Comments on: IPCC/RCM climate data

Question: IPCC/RCM climate data

allelopath — Tue, 19 Jun 2007 17:29:48 -0800

Need IPCC/RCM climate data.

I am looking for climate data from IPCC 2007 for RCM (Regional Climate Models). I am told there are about 20 model runs. I need scalar data of precipitation (as a multiplier) and temperature (+/-) for the southwestern US. The data may be here somewhere: http://www.ipcc-data.org/, but I'm not finding it.

By: Pinback

Pinback — Tue, 19 Jun 2007 19:59:39 -0800

I dug into this a few weeks ago for an assignment, and ran up against the issues mentioned in this FAQ - specifically, the entries on RCMs and downscaling.

Looks like you can't get the data from the IPCC. You either have to go direct to the source (your local weather service records) for all that you want, or take what local info you can get and use it to downscale from the GCM. Either way, you'll be wanting some extremely fine-grained data, and your error window will be high.

(In the end we couldn't get the data required in time / budget, and it was largely an interest exercise anyway, so we just ran with the various GCMs. The RCMs are somewhat moot anyway, until you have enough of them spanning the whole world to build up / compare to various GCM scenarios.)

By: carmina

carmina — Tue, 19 Jun 2007 20:22:18 -0800

This is exactly the type of question I never expected to see on AskMe. AskMe never ceases to amaze me.

The main repository for the IPCC-AR4 model runs (the 20 something models you are mentioning) is hosted by PCMDI at Lawrence Livermore Laboratory. The data portal is here. The new users orientation site is this. They want you to email them and register your project so you can get a password and the ability to ftp/download the data. They are very nice people and helpful. Check the subprojects website for already published and ongoing work in the area of your interest.

However, these are global runs, not regional. Regional models are typically run "downscaled" from these IPCC runs. IPCC AR4 does not have a main repository for regional model output. You may want to check with other people who do such runs independently. Some resources to start with:
University of Arizona, Berkeley, NCAR, a recent AGU meeting talks. The way to go is contact people and ask for their model output. One of the best people in the field (coordinating author for IPCC -like uberoverlord) has done work on the SW US future scenarios.

Beware though that there is some controversy on how regional models are run for climate prediction.

By: carmina

carmina — Tue, 19 Jun 2007 20:24:07 -0800

Here are the AGU session talks.

By: carmina

carmina — Tue, 19 Jun 2007 20:26:28 -0800

damn. Gerry Meehl is said uberoverlord.

By: allelopath

allelopath — Wed, 20 Jun 2007 04:48:30 -0800

>>You either have to go direct to the source (your local weather service records) for all that you want
But the 'data' that I want is not data in the purest sense, but rather output of the RCMs, ie predictions of precipitation and temperature.