Predicting Pressure from Environmental Factors in London based on ARIMAX Model

Abstract

Pressure is an important index in atmospheric science as it influences various aspects of both the natural environment and everyday life. Although the significance of pressure is widely recognized, accurately measuring and recording pressure data remains a crucial issue. This is due to the fact that weather prediction involves fitting a large number of nonlinear factors, which poses significant challenges in establishing precise predicting models. Therefore, the current focus of most research on pressure lies in the innovation and improvement of algorithms for pressure sensors and other related devices. This paper makes a statistical analysis on the raw climate data of London over a period of 13 years starting from 2008, applying an ARIMAX (Autoregressive Integrated Moving Average with Exogenous Variables) dynamic regression model to fit 4 exogenous variables to the pressure data as linear factors. The fitting results showed that the variation in pressure is associated with multiple environmental factors. Furthermore, it is observed that the pressure in London is projected to be slightly higher than normal one in the upcoming months. Due to the numerous parameters in the ARIMAX model, overfitting and underfitting phenomena are prone to occur, therefore, for parameter adjustments, further processing is required.