Theme: Data Assimilation
This Theme aims to pursue successful new approaches for data observation into energy and/or water balance models. The assimilation of a range in situ and satellite observations of atmosphere, ocean and land is a critical factor in the success of weather forecasting. Statistical climate and streamflow predictions also entrain current observations. Retrospective climate analysis (re-analysis) and water balance estimation through the AWRA system also rely on data assimilation techniques. Flood forecasting currently relies on the implicit assimilation of weather and streamflow observations in the mind of the forecaster, but formal data assimilation procedures are being developed. Finally, climate and water scenario prediction models are informed by observations in a range of ways (e.g. for calibration, evaluation), some of which can be considered forms of data assimilation.
Data assimilation is a highly complex challenge, as it requires understanding across a range of areas: observation characteristics and errors; the relationship between observed and modelled quantities; the conceptual structure and equations of the biophysical model; the mathematical assimilation techniques available; and the constraints imposed by the computational software and hardware. As a consequence, the Australian data assimilation community tends to be highly specialised towards specific applications in terms of each of these aspects. While necessary, this has been an obstacle to the exchange of expertise and to the increasing the efficiency of research through collaboration.
This Theme is intended to strengthen and grow the Australian data assimilation community in the area of climate and water modelling. In particular, it provides a forum to exchange expertise and collaborate on topics of common interest in the areas mentioned. It is anticipated that experiments will be designed and carried out to address questions of common interest, such as the error in particular observations (e.g., remotely sensed soil moisture) or the effectiveness and efficiency of alternative assimilation techniques (e.g., Kalman techniques). The experiment may be tiered, e.g., be partly continental and partially focused on the Murrumbidgee catchment.