Since many areas of the West get little or no summer rain, winter snowpack is their lifeblood. The gradually melting mountain snow refills reservoirs and groundwater aquifers as thirsty people drain them, helping farms and cities survive months of hot, dry weather.
But as California was reminded during its long, five-year drought, deep winter snows are no guarantee. This seemed like an anomaly. But a new study warns that winters with substantially less mountain snow could become the norm.
The study, published in the peer-reviewed journal Nature in April, was led by a team of scientists in Canada and the U.S. It not only verifies that the snowpack has declined already in many areas of the West over the past three decades but also concludes that this decline was caused by greenhouse gas emissions. And it forecasts that additional declines in snowpack of up to 60 percent are possible over the next two decades.
That’s a major decline that would have huge impacts on Western economies.
To learn more about the findings, Water Deeply recently interviewed the study’s lead author, John Fyfe, a senior research scientist at the Canadian Centre for Climate Modelling and Analysis in Victoria, British Columbia.
John Fyfe: There are two aspects in my mind. There was analysis of the historical observation record, where we established there was a certain decline – between 10 and 20 percent – in snowpack since the 1980s. That isn’t entirely new information. Other people have come up with that number in different ways. But I think this is the first study that comprehensively used all the observations that were available.
The second aspect was a formal statistical approach to detecting an anthropogenic signal in the observations and attributing it to a human influence. We are now able to make a strong detection and attribution statement that humans were involved in the observed decline in snowpack that we’ve seen since the 1980s.
Water Deeply: Has that connection to human influence been made before?
Fyfe: There was another paper by one of my coauthors, maybe eight years ago, that used earlier observations, less complete, and not as advanced computer models in order to make this statement. And it was more specific to smaller regions around the West. This is the first analysis that covers the whole entire Western United States. So it’s not 100 percent novel; it’s maybe 80 percent novel.
Water Deeply: You also make predictions about the snowpack.
Fyfe: Yes, that’s part two of the paper. The main emphasis of this paper has to do with being able to quantify the contribution due to human causes. So now we’ve established that, between now and 2040, we could see another 30 percent decline in the snowpack due to human activity alone. So if it was just humans acting on the climate system, we’re predicting a 30 percent decline in snowpack between now and 2040.
There is also natural variability. The system does bounce around and change independent of humans. We were able to quantify the internal variability contribution using an unprecedented large set of climate model simulations.
It was a set of simulations that took several years to complete. And it’s that new set of simulations that allowed us to make what is the most notable statement in this paper, and that is that we’re predicting up to a 60 percent decline in snowpack over the Western U.S. That’s a combination of 30 percent from human impact, and possibly an additional 30 percent decline due to this natural variability. That’s the key result. It is a big number. It’s frightening.
But that does cut both ways. Internal variability not only can add to the reported loss. It can also subtract from the human contribution. So there’s a range of losses – from a 60 percent loss up to a very small increase through 2040.
Water Deeply: Does that large range you detected suggest a loss in snowpack is more likely?
Fyfe: It’s very likely we will see a 30 percent loss. It’s less likely that we’ll see a 60 percent loss or an 8 percent gain.
That 60 percent loss is toward the extreme end of the losses we might see. But it can’t be ignored either. It could possibly be headed in that direction.
Water Deeply: Can you clarify exactly what you mean by the “human influence” on snowpack?
Fyfe: This is a computer modeling exercise, so we use state-of-the-art computer models. These are global climate models that include oceans, land, atmosphere, ice and carbon. These are actions between carbon and the climate system.
A typical experiment is over a historical period where we have observations of how much CO2 humans have put into the atmosphere, how many aerosols, how many pollutants, and you run experiments that include all of those human additions to the atmosphere. Then we do another experiment where we run the model without human influence, only natural internal influences. Then we compare the two sets of experiments.
And the conclusion is, over the historical period, that we cannot reproduce the actual observed evolution of snowpack over the region unless we introduce the human contribution to the chemistry of the atmosphere. Without human influence, it does not produce a loss in snowpack. Primarily, the driver is increasing greenhouse gases. It’s only when we introduce the observed amount of greenhouse gases that humans have put into the atmosphere that we can reproduce the observational record.
Water Deeply: In the study, did you examine the changes in precipitation from snow to rain?
Fyfe: We did not do that in the observations. But we did do it in the modeling. We analyzed the precipitation changes between now and 2040, and in the model, in our predictions, there is predicted to be more precipitation falling as rain rather than more snow.
The combination of the warming with increased greenhouse gases and the increased precipitation falling as rain rather than snow is a double whammy for the region. We saw an increase in total precipitation falling as rain rather than snow. Rain falling on snow is not a good thing.
Water Deeply: What are the implications of that?
Fyfe: We haven’t looked at that in any detail. It’s apparent that if you are to lose 60 percent of the snowpack between now and 2040, that is going to have an impact on the landscape. The main thing I would think we want to worry about is the impact on streamflow – the amount of water going down the major rivers.
And we actually are pursuing those questions right now. We know now, in a parallel study, that these losses in snowpack – historical and future – are tied to declines in Columbia River streamflow. That work is not published yet, but it will came out. It’s the same on the Colorado River. There is an impact on the Colorado as well, in our new calculations.
Water Deeply: Can you explain the role of a key weather system – the Aleutian low – in the fate of the Western snowpack?
Fyfe: That’s the internal variability story. If we do see a 60 percent loss, what we’re predicting is that it’s the consequence of the Aleutian low. It’s sort of the main circulatory feature in the Northern Pacific. It has a huge impact on weather and climate over Western North America. If it deepens [develops even lower pressure] over a decade or two – and it’s done this in the past and will in the future – it’s that situation that can cause this enhanced snowpack loss.
Associated with deepening of the Aleutian low is the circulation and wind. The circulation would change such that warm air over the tropical Pacific would be carried by the winds over Western North America. It’s those warm temperatures that would drive the loss in snowpack, partly. On top of that, the tropical Pacific is a very moist region. These winds carry not only temperature, they also carry moisture. So you get warmer temperatures and you get more moisture. The moisture, we predict, will fall as rain rather than snow. Here’s the double whammy: You get warmer temperatures driving snowpack loss, and you get more rain driving snowpack loss.
Water Deeply: How confident are you in these results?
Fyfe: In terms of the historical parts of this paper – looking to snowpack changes in the past – I would say we’re virtually certain the snowpack has declined. And we’re maybe more than 90 percent certain that humans have had an impact on that.
And going into the future, it’s highly likely there will be a loss of snowpack, and it’s likely that loss could be as high as 60 percent. We could attach probabilities. There are uncertainties and we can quantify those. We’re highly confident both in our understanding of what has happened in the past and our projections into the future. I would eat my hat if that wasn’t the case.
I did write a paper I think in 1998 – a long time ago – where we were looking at climate change at high elevations over the mountains. It was one of the first studies that did that. Back then, we did predict in a similar kind of way where we are right now. So we have verified our earlier results.