Undergraduate Program in Geography, Geographic Information Systems II, York University, Toronto, Ontario, Canada By: Francis Lavigne-Theriault
INTRODUCTION
The purpose of this study is to investigate the proximity of tornadoes to population centres across Canada and to provide insight into whether the frequency of tornadoes are higher near cities, due to the increased likelihood of detection. To do so, the focus will be narrowed down on the provincial scale to investigate the proximity of verified tornadic events across the Province of Manitoba and population centres. This study will hopefully provide insight in the ongoing concern on whether tornado frequency is increasing or simply population density is increasing, therefore leading to a higher frequency of tornado reports.
METHODS
Data acquisition
Statistics Canada 2016 boundary files were downloaded (publicly available at the website (https://www12.statcan.gc.ca/census-recensement/2011/geo/bound-limit/bound-limit-2016-e ng.cfm) and inputted into ArcMap. The boundary files downloaded were provinces/territories and population centres in the shapefile format. The national verified tornado events dataset was downloaded from Environment and Climate Change Canada (publicly available on their website (http://data.ec.gc.ca/data/weather/products/canadian-national-tornado-database-verified-eve nts-1980-2009-public/?lang=en) and imported into ArcMap as a shapefile. The dataset has 1840 tornado events as point data across Canada between 1980 and 2009.
Data manipulation
Figure 1 depicts the combination of population centres, provinces and tornado event shapefiles. As can be observed, it is very difficult to discern the population centres and this is the main reason for picking an individual province as our study area rather than the whole country. In the shapefile for Canadian provinces/territories, the Province of Manitoba was selected and exported into a separate shapefile for analysis using the select by attributes tool. The Canadian population shapefile and tornado events shapefile were then clipped to the newly created Manitoba shapefile using the clip tool in ArcMap to display tornadoes and population solely for the Province of Manitoba. This yielded 284 tornadoes for the province between 1980-2009.
Figure 2 depicts the result of clipping the tornado events and population centres to the Manitoba shapefile previously created. However, the population centres are still barely visible. Since we wanted to focus on population centres, the study area was narrowed down even further to southern Manitoba. Therefore, our resulting study area is shown in Figure 3 and depicts noticeable population centres such as Brandon, Portage La Prairie and Winnipeg.
Buffers
We are interested in discovering the proximity of tornadoes from major cities in southern Manitoba and more specifically Portage La Prairie, Winnipeg, Brandon and Morden/Winkler. The icon for verified tornado events in Figure 3 was modified when compared to Figure 1 and 2 for display purposes. The buffer tool was used on the resulting southern Manitoba shapefile (Figure 3) to determine proximity between the verified tornadoes and population centres. A 5 kilometre buffer was created around each population centre across southern Manitoba using the population shapefile. That buffer was then used to clip to the Manitoba tornadoes shapefile, to get only the tornadoes within a 5 km radius of a city or inside a city itself.
RESULTS AND DISCUSSION
Buffer results
The buffer resulted in 64 tornadoes being in close proximity (5 km) or within a city across southern Manitoba between 1980 and 2009 (Figure 4). The detailed results are shown in table 1, which shows the different characteristics of the retained 64 tornadoes, including their Fujita ratings (the F-scale ranges from 0 to 5, 5 being the strongest), the latitude/longitude coordinates of each tornado and their nearest cities. It is interesting to note that the strongest tornadoes retained are only F1 tornadoes. Therefore, no violent F3+ were retained by the 5 km buffer. This is a bit strange due to the large dataset of 283 tornadoes across Manitoba between 1980-2009. Also, Canada’s strongest tornado and Canada’s only F5 tornado occurred in Elie, MB, but it was not accounted for by the buffer. It is possible that our restraints to just southern Manitoba provided some limitations or our buffer was not large enough. It is also possible that the 2016 census shapefile did not account for small communities as well.
The highest frequency of tornadoes is therefore observed near the City of Winnipeg with 15 tornadoes, Morden/Winkler with 8 tornadoes, Brandon with 5 tornadoes and lastly Portage La Prairie with 3 tornadoes. The remaining 33 tornadoes are spread across smaller cities across the province. Based on these simple observations, one could stipulate that the larger the city, the larger the frequency of verified tornadoes. However, this is nowhere near conclusive evidence considering only 64 out of the 284 Manitoba tornadoes between 1980-2009 were within 5 km of a city or within a city, this means that only 22.54% of verified tornado events occurred within such parameters. This number is much lower than what was expected.
CONCLUSIONS
Tornadoes are fascinating, mysterious and are a deadly phenomenon of nature. With the changing climate and our growing population, it brings forth a simple question: is the tornado frequency increasing across Canada or are we simply reporting them more efficiently? To provide a potential answer to this question, we employed data from Environment and Climate Change Canada and Statistics Canada to conduct a simple GIS analysis. By clipping our large dataset into a more usable province-wide format, rather than nationwide, allowed us to focus on the 284 verified tornadoes across Manitoba between 1980-2009. Since most of Manitoba’s population is in southern Manitoba, we looked into major population centres and their proximity to tornadic events by employing a simple 5 km buffer on each population centre. This yielded 64 tornadoes which were within 5 km of a population centre or within the city limits. Of those 64 tornadoes, the highest rated tornado was F1 and 15 of them (the highest frequency) were reported in or near the City of Winnipeg. Further research will be needed to answer whether the frequency of tornadoes have increased due to a changing climate or due to an increasing population. Future research should compare every province and the population density to see if there is a correlation between the two variables. We would also recommend future studies to use a larger dataset. For example, the United States dataset for tornadoes goes back to 1950 and is updated up to 2020. Since our population has been increasing exponentially in the last 30 years, a more recent dataset could provide better conclusions.
References Environment and Climate Change Canada. (2018). Canadian National Tornado Database: Verified Events (1980-2009) - Public [Data set]. ECCC http://data.ec.gc.ca/data/weather/products/canadian-national-tornado-database-verified-even ts-1980-2009-public/?lang=en Statistics Canada. (2019). 2016 Census - Boundary Files [Data set]. Statistics Canada https://www12.statcan.gc.ca/census-recensement/2011/geo/bound-limit/bound-limit-2016-en g.cfm