I always look forward to the Olympics and the chance to spend a couple weeks watching sports that, generally speaking, you don’t get the chance to watch all that often. Obviously in the Winter Olympics, many of the events can be significantly impacted by weather conditions as they take place on ice and snow. With the XXIII Olympic Winter Games in Pyeongchang, South Korea quickly approaching, I decided to take a look at what the weather was like during past Winter Olympics.
I examined the Global Historical Climate Network (GHCN) Daily Summaries dating back to the first Winter Games in 1924 in Chamonix, France, looking for historical daily weather data records that were in reasonable proximity to the host city, both in distance and elevation. In many cases, there was not a weather observation site with a long period of record in immediate proximity to the location of the Opening Ceremony. However, I generally considered a site within 100km (62mi) distance and 500m (1640ft) elevation to be a reasonably representative approximation for comparison purposes. Please keep that in mind; in other words, there are undoubtedly some error bars around the values I present here. However, some trends are strong enough to make some basic conclusions. What you see below is thus a snapshot of the average weather during each Olympics, rather than a climatological normal for the host cities.
For three of the Winter Olympics (St. Moritz ’28, St. Moriz ’48, and Cortina ’56), no observing site existed that was reasonably close in elevation. All of the closest observing sites within 100km of the host city were over 1000m different in elevation. Therefore, the above table is not complete, but the vast majority of the Winter Olympics have a representative observation with which to compare. It’s also worth noting that the closest observation to Turin (2006) was Milan, which is just over 100km away. However, the climatological normals for both cities in February are very close, and they sit at a similar elevation in the same broad valley in northern Italy, so Milan was considered a close approximation. Although a climate site does exist in Turin, the data was unavailable for this time. The only other instance of incomplete data was Nagano (1998), which had some missing minimum temperature data. Other nearby sites in Japan also had some missing minimum temperature data, so an average was just computed from what was available. Put an asterisk next to those two sites if you want.
The other note about the data is that a GHCN site with representative data was not found for Sochi. However, I decided to replace that with archived METAR observations from nearby Adler (URSS), which is actually closer to the Sochi Olympic Park than the Sochi observation itself. This data was retrieved from Weather Underground and seemed representative of the conditions in the area.