Utah is a major tourist destination for skiers, boasting not only ample snowfall and a number of world class resorts, but claiming to be “the greatest snow on earth.” Due to its nature as the Eastern edge of the Great Basin, the mountains of Utah often record hundreds of inches of snowfall annually. I was curious to find out just where all the greatest snow was falling and how much snow was accumulating.
There is no shortage of ski resorts across the state of Utah, however just being in the state does not necessarily guarantee sufficient snowfall, I discovered that outside of the mountain ranges in Utah, most recording stations receive a small fraction of the amounts recorded in the mountains.
Overall snowfall totals have been decreasing in the last 20 years, with global warming often being attributed to this change. Increasingly warm temperatures are shortening the length of the winter season from Mid-October to mid-April, to early November to beginning of April. There are a number of activist groups working to address this change, most notably Protect Our Winters (POW) which has been working to organize and mobilize the influence of skiers and snowboarders. While outside the scope of this project, an interesting analysis of the effects of global warming would be comparing snowfall totals from January and February against December and March to determine the consistency of rising temperatures.
My first step in this analysis was to find and then subsequently clean the data. As I was working with weather data I was able to obtain it from NOAA’s National Centers for Environmental Information. The data download was relatively straightforward as the data came in a CSV and you were able to filter by state and month.
The next step of cleaning the data was done using openrefine and while this part was not difficult, there were a number of decisions that needed to be made with regard to the data. As most of Utah is considered high desert, there is often not much, if any, precipitation recorded at weather stations thus resulting in a substantial number of 0″ recorded as well as more than a few M, standing for missing data. Additionally, T was used for trace amounts of precipitation. After I had transformed the data from the short format to a long format I needed to determine whether the zero, missing, and trace values should be included. I did a brief visualization and keeping these values included was overwhelming and did not help to illustrate snowfall in Utah. After removing the unnecessary data points, snowfall amounts and concentrations became much more easily viewable.
Results & Reflection
Winter storms come in waves, also known as storm cycles and it is during these periods that substantial amount of snow will fall. There is no standard time span for the storm cycles thus some are longer than others, but generally they are at least 2 days of snowfall.
Overall snowfall was concentrated in the Northern half of the state with most of the snowfall being recorded in the area around Salt Lake City. Areas in the south and at lower elevations received minimal, if any, snowfall over the time frame analyzed. This brings up a major flaw with weather data, as precipitation measurements come from weather stations and weather stations are located only where people have settled there is likely more locations with snowfall, particularly within National Forests which are sparsely populated.
Additionally the format of Tableau was somewhat limiting, with the small view of the dashboard particularly limiting and I was in particular frustrated by the inability for the dashboard to expand downwards, as it would have been much easier to format and show more detail with the different visualizations.