02 October 2009
Understanding Climate Modeling
Podcast with scientist Greg Holland
(begin text)
Narrator:
This is an America.gov podcast.
For print versions of articles, multimedia, and subscription information, visit www.america.gov.
Narrator:
America.gov recently spoke with Dr. Greg Holland, senior scientist and acting director of the Earth and Sun Systems Laboratory at the National Center for Atmospheric Research in Boulder, Colorado. Dr. Holland is an expert on climate modeling, the primary way that scientists try to understand the enormous complexity of the Earth's climate and how climate change may affect the planet in the future. Climate models themselves have grown more sophisticated in the last few years. We asked Dr. Holland how these models have given policymakers and planners a window into the impact of climate change.
Dr. Greg Holland:
Well first let me say that climate science, like anything to the weather, is definitely inexact. But having said that, if you look at the way that weather forecasts have improved over the past two decades, they've improved entirely because of the computer models. And it's the same with climate science. The problem with climate science is that it takes you a while to actually prove they're right because you have to wait 30 years. But it’s exactly the same process, and the climate models have gotten to the point where the information is real and there are uncertainties associated with it which you can take into account. The biggest single exciting thing is the fact that over the past 10 to 15 years the climate models have gone from being very much maligned and very much discussed as not really providing information, to now you have the IPCC saying global warming is unequivocal. The more exciting things which are just starting to happen now are that up until quite recently climate has been described in its global context: there will be a global temperature rise of x, there will be sea level rise by x feet. Now, we're actually zoning down to be able to say, given that, what's going to happen to the weather in Washington, what's going to happen to hurricanes in the Gulf of Mexico. But the big problem there is that in the short term, the uncertainties go up because the finer the scale you come down to, the harder is it to get really hard information. So it's an exciting time if you're working in the field because this is where the real emphasis on climate should be, on the regional and local. By the events to date, with the uncertainties, in some cases they've under-predicted what we've observed to happen. In some cases they've over-predicted what we've observed to happen. But all of the things that we're seeing now, with one or two exceptions, are within the error bars that were on the models back then, in other words we're on-line given the uncertainties. The exceptions, the biggest exception is the Arctic, which is melting far faster that what was originally thought. With a caveat: it was always known that when the Arctic started to warm, it would go very fast. But most of the models had it going a little bit longer before it went through that fast cycle. Unfortunately, the fast cycle started earlier.
Narrator:
Regional models of climate change are one of the most useful advances of the past few years. But the diversity of climates poses challenges as more detail is added into models. Although climate change is a global process, the effects are felt in different ways at local levels.
Dr. Greg Holland:
Current climate models— let me give you an idea how it works. Current climate models simulate everything you can see outside your window. It's got the atmosphere; it's got the rain; it's got the sun and everything else; it's got the clouds; it's got the trees, and they grow; it's got the oceans, and it's got fishes in the ocean and they actually swarm around — I mean it's got all those things in there and as they add more, and as we come down to finer resolution, the uncertainties go up simply because at smaller resolution, you can't be anywhere near as certain. I can be a lot more certain about what's going to happen to the globe from a radiation change that I can be what's going to happen in Washington, D.C.. Firstly, the models themselves are not perfect. And secondly, when you have a rapid feedback cycle, it's to a large extent unpredictable when it starts. It's sort of like a thunderstorm: I can tell you there's going to be thunderstorms this afternoon, but I can't tell you there's going to be one forming at 3:30. It might form at 4:30. But once it gets going, it's going to go very quickly, you know that. The Arctic warming is in that sort of “goes very quickly” — the ice melts to a certain point, then the incoming radiation that which used to be almost entirely reflected back to space by the ice is now being almost entirely absorbed by the ocean, and bang — away it goes. The next really big one which we know is likely to happen, but we're not sure when or where, is if the permafrost starts melting. Because you put the worst greenhouse gases in the world into the atmosphere in enormous amounts, methane.
Narrator:
So how are climate models being received in the government, business, and civic communities? In the past, climate models offered information that was difficult for managers to apply to their own organizations. But now, the models have improved to where changes in organizational planning are taking climate change into account.
Dr. Greg Holland:
I think it's fair to say that most reasonable governments around the world now accept that there's a problem. The climate services are becoming a big issue. NOAA's never had a climate service; they're now in the process of developing one. The Europeans have never had a coordinated climate service; they're now in the process of developing one. Right now we're at the stage where major industries, state governors, and people like that are actually starting to use the information because they think there's sufficient information to start to make rational decisions. And we're now working towards a stage where we'll also take account of the multidecadal variability in the climate. Again, the devil is how they react to it. You mention China, and actually China's starting to do some very good things with renewable energy. But at the same time, they're still opening between 2-3 coal-fired power stations a week. And it's really hard to sort of point the finger because if you live in the U.S. or Australia or Europe, we've already spent that carbon budget if you like, and countries like China and India are saying, well, why should you be rich and we shouldn't be rich, so it's a difficult problem. With regard to the future, the way climate modeling and climate forecasting is going, they're going in one very important direction. Up until the present, what you have seen is climate projections. In nowhere in the IPCC or anywhere else will you find the word prediction. What they'd say is here is what we project, given these conditions and given what we've heard. Well from here on in, you're going to be getting climate predictions, and the predictions are going to be on 10-, 20-, 30-, 40-, 50-year timeframes — in other words the timeframe that most people plan on.
Narrator:
The United States is now among the leaders in seeking solutions to climate change and preparing for its impact. But Dr. Holland reminds us that halting climate change will require skills and understanding of how climate models work and the best ways to use the information provided. Dr. Holland gave us some details on what those requirements are.
Dr. Greg Holland:
Two or three things. The first thing is those that actually need the information should be looking quite carefully at what's available now and should also remember what's available now will be improved next year and the year after. That's the pace of change at present. The high-resolution, regional stuff that we're doing now couldn't have been done two or three years ago. So there's going to be a rapid sort of improvement. But they also need to take good account of the uncertainties as it goes along. The second thing is that what's going to hold climate prediction back is partly just simply a lack of resources. You need the biggest and fastest computer on earth; that's happening about as fast as it could go. But you also need a reasonably good cadre of especially young people that are going to come in and make a career out in it, and there's a real need to bring in younger people, get them sort of trained up so that they can carry through over the next few decades in the process. And I guess it really does behoove planners, be they commercial or government, to start getting expertise on their staffs, because it is going to be the fundamental basis of planning.
Narrator:
This podcast is produced by the U.S. Department of State's Bureau of International Information Programs. Links to other Internet sites or opinions expressed should not be considered an endorsement of other content and views.
(end text)
Listen to this Podcast (MP3)