On the study of a major snowstorm in a global warming context

Winter storms, especially those happening when the near-surface temperature is about 0°C, can lead to power outages, ground and air transport disruption and sometimes serious injuries. In this regard, it is crucial to better understand the processes implied in those events. The impact of global warming is not negligible for many physical processes since the temperature will probably be 2°C to 10°C higher in about 100 years. One could imagine a snow storm that happened around -10°C to -5°C to become a near 0°C storm with occurrences of freezing precipitation. For those reasons, investigating the effect of a warmer climate on winter storms can lead to a better understanding of which processes might be responsible for changes in the precipitation types and quantity.

In this study, we look at a snowstorm that hit the eastern part of Canada on December 27 2012 during which about 50 cm of accumulated snow was reported in Montreal (some places in this area got almost 1 m of snow). Liu et al. (2016) performed continental-scale convection-permitting climate simulations (Δx = 4 km) over 13 years with the WRF (weather research and forecasting) model using a pseudo-global warming (PGW) approach. This method consists in adding perturbations to the re-analysis data driving the model. Those perturbations are taken from CMIP5 ensemble simulations (GCMs) following the RCP8.5 emission scenario. Here, we are using results from those simulations in both current (CTRL) and future (PGW) climate to investigate the possible impact that global warming would have on this snow storm.

The accumulated precipitation during the event is shown in the figure attached. We can see a major increase in total accumulated precipitation in a warmer climate, which is mainly in the form of rain, for the exception of a localized IP event northeast of Lake Ontario. This could be caused by a combination of an increased warm air advection at 850 hPa (due to stronger winds) and an increase in the surface temperature (from -5°C in CTRL to 0°C in PGW), favoring the formation of IP. Also, the snow event in the Montreal area is slightly enhanced (up to 12 cm more snow) and is confined in the Saint-Lawrence River Valley (SLRV); the event is not shifted poleward, as one could expect. The increase of temperature implies an increase in humidity, which leads to an increase in precipitation even if the temperature is below 0°C. This is possibly the reason why there is more snow in this area.

There are plenty of features and processes that we haven’t looked at yet for this case study. For example, does the wind channeling in the SLRV have an influence on the IP event northeast of Lake Ontario? Is there any more/less moisture transport in PGW than in CTRL? Is there any change in the wind direction near some sites of interest in the warmer climate? We will also be looking at an ensemble of near 0°C storms (CTRL and PGW) in the future, which could lead to a nice dataset to study the mesoscale and microphysics features of those events. I will keep you posted on future results.

Figure: Accumulated precipitation (mm of water-equivalent) on December 26-28 2012 in CTRL and PGW, as well as the bias PGW-CTRL. Rain (R), freezing rain (ZR), ice pellets (IP), graupel (G) and snow (S) are shown on those plots.

Add new comment