Previously I had a quick look at the coastal weather stations for a first idea of likely weather conditions we might encounter in Greenland, but at our last training we were discussing actual windspeeds, incidence of Piteraqs, and likelihood of severely cold conditions, and possibility of melt.
As a background, here are some thoughts from folks more familiar with Greenland than me …
Ruth Mottram (DMI scientist): It varies *a lot* in April, it probably has the biggest variability of any month in Greenland – this year (2016) we had record warm temperatures (and rain!) and then it got rather cold again before again record warm temperatures in May, but the year before was very cold with a record maximum sea ice extent in west Greenland and more snow than they could deal with in the East (which is kind of saying something when it comes to Greenland) so at this stage who knows!
Ahem, so based on that it seems that basically anything could happen. The good news is that (i) the DMI Piteraq forecast is quite reliable and should be able to help us prepare in the event of a storm, (ii) the DMI seasonal forecast is also considered a helpful resource now. We shall be using both this spring!
Anyway, to begin with I had another look at freely available data from a couple of PROMICE (Program for Monitoring of the Greenland Icesheet) weather stations near Tasiilaq (East, our starting point) and Kangerluusuaq (West, towards our destination), to look for answers to some questions. The weather stations I chose are shown in the table below. I hope the east stations should give us a good idea of the potential for wind storms and temperatures at the start, while the uppermost westerly station is quite and hopefully shows conditions somewhat characteristic for the inland ice, and the lower west station should indicate conditions as we are exiting the ice. I looked at data for April and May only, from 2010-2017 (8 years). As there is sometimes data missing from the hourly records, in the table below I calculated occurrences of extreme conditions over one or both of the months as a % of the available hourly data for these months over the whole 8 years.
Assuming we will be spending most of our time in conditions most similar to the station at 1840m (though we will go higher than this on the traverse as well), over these 8 years of Aprils and Mays the data suggest that average temperatures are -14°C, discarding the upper and lower 10% of the temperatures – i.e. excluding these extremes – the remaining 80% of the time, air temperatures lie between -25 and -4°C. I guess these might represent our typical conditions? And the winds are likely to be around 20km/hr a third of the time, and about three quarters of the time between 10-30km/hr. Below is a pair of probability histograms of temperature(left) and wind(right) conditions at the KAN_U weather station.
I also wondered if there is anything interesting in seeing how these conditions evolve over real time for the last 8 years of April/May, so I plotted windchill and windspeed at all 4 stations in a different colour for each year going from pale yellow through to the purple and black tones. Note that the higher eastern station did not have valid data for the first years, and was retired in 2015, and that the lower station did not have valid temperature data for the final two years. I suspect these operational failures are, in themselves, a sign the conditions might be at least on occasion more savage than those recorded in the data.
What we can see from this is that the temperature and windchill conditions really improve substantially into May, and that while we might have melting conditions towards the end of our traverse the inland ice will rarely have windchill conditions above 0°C. That said when surrounded by snow and its a sunny day its a bit like being in a solar cooker. Most years experience at least one dangerous wind event in April on the eastern side of the icesheet, but they are typically not protracted. Some years however have periods of repeat storms, 2013 and 2014 being notably stormier in the dataset available here.