Implementação computacional de interpolação espacial de dados climáticos nas Ilhas Britânicas: uma abordagem interativa com Shiny

Abstract

This study focuses on the development of an interactive Shiny app and the spatial interpolation of temperature data in the British Isles. The main goal is to offer users a tool that allows them to explore a variety of model and error combinations, as well as to understand how spatial interpolation contributes to more insightful representations.

In the Shiny Application Development section, an interactive application was built using the R language and the Shiny package. This app allows users to choose different models and error metrics to visualize the resulting predictions in an accessible and interactive format. The process involved reading and preparing the data, followed by creating an intuitive user interface and interactive elements, such as drop-down menus and zoom, pan and tooltip features in the visualizations.

The Spatial Interpolation section plays a crucial role in this study. It involves preparing the grid data, fitting the variogram model to estimate spatial correlations and applying the ordinary kriging technique to estimate temperature values at locations where observed data were missing. Interpolated and continuous maps of the temperature predictions were generated and visualized using the ggplot2 library1.

Therefore, this work combines the development of an interactive Shiny app with advanced spatial interpolation techniques to offer a powerful tool in the analysis and visualization of climate data in the British Isles. The interactive features facilitate the exploration of complex information, while the spatial interpolation provides accurate representations of the model and metric combinations across the region.