Nature 512, 416–418 (2014)

Credit: JOHNNY LYE/THINKSTOCK

Snowfall is a key component of the Earth system that has important climatic, ecological and societal implications. Climate models indicate that mean snowfall will decline with warming in most regions, but with some increases in very low-temperature regions. However, the response of heavy (extreme) snowfall events — which have particularly large human impacts — remains less clear.

Paul O'Gorman from the Massachusetts Institute of Technology uses a climate model ensemble (taken from the Coupled Model Intercomparison Project Phase 5, under a high emissions scenario) and physically based theory to investigate changes in average and extreme Northern Hemisphere snowfall in response to twenty-first century climate change.

He finds that average snowfall declines with warming — in line with expectation — and snowfall extremes also decline in most regions, but the response is generally weaker than for the mean. This is explained by snowfall extremes occurring close to an optimal temperature that is insensitive to climate warming, which leads to smaller fractional changes at higher daily snowfall levels. These findings suggest that heavy snowfall impacts may not decline substantially under climate change and that snowfall extremes may be a poor candidate for climate change detection.