Climate change and biofuel wheat production in southern Saskatchewan: Long-term climate trends versus climate modeling predictions Sierra Rayne a,∗ , Kaya Forest b a Chemologica Research, 318 Rose Street, PO Box 74, Mortlach, Saskatchewan, Canada, S0H 3E0 b Department of Environmental Engineering, Saskatchewan Institute of Applied Science and Technology, Palliser Campus, 600-6th Avenue NW, PO Box 1420, Moose Jaw, Saskatchewan, Canada, S6H 4R4 Abstract Climate modeling work has suggested biofuel wheat production in southern Saskatchewan, Canada, during the mid- 21 st century will be influenced by increasing annual precipitation, including precipitation increases in every month except July and August, increasing daily mean, minimum, and maximum air temperatures throughout the year, and substantial increases in the risk of wheat heat shock (temperatures>32.0 ◦ C). In the current study, we compare prior modeling predictions to historical trends in the number of days with maximum temperatures >32.0 ◦ C during July and August, the number of hours with maximum temperatures >32.0 ◦ C during July, as well as monthly and annual total precipitation, mean daily temperatures, and mean maximum daily temperatures for climate stations throughout southern Saskatchewan. We find no evidence of increasing trends for wheat heat shock days or hours during the mid-summer period in this region. In contrast, the majority of stations exhibit significantly declining temporal trends in wheat heat shock days and hours. Historical precipitation and temperature trends for the climate stations under consideration in southern Saskatchewan display significant inter- and intra-station heterogeneity throughout the year in terms of whether or not trends are evident, as well as their magnitude and direction. Consequently, caution must be exercised when extrapolating any case study analyses at a particular location to larger geographic areas of the province. Based on our analyses of historical climate data for southern Saskatchewan, it is unclear whether climate models are accurately predicting future climate change impacts on biofuel wheat production for this region in the mid-21 st century. Keywords: Climate change, Biofuels, Wheat, Southern Saskatchewan Approximately 15% of fermentation derived ethanol in Canada is from wheat, with various proposals currently under consideration to substantially increase production rates - particularly in the two western prairie provinces of Saskatchewan and Manitoba [1]. In a recent study, Wang et al. [2] assessed the potential impacts of climate change on biofuel wheat production in southern Saskatchewan. Using Swift Current in the southwestern region of the province as a case study site, the climate modeling ef- forts of these authors concluded that, by the 2050s and relative to a 1961-1990 baseline period, annual precipita- tion is expected to increase, precipitation increases are ex- pected in every month except July and August (two of the three climate scenarios predicted increasing precipita- tion in June, with the other scenario predicting a decrease in this month by the 2050s), daily mean, minimum, and maximum air temperatures are expected to also increase throughout the year (with the greatest increases by season in the following order: winter>summer>spring>fall), and wheat heat shock (temperatures>32.0 ◦ C; mostly occur- ∗ Corresponding author. Tel.: +1 306 690 0573. E-mail address: sierra.rayne@live.co.uk (S. Rayne). ring in July) was expected to increase substantially under all three climate scenarios. In the current study, we compare these climate mod- eling predictions by Wang et al. [2] to trends observed over the available climate records throughout southern Saskatchewan (defined as south of approximately Prince Albert [latitude 53.2-53.3 ◦ N] [3]; see Table 1 and Fig- ure 1 for details and locations of the climate monitoring stations under consideration), which in some cases date back to the 1880s. Monthly and annual means of daily mean temperature and daily maximum temperature, as well as monthly and annual total precipitation, were taken from the second generation homogenized temperature and the second generation adjusted precipitation datasets of the Adjusted and Homogenized Canadian Climate Data archive [4, 5]. Trends were investigated using paramet- ric linear regression, non-parametric Mann-Kendall time trend analysis with Sen’s slope estimation for the slope [6–8], non-parametric Spearman rank correlation [9], and non-parametric Kendall rank correlation [7] with KyPlot (v.2.0.b.15) and the R statistical [10] software packages. Hourly and daily climate information was obtained from the online National Climate Data and Information Archive Preprint submitted to viXra March 7, 2013