8.3 CHANGES IN DAILY AND EXTREME TEMPERATURE AND PRECIPITATION INDICES RELATED TO DROUGHTS IN CANADA Eva Mekis and Lucie A. Vincent Climate Research Branch, Meteorological Service of Canada 1. INTRODUCTION Weather conditions related to temperature, precipitation, droughts, storms or other aspects of climate, can cause severe damage and large economic and social losses. Assessing changes through temperature and precipitation ground measurements gives us the opportunity to be prepared and adopt strategies to deal with these events. In particular, severe droughts can present serious threat to our society and environment. Over the Canadian Prairies, prolonged droughts have major impacts on agriculture, industry, recreation and aquatic ecosystems. Meteorological droughts often occur when there is below-normal precipitation for an extended period of time with very high temperature which increases evaporation. However, each drought is different and depends on the region affected, its duration and intensity, and the region’s capacity to adapt to water shortage. Earlier work of Vincent and Mekis (2006) reported on trends in daily and extreme temperature and precipitation indices for 1900-2003 and 1950-2003. Trends were generated and analyzed using homogenized data from 210 temperature and 495 precipitation stations across Canada. Results showed increasing trends in warm events and decreasing trends in cold events. For example, the number of frost days has decreased significantly over both time periods. The analysis of daily precipitation indices revealed more days with precipitation, a decrease in daily precipitation intensity and a decrease in the maximum number of consecutive dry days. No consistent change was found in most of the extreme precipitation indices except for a significant increase in the number of days with heavy precipitation. The present work is also based on homogenized temperature and adjusted precipitation datasets. The study period is extended until 2007 to include the more recent quality controlled climate observations. The objective is to investigate temperature and precipitation indices which are different from the conventional SPI and Palmer Drought Index but can still provide valuable information about droughts in Canada. The trends for five temperature and six precipitation indices are first analyzed over the country. Then the time series are examined for the Canadian Prairies in order to establish if these indices are good indicators of droughts over this region. ____________________________________________ *Corresponding author address: Éva Mekis, Climate Research Division, Environment Canada, 4905 Dufferin Street, Toronto, Ontario, Canada, M3H 5T4; e-mail: Eva.Mekis@ec.gc.ca 2. DATA 2.1 Temperature Indices The temperature indices presented in this study are based on daily maximum for 210 stations across the country. Homogeneity problems caused by station relocation and change in observing procedures were addressed using a technique based on regression model with surrounding stations (Vincent et al. 2002). Table 1 presents a list of the five selected temperature indices along with their definition. The annual and summer means of the daily maximum temperatures present the average condition over the year and the season. The warm temperature extremes are described by the number of hot days (days with daily maximum temperature above 30°C), the number of warm days during the summer (days with daily maximum temperature above the 90 th percentile) and the number of warm spells during the summer (three consecutive days with daily maximum temperature 5° above normal). The summer is defined as June, July and August. More details about these indices are provided in Vincent and Mekis (2006). 2.2 Precipitation indices The precipitation indices provided in this study are computed from adjusted daily rain, snow and total precipitation amounts for 461 stations. The precipitation datasets were recently updated to include more recent station’s observations, but all observations are still based exclusively on manual observations. The methodology follows the steps described in Mekis and Hogg (1999). Adjustments were applied on daily values for rain and snow separately. For each rain gauge type, corrections to account for wind undercatch, evaporation, and gauge specific wetting loss were implemented. The details of rain gauge corrections are further explained in Devine and Mekis (2008). For snowfall, density corrections based upon coincident ruler and Nipher measurements were applied to all ruler measurements (Mekis and Hopkinson, 2004). Adjustments obtained from homogeneity tests using information from surrounding station were applied for the joining of observations from several stations (Vincent and Mekis, 2008). In Canada, when a station is relocated, a new identification number is often given to the new location and the two station’s observations can be combined in order to create a long time series. Great care was as well given to properly account for the trace observation. The list of the precipitation indices included in this study is presented in the second part of Table 1. The annual