxerê (SC), Joaçaba (SC) and Bento Gonçalves (RS) stations, during January to March 2017. The results demonstrate that the model responds satisfactorily to the prediction of the ITU, and that as the forecast horizon increases, the R² values decrease according to the station. The probability of success decreases from 80% in the first day to 60% on the sixth day, demonstrating that the WRF model can be used to predict ITU as a tool to mitigate environmental impacts in the production of broiler chickens.