Refining rainfall projections for the Murray Darling Basin of south-east Australia—the effect of sampling model results based on performance

Abstract

One of the aims of developing new climate projections is to better address the requirements of stakeholders-particularly those who require less uncertainty and/or probabilistic information to work with. Projections are continually updated over time as more, and newer, climate model simulations of the future become available but this can introduce problems when it comes to interpreting large samples with differing results. Regional projections of rainfall are characterised by a high level of uncertainty, partly because of different sensitivities of the different models. Some models can be demonstrated to perform relatively poorly when assessed by their ability to simulate present-day means and variability and here we show that the uncertainty in model projections can potentially be reduced when the projection from these models are either discounted or ignored entirely. When applied to the Murray Darling Basin of south east Australia, it is possible to demonstrate a clustering of the results from the better performing models. These indicate that the rainfall changes to be expected as a result of increased greenhouse gas concentrations into the future are more likely to be at the drier end of the full set of model results. This occurs because the better performing models indicate decreases in winter and spring which are significantly different to the changes indicated by the other models. These results suggest that there are compelling reasons for discounting, if not entirely dismissing, some model results based on their failure to satisfy some basic performance criteria.

Projected climate change in Australian marine and freshwater environments

“…Unfortunately, the observational data to validate models are often not available for aquatic biologists, nor are they within their typical skill set. Thus, using a suite of models to make an ensemble average will be the best available option for marine and freshwater researchers, although there is some debate as to whether this approach provides the most robust projections for regionally specific climate-change projections, because inclusion of poor models can bias the ensemble average (Perkins et al 2009;Pierce et al 2009;Shukla et al 2009;Smith and Chandler 2010). Even though the resolution of climate models is improving, model scale is still considered coarse (,100-200 km) with regard to representation of the environmental and biological processes that many aquatic biologists are interested in (,10 km), such as reef-specific recruitment or growth.…”

“…A challenge given the computing time this involves is to generate a suitable suite of scenario-model combinations so that future conditions and the associated uncertainty can be assessed. To improve confidence in model-based projections, historical validation is important (Smith and Chandler 2010). In the case of Australia, this means a commitment to maintaining critical in situ ocean-climate observations, now realised under the IMOS program (www.imos.org.au, accessed 24 August 2011).…”

The relative performance of Australian CMIP5 models based on rainfall and ENSO metrics

“…However, as noted by Knutti (2010), 'metrics and criteria for model evaluation must be demonstrated to relate to the projection'. If we consider just the Australian region, previous model assessment studies include: Hope (2006a, b), who pointed to problems with simulations of the winter trough over the southwest of the continent; Suppiah et al (2007), who assessed the performance of models with respect to how well they reproduced patterns of seasonal average temperature, mean sea level pressure (MSLP) and rainfall; Watterson (1996Watterson ( , 2008, who used a statistic determined from simulated and observed patterns of seasonal average temperature, mean sea level pressure (MSLP) and rainfall over the continent; Perkins et al (2007) and Maximo et al (2007), who considered the ability to simulate daily rainfall and daily minimum and maximum temperatures for different regions; Charles (2007), who focussed on the ability to simulate both daily MSLP patterns and the seasonal cycle of monthly average MSLP; Colman et al (2011), who noted large model biases in simulating features of the monsoon; and Smith and Chandler (2010), who focussed on the Murray Darling Basin of southeastern Australia and assessed models in terms of their ability to simulate key features of both rainfall and the El Niño Southern Oscillation (ENSO).…”

Adaptations Assessment on the Impacts of Flooding Under Current Condition and Climate Change Scenario, Case Study: Centini Village, Indonesia

“…Others are using government data such as from statistical bureau to validate the historical conditions such as population numbers, migration, crude death rate (CDR) and numbers of housings. Especially for the rainfall pattern in the future, I compare relevant six global climate change models (GCMs) for the Australasia context (Smith and Chandler, 2010;Katzfey et al, 2010). Those GCMs are GFDLC2.1, GFDLCM 2.0, ECHAM 5, HadCM, Micro 3.2 and Mk 3.5.…”