Desenvolvimento de um sistema web para estimativa numérica de dados meteorológicos do Rio Grande do Sul

Abstract

Climate influences a large variety of human activities and the real-time access to climatic data aims for providing information that are fundamental in several human activities, mainly for agriculture. Nowadays there are few weather stations operating, which causes a lack of information at worldwide level, in different regions. Keeping in mind that the state of Rio Grande do Sul has a large part of its economy based on agriculture, as well as the climatological information's relevance, this study aims at developing a web system for numerical estimate of meteorological data in the RS's state, based on the automatical and superficial weather stations from the National Institute of Meteorology (INMET), in order to make the data available to the regions where there are no stations. The inverse distance weighting was used as an interpolation model, applying the exponents from zero (0) to five (5) and, later, comparing them through the coefficients of correlation, regression and the performance index. It was made use of programming languages as PHP, HTML and JavaScript to develop the web system, with support for MySQL database. The programs used were Macromedia Dreamweaver 8.0 and HeidSQL; the first was used for web programming whereas the second was used to manage the database. Among the nine variables that was analysed, just four of them showed a great performance. They are: temperature, relative humidity, atmospheric pressure and dew points. The interpolation model with the exponent five (5) has shown the best performance regarding to the four variables. After defining the best method, it was created a MySQL database called SWIM (Meteorological Interpolation's Web System) and through it the web system was developed, which has offered quickness, security and reliability to the application.