12 September 2008

Model Resolution

Policy makers are putting a lot of weight on climate change models, so it's important that they be as accurate as possible. Accuracy is taking some time to achieve:
Previous studies have indicated that climate change will cause major water shortages in the Middle East.

[snip]

The study based on global models, published in Climate Change last month (29 July), predicted a loss of about 170,000 square kilometres of viable rain-fed agricultural land, a longer dry season and changes in the timing of maximum rainfall.

However, a second study using a regional model indicated an annual rainfall increase of over 50 per cent around the Euphrates-Tigris watershed, which feeds the two largest rivers in the Middle East...

"There are regional scale phenomena not captured by the global climate models that can substantially change the projected climate change for particular regions," Evans told SciDev.Net. "So while the global models provide good estimates of the change at large scales, they can be quite wrong in particular locations."

"To capture these particular phenomena the resolutions need to increase by about ten times," Evans says
Bottom Line: The best science is confirmed by replication, and these models (and data used for input and confirmation!) are working in the right direction. Quickly, please.

3 comments:

Bob Murphy said...

I'm really not trying to be a jerk here: What do you mean they're "working in the right direction"? What would it look like if they were still hopelessly muddled? Wouldn't you get one model saying the exact opposite of the other?

David Zetland said...

"in the right direction" in terms of finding common ground in results (NOT by deciding in advance to reproduce others' results). Once we agree (e.g., gravity), we can make progress...

Bob Murphy said...

So what was the common ground here? They both agreed there will be rainfall in the Middle East, and just disagree about how much?

I realize I'm coming off sarcastically, but I still don't get how this episode illustrates your point. If one physicist said gravity causes an object's height to go way down, while another said gravity will cause its height to increase 50%, I don't think you would have cited it as an example of common ground.