Change point analysis of precipitation indices of Western Canada

Abstract

The Markov chain statistical technique was employed to fill few missing values of daily precipitation observed at nine long-term (≥50 years of data) climate stations of Western Canada. Statistical properties of the gap filled precipitation data are compared with those of the original data to ensure that this approach preserves the statistical characteristics of the historical data. After gap filling missing data, 10 precipitation indices that represent magnitude and frequency of precipitation properties were computed from those 9 stations and the other 20 stations without missing records. Next, the Wilcoxon rank-sum test was used to detect possible change points of these precipitation indices. Based on the analysed results, there had been an increasing trend in the maximum number of consecutive wet days and the annual maximum 1-day precipitation and a decrease in the maximum number of consecutive dry days over most of British Columbia (BC). The annual total precipitation index of many stations had increased in southern BC and decreased in central parts of Canadian Prairies. Similar change was detected in the number of days when precipitation is equal and greater than 10 or 20 mm, and indices for the annual monthly maximum 1-day precipitation – precipitation that exceeds the 95 and 99 percentiles – and the simple precipitation intensity index. On the basis of the above results, the diurnal temperature range (DTR) of Western Canada has decreased, and predominant negative correlations between precipitation indices and DTR, it seems that Western Canada has generally become wetter since the middle of the 20th century partly because increase in the minimum daily temperature has been higher than the increase in the maximum daily temperature, leading to a decrease in DTR, which is likely related to climatic warming.