Evaluating the representation of Australian East Coast Lows in a regional climate model ensemble.

Di Luca, A., J.P. Evans, A. Pepler, L.V. Alexander and D. Argueso
Journal of Southern Hemisphere Earth Systems Science, 66(2), 108-124, 2016.


Due to their large influence on both severe weather and water security along the east coast of Australia, it is increasingly important to understand how East Coast Lows (ECLs) may change over coming decades. Changes in ECLs may occur for a number of reasons including changes in the general atmospheric circulation (e.g. poleward shift of storm tracks) and/or changes in local conditions (e.g. changes in sea surface temperatures). Numerical climate models are the best available tool for studying these changes however, in order to assess future projections, climate model simulations need to be evaluated on how well they represent the historical climatology of ECLs. In this paper, we evaluate the performance of a 15-member ensemble of regional climate model (RCM) simulations to reproduce the clima- tology of cyclones obtained using three high-resolution reanalysis datasets (ERA- Interim, NASA-MERRA and JRA55). The performance of the RCM ensemble is also compared to results obtained from the global datasets that are used to drive the RCM ensemble (four general circulation model simulations and a low resolu- tion reanalysis), to identify whether they offer additional value beyond the driving data. An existing cyclone detection and tracking algorithm is applied to derive a number of ECL characteristics and assess results at a variety of spatial scales. The RCM ensemble offers substantial improvement on the coarse-resolution driving data for most ECL characteristics, with results typically falling within the range of observational uncertainty, instilling confidence for studies of future projections. The study clearly highlights the need to use an ensemble of simulations to obtain reliable projections and a range of possible future changes.

Key Figure

Figure 2 Scatter plots of the number of events Vs. their mean duration (a and b) and the event-mean 200-km SLP gradient Vs. event-mean size (c and d). Top panels show results for the warm season events and bottom panels for the cold season events. Ensemble mean values are shown in red, black and grey for the reanalyses, the RCM and the GDD ensemble respectively. The RCM and GDD ensemble mean values exclude results using the NCEP/NCAR reanalysis data. Only results obtained using the low resolution SLP fields (i.e. at the common 300-km grid mesh) are shown.

UNSW    This page is maintaind by Jason Evans | Last updated 29 November 2013