Group Comparison, Trends and Cluster Analysis to Understand Historical Precipitation

Abstract

Several studies alert about increasing temperatures in different sites of the world. Most of the studies are based on climate models, but the use of historical climate data also presents similar results. These increasing temperatures are not the only climate element suffering change. Shifting in precipitation patterns (frequency and intensity) is another important change that needs to be better understood, as it may result in different environmental and social impacts. Understanding how climate is changing, as well as the potential impacts of these climate changes, is crucial for the determination of adaptation policies towards mitigation of the impacts of climate change. Relatively long historical climate datasets are available for many cities and sites, but little use of this information is made towards a better understanding of climate change impacts. The high cost of installation and maintenance of the climate monitoring stations, throughout decades, could be better availed and deeper information about the dynamics of regional climate change could be obtained from the datasets. For this study, three different approaches were followed to understand local precipitation from historical data: (1) a general assessment of annual totals with comparison of past and recent periods; (2) trend analysis; and (3) the application of cluster analysis to analyze monthly and seasonal changes. Fifty-five years of monthly rainfall and snowfall data were obtained from Environment Canada (Canadian National Climate Data and Information Archive) for the municipality of Sudbury (Ontario, Canada). The dataset suggests that annual rainfall is increasing in Sudbury while annual snowfall is not changing significantly. Trend analysis detected annual, seasonal and monthly changes in rainfall, highlighting that the annual increasing trends observed for rainfall volumes were not uniform throughout the year. No significant trends were detected for snowfall. However, cluster analysis detected changes in snowfall distribution for December, January and February, the 3 months of higher snowfall values in the site. The clusters also evidenced that, despite the annual increases in rainfall, decreases could be observed in some months, especially in summer, the rainier season in Sudbury. This study did not intend, by any means, to prove that the methodology herein presented is better than the overspread use of climate models. The results highlighted the importance of also taking into account these different approaches, and combinations of the three methods should be considered to provide a more detailed perspective of monthly and seasonal changes that could be hidden in the usual high variability of some climate elements, such as precipitation.