Pielke Sr., R.A., and R.L. Wilby, 2011: Regional climate downscaling – what’s the point? EOS. January 31 2012 pages 52-53
has appeared [for those who do are not members of the AGU, a copy of the manuscript can be obtained here).
This article presents a list of reasons why multi-decadal regional climate forecasts are unable to provide skillful predictions to the impacts community. These can be summarized as:
As a necessary condition for an accurate prediction, multidecadal global climate model simulations must include all first-order climate forcings and feedbacks. However, they do not.
Current global multidecadal predictions are unable to skillfully simulate regional forcing by major atmospheric circulation features such as from El Niño and La Niña and the South Asian monsoon.
While regional climate downscaling yields higher spatial resolution, the downscaling is strongly dependent on the lateral boundary conditions and the methods used to constrain the regional climate model variables to the coarser spatial scale information from the parent global models. Large-scale climate errors in the global models are retained and could even be amplified by the higher- spatial- resolution regional models. If the global multidecadal climate model predictions do not accurately predict large-scale circulation features, for instance, they cannot provide accurate lateral boundary conditions and interior nudging to regional climate models.
Apart from variable grid approaches, regional models do not have the domain scale (or two-way interaction between the regional and global models) to improve predictions of the larger-scale atmospheric features. This means that if the regional model significantly alters the atmospheric and/or ocean circulations, there is no way for this information to affect larger scale circulation features that are being fed into the regional model through the lateral boundary conditions and nudging. For example, recent research indicates that terrestrial evaporation from the Eurasian continent contributes 80% of China’s water resources. In this case, the regional model domain has to be large enough to include areas that are connected by soil moisture feedbacks.
The lateral boundary conditions for input to regional downscaling require regional-scale information from a global forecast model. However the global model does not have this regional-scale information due to its limited spatial resolution. This is, however, a logical paradox because the regional model needs something that can be acquired only by a regional model (or regional observations). Therefore, the acquisition of lateral boundary conditions with the needed spatial resolution becomes logically impossible. Thus, even with the higher resolution analyses of terrain and land use in the regional domain, the errors and uncertainty from the larger model still persist, rendering the added simulated spatial details inaccurate.
There is also an assumption that although global climate models cannot predict future climate change as an initial value problem, they can predict future climate statistics as a boundary value problem. However, for regional downscaling (and global) models to add value (beyond what is available to the impacts community via the historical, recent paleorecord and a worst case sequence of days), they must be able to skillfully predict changes in regional weather statistics in response to human climate forcings. This is a greater challenge than even skillfully simulating current weather statistics.
Multi-decadal regional climate predictions (i.e. regional climate downscaling) has practical value but with the very important caveat that it should be used for model sensitivity experiments and not as predictions.
Our bottom-line conclusion is that
It is therefore inappropriate to present [multi-decadal climate prediction] results to the impacts community as reflecting more than a subset of possible future climate risks.
As I have written on my weblog, e.g. see
the huge expenditure of funds to provide to the impacts community, what are claimed to be skillful predictions of regional climate decades into the future, and to present them as the envelope of what could occur, is a waste of money and people’s time. It is also seriously misleading policymakers on the risks we face from climate and other environmental threats in the coming decades.
I welcome guest weblogs from climate scientists and the impacts community that either seek to refute our findings or to provide additional reasons on the failure of multi-decadal climate downscalings as a skillful prediction tool.