Temperature Varations in Southern Saskatchewan, 1921-1970: Regional Identification of Trends
Magill, B. L.
Saskatchewan's vulnerability to variations in the thermal climate could have serious ramifications for agricultural production, energy requirements, economic activities, transportation and recreation. Therefore, a knowledge of temperature change is important. The objectives of this study were to determine the temporal and spatial changes of the thermal regime in Saskatchewan for the fifty year period from 1921-1970. A completeness index was used to measure the reliability of the available temperature records in terms of record length and the number of missing observations. The spatial pattern of a correlation analysis and the completeness index were employed for the evaluation of the network of observing stations. Methods for temporal estimation were applied and the results tested. It was found that monthly temperature data could be estimated satisfactorily but that daily temperature data could not be as adequately estimated. The improved data base was then used to evaluate the temporal and spatial variability of the thermal climate of southern Saskatchewan. The annual, April-October, January, April, July and October mean monthly, mean maximum monthly and mean minimum monthly temperature regimes were examined. An eleven-year binomially weighted running mean was used to analyze the temporal variability of the thermal regime. As the average temperature for each station varied widely throughout the study area, the cumulative sums of the differences between consecutive running means were computed for each station and graphically plotted. These cumulative temperature change curves were then examined for evidence of trend. In order to quantify the established trends, regression lines were calculated on the cumulative temperature change data at each station. The geographical distribution of trend was examined by plotting the slope of the regression line of temperature change at each station. Several intriguing features were disclosed. Statistically significant trends were observed for each of the seasonal, April-October and annual monthly temperature regimes from 1921 to 1970. A predominant cooling trend was evident for January, April and July mean, maximum and minimum temperatures from 1921 to 1970. For the annual and April-October temperature regimes, a cooling trend was evident for the mean and maximum temperatures only. A major cooling trend in annual minimum temperatures was ascribed to approximately half of the stations only. In sharp contrast, a major warming trend dominated the October temperature regimes and the April-October minimum temperatures. Although the dominant trend, whether warming or cooling, generally typified the overall direction of trend for each of the mean, maximum and minimum temperatures, a reversal of trend was exhibited by some stations. It was also concluded that although the study area indicated a general uniformity in the overall trend of temperature, distinct regional variations in the magnitude of trend were evident. The spatial distribution of trend values revealed that stations located within a central corridor aligned northwest to southeast incurred the least rate of temperature change over the period of record analyzed in comparison to stations located in the southwest and northeast.