Home

Back

page regenerated:
15 Jun 2009

Climate Change in Queensland under Enhanced Greenhouse Conditions (Annual Report 2002-2003)

A CSIRO report on research undertaken for Queensland Departments of State Devlopment, Main Roads, Health, Transport, Treasury, Public Works, Primary Industries & Fisheries, Environmental Protection, Natural Resources & Mines, Queensland Rail, SunWater, CS Energy.

Wenju Cai, Steve Crimp, Kathy McInnes, Barrie Hunt, Ramasamy Suppiah, Mark Collier, Tracy Elliot, Kevin Hennessy, Roger Jones, Cher Page, and Penny Whetton

[Acrobat versions available at the end of this page]
Hard copies are available by emailing to rouseabout@nrm.qld.gov.au

Summary for Policymakers
This report summarises the work of the first year of a second four-year project examining the impacts of climate change in Queensland. The initial focus of the contract was to obtain a definitive sign and/or narrower range for projected rainfall changes over Queensland using both the improved CSIRO Mark 3 coupled climate model and the application of a multi-model consensus technique. A series of milestones were determined as part of the contract agreement. All aspects of the work associated with these milestones have been completed. The following provides a summary.

Task 1
The first task of this contract was to perform a detailed assessment of the climatology for Queensland as simulated by the Mark 3 coupled model for current conditions. This is a necessary task prior to making any assessment of climatic changes under enhanced greenhouse conditions. Comparisons were also made with the climatology from the earlier Mark 2 coupled model in order to illustrate improvements. We find that:

1. There has been a significant improvement in the ability of the model to simulate both realistic ENSO cycles and global ENSO-rainfall teleconnections. The Mark 3 climate model is one of the few in the world that can do so.
2. The model still suffers from a common "cold tongue" problem, in which the cold water zone of the central and eastern equatorial Pacific extends too far west. While this is problem common to almost all coupled models, deficiencies in several fields (e.g. rainfall and evaporation) appear to be associated with this bias. CSIRO is one of several leading modelling groups in the world that are endeavouring to overcome this problem.
3. The model is also able to simulate tropical cyclone events with realistic intensity, spatial scale and frequency. Until now, this has been extremely difficult.
4. The simulated hydrological fields from the model are generally improved but in some fields (soil moisture, runoff) improvements are difficult to demonstrate. This is believed to be due to the fact that the resolution is still unable to resolve the appropriate spatial scales.
   

Task 2
The second major task was to analyse projected climate change to 2100 AD and beyond from a transient CO2 simulation that included other forcings such as aerosols. The CO2 forcing is based on the A2 projection of the Special Research Emission Scenarios (SRES) and follows the observed evolution from 1870 to 2000, then follows the projected CO2 levels of the A2 scenario from 2001-2100. By 2100, the equivalent CO2 reaches a level that is more than three times the level of 1870 (concentration ppm). Thereafter, both the CO2 and aerosol levels are held constant and the model integrated for another 150 years. This simulation was then contrasted with a control simulation, in which the CO2 and aerosols were kept at 1870 levels. The focus of the analysis has been the response of hydrological processes at regional scales, an assessment of the influence of multi-decadal variability, and the response of simulated ENSO events. We find that:

1. The model produces distinctively different surface temperature responses in different catchment areas, with a small warming rate in coastal areas increasing toward inland regions.
   
2. From early in the 21st century, the model produces a decreasing rainfall trend in all catchment areas (in the range of 5-20% by 2100 for summer season). This is particularly clear after interannual signals are removed. The decrease in rainfall is associated with the development of an El Nino-like warming pattern in the equatorial Pacific Ocean. This result is consistent with those previously found in an ensemble of results from the previous Mark 2 model.
   
3. Alongside the rainfall decrease, actual evaporation, soil moisture and runoff all exhibit decreasing trends. It is important to differentiate potential and actual evaporation. Potential evaporation increases with temperature at a rate of between 2-8% per degree warming in global mean temperature.
   
4. The results suggest a tendency for a more variable climate, with more extreme flood and drought events, and hence stronger variations in soil moisture and runoff. They also imply that the impacts of drought events will be more severe, being exacerbated by the increased temperature and evaporation.
   
5. During summer, the impact of decreased rainfall on soil moisture is disproportionately large compared to winter since temperature represents the major driver of soil moisture retention.
   
6. Referenced to the present day (control) climate, there is a tendency for slightly increased amplitude of ENSO events.

Task 3
The third major task was to investigate a range of possible climate change outcomes by pooling together the results from the CSIRO model with those of its global peers. The focus was to be on the results for rainfall and other hydrological processes. We find that:

1. From a total of 12 sets of model results, 7 show a decrease in annual mean rainfall. The remainder shows either no change or else a small increase. Eliminating those models which are unable to produce a reasonable seasonal cycle of rainfall over Queensland yields 7 out of 10 showing a decrease.
   
2. Of the 9 models that provided results, annual mean potential evaporation increases. The increase for the Queensland region is in the range of 2-8% per degree warming in global mean temperature.
   
3. As a consequence of increased potential evaporation/decreased rainfall, the water balance deficit increases in all 9 models.
   

Task 4
The fourth major task was to analyse the results from an ensemble of climate change simulations using the same model. The aim here is to better identify the climate change signal above the climatic "noise"or "natural" climate variability. Computational demands made this impractical with the Mark 3 coupled model but achievable using the coarser resolution Mark 2 model. This model was forced by a range of SRES emission scenarios, for a total of 12 experiments. In addition, the Mark 3 atmosphere-only model was used to generate an ensemble of 10 sets of results based on forcing by observed sea surface temperatures over the past century. We find that:

1. The ensemble mean result of the 12 SRES experiments shows a decrease in annual mean rainfall, although the trend is weak when compared with the amplitude of interannual and inter-decadal rainfall variability. The decrease is also discernible in the majority of the individual experiments.
   
2. The decrease in rainfall trend is well "registered"in the response of soil moisture, which shows a corresponding decrease in the mean of the 12 ensemble experiments, and in the majority of the individual experiments.
   
3. The results from the 10-member ensemble using the Mark 3 atmosphere-only model indicate that the basic features of observed extreme events over recent decades were reproduced in a majority of the experiments.
   

A significant achievement of the first Consultancy year was the development of more certain rainfall scenarios over Queensland. On average, annual rainfall is projected to decrease under enhanced greenhouse conditions, and this is the case with the majority of other global climate models. This finding represents both opportunities and challenges. It is an opportunity because we can now start to assess its likely impacts. The challenge is to quantify the extent of the reduction on regional scales. This is where regional models of enhanced spatial resolution are required. There is also a need to better define the human-induced climate signal from that due to natural variability. These issues will be addressed during the next phases of the Consultancy agreement.

Acrobat document(s). You may need to download the free Adobe Acrobat Reader.