Report on AWAP / AWRA / CABLE modelling workshop
by Natasha Herron (BoM)
On 20 December 2013, the Bureau of Meteorology organised a workshop to discuss current and planned developments of AWAP, AWRA and CABLE modelling systems, with a view to rationalising ongoing development of hydrologic components and making best use of outputs generated by the models for Bureau water modelling applications. The meeting brought together 24 individuals representing the main developers and operational or research users of the models, as well as individuals representing the ACCESS modelling system and OzEWEX.
The workshop started with overview presentations on the model systems.
- AWAP. Matt Paget (CSIRO) provided an overview of the system, which is no longer being developed but fully transferred into Bureau operations. The presentation clearly demonstrated the strong impact of AWAP in industry as well as national and international research, due in important part to the ease of access to model data.
- AWRA. Russell Crosbie (CSIRO) presented the current status of the model, for which a new release (version 4) will soon bring several conceptual and calibration changes into Bureau operations. The potential use of forecasting was discussed, as well as the potential value of the observations used to calibrate AWRA for the evaluation of CABLE.
- CABLE. Rachel Law (CSIRO) highlighted some areas of current development in the research model version, including more sophisticated descriptions of soil and groundwater processes and satellite data assimilation applications. In discussion the importance of improving the model for global climate applications was emphasised, as many developments in CABLE are tested offline and for regional applications (e.g. within BIOS2).
- Model performance. Guided by presentations from Andrew Frost (Bureau) and Rachel Law (CSIRO), the discussion highlighted that it is important to consider whether the outputs and geographical area of interest were given special consideration in model development and/or calibration. The value of a shared approach to benchmarking was discussed, which highlighted the need to be clear about the meaning benchmarking and its intended purpose.
- OzEWEX. Albert van Dijk (OzEWEX chair) gave an overview of OzEWEX activities that are relevant to the workshop, particularly regards development of the PALS (Protocol for the Analysis of Land Surface models), data assimilation, and (hydrological) forecasting. Global applications of the AWRA system (W3RA) and a new carbon dynamics component in AWRA were also presented.
The subsequent discussion was around 6 themes:
- Multi-model ensembles. It was agreed that these are an increasingly important means to improve estimation and should be the focus of a separate future meeting.
- Forecasting. In particular soil moisture forecasting is of interest to the Bureau and some research is currently in train in this area. Access to Bureau weather forecast data is important for these.
- Benchmarking and Evaluation. From Bureau perspective, the PALS is an appealing infrastructure for AWRA as well as CABLE. Publishing and sharing relevant observations such streamflow, groundwater recharge and flux tower data was considered a priority and will be followed up in several ways.
- Hydrological processes in global models. Better representation of groundwater, river routing and flows to sea in CABLE is important and it was decided to assess whether AWRA expertise can help improve this.
- Community use of AWRA. It was agreed that moving towards a community approach to AWRA model development was desirable, but timing and overheads needed to be thought through. WIRADA management will need to formulate a position on transitioning AWRA-LG to a community model.
- OzEWEx. Bureau reiterated their support for engagement in OzEWEX consistent with the Bureau’s objectives.