CMIP5 Updated Climate Projections: Our Best Global Effort – An interview with Imtiaz Rangwala
[CMIP5] models represent our best global effort and a huge amount of money and resources. Basically, we’re trying to constantly improve the representation of all the processes in the earth’s climate system.
Many western water managers use the CMIP3 downscaled climate projections to plan for the effects of climate change on their water systems. Now that the CMIP5 projections have been released, water managers are wondering what this means for their investment in CMIP3. In this interview, we explore the topic with climate expert Imtiaz Rangwala.
What are the major differences between the CMIP3 and CMIP5?
CMIP5 includes new emissions scenarios, more model simulations, and many more archived climate variables. Many of these are available at daily time steps, so we have much more input available to drive impacts models, compared to CMIP3.
Globally and regionally, the two projects show very similar projections, which is reassuring and generally in accord with our basic understanding of climate physics. The similarities are consistent at seasonal scales, and the magnitude of change is generally a function of the emissions scenario: the higher the emissions, the greater the change. For example, CMIP5 includes RCP 8.5, a very high emissions scenario with no analog in CMIP3. If you consider that scenario, you’ll see much more rapid climate change even by mid-century, which could definitely change your previous numbers.
Do organizations that have used CMIP3 projections need to start their assessments all over?
Not necessarily. CMIP5 projections do not entail any major shift in thinking—rather, they give water managers more details they can add to their projections and integrate into the planning they’ve already done.
Broad-scale patterns remain the same between CMIP3 and CMIP5, but we notice some fine-scale differences in some regions. For example, in the upper Colorado River Basin, the spread of CMIP5 projections is slightly more wet (or less dry) compared to CMIP3 during spring and summer, which tends to make the multi-model mean about 5% wetter in CMIP5. Although several large-scale atmosphere-ocean circulation processes are better represented in the CMIP5 generation of models, at this point we cannot say whether CMIP5 projections are more credible than CMIP3 for any particular region. Inter-model differences in precipitation projections are equally large between CMIP3 and CMIP5. Furthermore, CMIP5 GCM resolution is still pretty coarse, so the mountains and their effects are not represented adequately.
Are physical improvements in CMIP5 causing wetter projections in the American Southwest? That’s an interesting question that no one has taken up yet. I believe the Bureau of Reclamation is talking about pursuing it. But as a water manager in the Southwest, if you have already planned for the slightly drier climate that CMIP3 projected, you are probably well prepared.
If you were making a high-level case to water utility customers or stakeholders for upgrading projections using CMIP5 scenarios, what would your major points be?
You keep your computer software up-to-date, right? You have to upgrade to new information in much the same way. For the next seven years or so, people will be talking the language of CMIP5, so you should be aware of that, and at least document the major differences between CMIP3 and CMIP5 projections for your region.
CMIP5 could be especially helpful as we attempt to understand and quantify extreme climate events, primarily because of the daily data that has been archived. Already, new kinds of extreme indices are coming out of CMIP5 projections. Much of the climate impact assessment based on CMIP3 was developed using mid-range emission scenarios, but considering a more extreme emission scenario, such as RCP 8.5, will likely show more aggravated impacts. Water managers and policy makers should be at least aware of that possibility.
At the same time, the complexity involved in generating these projections is huge, so we must use the models intelligently. We should be aware of the major sources of uncertainties related to the projections relevant to our region. Nonetheless, the models represent our best global effort and a huge amount of money and resources. Basically, we’re trying to constantly improve the representation of all the processes in the earth’s climate system. CMIP5 represents that effort.
Imtiaz Rangwala is a research associate with the Western Water Assessment and CIRES at the University of Colorado, Boulder. He is a climate scientist trained in assessing and diagnosing regional scale climate change, with a focus on high elevation regions such as the Colorado Rocky Mountains and the Tibetan Plateau.