Climate Dynamics Group

Group Leader: C.S. Frederiksen

Objectives: To use climate models to improve understanding of climate predictability, variability and change, and to improve the performance of climate models through evaluation and diagnostic studies.

Research activities

Climate model evaluation and experimentation

The group continued its involvement with design of and participation in international model experiments, including the Atmospheric Model Intercomparison Project (AMIP), the Coupled Model Intercomparison Project (CMIP), the Cloud Feedback Model Intercomparison Project (CFMIP), and the Climate of the 20th Century (C20C) project. An AMIP-2 project examining the potential role of soil moisture and its effect on Australian surface temperature variations and predictability was completed. The tropical diurnal variation of clouds and OLR has been evaluated using International Satellite Cloud Climatology Project (ISCCP) observational data and results from fourteen AMIP-2 models. A number of deficiencies in the ability of models to simulate aspects of the diurnal cycle, especially in the tropics have been identified and documented in an internal report.

Membership of AMIP and CMIP Science Panels has involved contributions to revisions of the AMIP experimental protocol, a new protocol for a coupled 20th century experiment, and input to the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) including two IPCC scenarios with stabilisation and a standard model variable output list.

Further work has been done on the C20C project, with additional forcings (observed time series of CO₂, volcanic aerosols, and observed trends in stratospheric ozone) added to BAM3 with the help of the Model Development group. An ensemble of ten runs for the period 1948-2002 has been completed and a standard set of C20C diagnostics was prepared for a workshop in Trieste, Italy, during April 2004. Scripts have been written to analyse all C20C models in two sub-projects ((a) validation of predictable and chaotic components of interannual variability and (b) patterns of predictable and chaotic variability) which the group is leading.

Simulations with the BMRC climate model coupled to the Chameleon Surface Model (CHASM) were used to study the strength of land-atmosphere coupling in models, and to determine the impact of varying land surface energy balance complexity on the surface climate simulated under AMIP-2 experimental conditions.

Collaboration with the Model Development group on development of BAM 4.0 assisted with diagnosing and reduction of climate biases.

Climate variability and predictability

A new technique, for extracting interannual patterns of the climate signal (predictable) and intraseasonal weather noise (unpredictable) components of seasonal means in observations and climate models, was further developed. A study of Northern Hemisphere geopotential variability, using this new method on NCEP and ERA40 data, was documented. Scripts for evaluating interannual variability in climate models from the first stage of the C20C project have been written and applied to BAM, Meteorological Research Institute of the Japan Meteorological Agency (MRI), CSIRO and Hadley Centre results.

A comprehensive study of the inter-annual variability of Australian surface air temperature and the relative importance of the predictable and weather noise components was conducted, using observations and climate models.

Close collaborations have been developed with Chinese scientists at the Laboratory for Climate Change, China Meteorological Administration (CMA) to study the interaction of the Australian-Asian monsoon, the role of Eurasian continental processes, and the impacts of Chinese land-use on local and regional climate. A number of observational datasets (snow, soil moisture and frozen soil etc.) have been collected for further observational study and model validation and some possible collaborations in climate and climate change studies have been discussed.

Climate change

The group continues its involvement with specific climate change research projects for the Australian Greenhouse Office.

Major developments in the study of feedbacks in climate change experiments include: (a) completion of bringing the feedback code up-to-date with BAM 4.0; (b) extension of feedback code for analysis of CFMIP results. This included generalising the code to allow for changes in model resolution, frequency of archiving, and handling differently specified cloud fields; (c) extension of feedback codes to permit choice of the Sun Edwards Slingo radiation code as an alternative to the Fels/Schwarzkopf and Lacis/Hansen codes. This will permit testing of the hypothesis that calculated climate feedbacks should not be sensitive to a choice of analysis radiation scheme different from that used in the original experiment; (d) extension of feedback code to permit calculation of horizontal and vertical components of temperature and water vapour feedback corresponding with uniform temperature change, changes in horizontal and vertical temperature gradients, and changes in relative humidity. This permits the calculation of conceptually simplified divisions of water vapour and temperature into components.

The group contributed to the work across BMRC on modifying and extending the BMRC coupled model as used for seasonal forecasting to make it suitable for climate variability and climate change studies.

A study was performed on the relative importance of complexity of land surface in climate simulations. Other studies undertaken include instabilities, interactions and circulation anomalies, in particular examination of the effects on the Indian monsoon, intraseasonal oscillations, Southern Hemisphere storm track locations and the dynamics of blocks and northwest-cloudband (NWCB) disturbances. An initial instability analysis, using NCEP reanalysis data, of the winter periods during 1949-1968 and 1975-1994 has been conducted to try to understand the climate changes that occurred in the early 1970s, especially over south-west Western Australian. Changes in the mean circulation between these two periods have been shown to have quite large impacts on the storm tracks. These changes are consistent with the observed decrease in precipitation in the latter period. In particular, our results show that cyclogenesis over WA is around 25 per cent weaker in this latter period with the storm track shifted further downstream.