|
Abstract:
|
The purpose of our presentation is to investigate the accuracy of weather forecasts. Particularly, maximum and minimum temperatures and rainfall are considered. We examine the distribution of the errors of these forecasts and how the errors change with the closeness of the forecasts. A regression model is also used to examine other effects on forecast accuracy, such as levels of temperature and rainfall and the previous day's weather. We use an interactive map to visualize our findings, displaying the errors in temperature and precipitation forecasting by month. Interpolation is used to improve the map's ability to generalize to other cities not included in the data. Our project was mostly coded with R, utilizing packages like Shiny, Plotly and Ggplot.
|
