Carol McSweeney

Selecting Ensemble Members to Provide Regional Climate Change Information

AbstractClimate model ensembles, such as the Coupled Model Intercomparison Project, phase 3 (CMIP3), are used to characterize broadscale ranges of projected regional climate change and their impacts. The 17-member Hadley Centre perturbed physics GCM ensemble [Quantifying Uncertainty in Model Predictions (“QUMP”)] extends this capability by including data enabling dynamical downscaling of these ranges, and similar data are now being made available from the CMIP phase 5 (CMIP5) GCMs. These raise new opportunities to provide and apply high-resolution regional climate projections. This study highlights the importance of employing a well-considered sampling strategy from available ensembles to provide scientifically credible information on regional climate change while minimizing the computational complexity of ensemble downscaling.A subset of the QUMP ensemble is selected for a downscaling program in Vietnam using the Providing Regional Climates for Impacts Studies (PRECIS) regional climate modeling system. M...

Selecting CMIP5 GCMs for downscaling over multiple regions

The unprecedented availability of 6-hourly data from a multi-model GCM ensemble in the CMIP5 data archive presents the new opportunity to dynamically downscale multiple GCMs to develop high-resolution climate projections relevant to detailed assessment of climate vulnerability and climate change impacts. This enables the development of high resolution projections derived from the same set of models that are used to characterise the range of future climate changes at the global and large-scale, and as assessed in the IPCC AR5. However, the technical and human resource required to dynamically-downscale the full CMIP5 ensemble are significant and not necessary if the aim is to develop scenarios covering a representative range of future climate conditions relevant to a climate change risk assessment. This paper illustrates a methodology for selecting from the available CMIP5 models in order to identify a set of 8–10 GCMs for use in regional climate change assessments. The selection focuses on their suitability across multiple regions—Southeast Asia, Europe and Africa. The selection (a) avoids the inclusion of the least realistic models for each region and (b) simultaneously captures the maximum possible range of changes in surface temperature and precipitation for three continental-scale regions. We find that, of the CMIP5 GCMs with 6-hourly fields available, three simulate the key regional aspects of climate sufficiently poorly that we consider the projections from those models ‘implausible’ (MIROC-ESM, MIROC-ESM-CHEM, and IPSL-CM5B-LR). From the remaining models, we demonstrate a selection methodology which avoids the poorest models by including them in the set only if their exclusion would significantly reduce the range of projections sampled. The result of this process is a set of models suitable for using to generate downscaled climate change information for a consistent multi-regional assessment of climate change impacts and adaptation.

No consensus on consensus: the challenge of finding a universal approach to measuring and mapping ensemble consistency in GCM projections

Communicating information about consistency in projections is crucial to the successful understanding, interpretation and appropriate application of information from climate models about future climate and its uncertainties. However, mapping the consistency of model projections in such a way that this information is communicated clearly remains a challenge that several recently published papers have sought to address in the run up to the IPCC AR5. We highlight that three remaining issues have not been fully addressed by the literature to date. Allen and Ingram (Nature 419:224, 2002) While additional information about regions where projected changes in rainfall are not ‘statistically significant’ can provide useful information for policy, the spatial scale at which changes are assessed has a substantial impact on the signal-to-noise ratio, and thus the detectability of changes. We demonstrate that by spatially smoothing the model projections we can provide more information about the nature of the signal for larger regions of the world. Christensen et al. (2007) Combining information about magnitude, consistency and statistical significance of projected changes in a single map can cause reduced legibility. We demonstrate the difficulty in finding a ‘universal’ method suitable for a wide range of audiences DEFRA (2012) We highlight that regions where projected changes in average rainfall are not statistically significant, changes in variability may still cause significant impacts. We stress the need to communicate this effectively in order to avoid mis-leading users. Finally, we comment on regions of the world where messages for users of climate information about ensemble consistency have changed since AR4, noting that these changes are due largely to changes in the methods of measuring consistency rather than any discernable differences between the CMIP3 and CMIP5 ensembles.