Preliminary climatology and improved modelling of South Indian Ocean and southern ocean mid-latitude cyclones

The intense mid-latitude cyclones that traverse the southern waters of the Indian Ocean, between South Africa and southwestern Western Australia, are among the strongest depressions found anywhere in the world, outside tropical waters. Near-surface winds that exceed storm force (i.e. 24 m/s or 48 knots), and central pressures of 960 hPa and lower, are relatively common for these systems. They pose a constant threat to both open ocean and coastal shipping, and regularly generate severe weather over the populated southwestern corner of Australia. Large ocean waves and swell produce extensive coastal inundation and erosion. There were two main aims in this study. The first aim was to develop a preliminary climatology of these intense mid-latitude cyclones, for the region 20-60 degreesS, 30-130 degreesE. The climatology, which is the first that we are aware of for this notoriously data-sparse region, is based largely upon satellite observations, particularly scatterometer data, and is supplemented by ship, buoy and all available land observations. The climatology revealed that, historically, the frequency and intensity of the mid-latitude cyclones in this domain have been significantly underestimated. This underestimation has resulted in analyses that have serious flaws, and the resultant operational forecasts provided to the duty forecasters in the regional forecast centre located in Perth, Western Australia, are of highly variable quality. A number of other climatological features of these storms are discussed in this article. The second aim was to identify the factors that can contribute to a significant improvement in model forecasts of these storms. So far, there have been very few studies of explosively developing cyclones over this part of the world. Results are presented here from a series of high-resolution numerical simulations of an intense cool season Southern Ocean cyclone that developed in 2003, using the HIRES numerical weather prediction model developed by L.M. Leslie. Here, we examine the sensitivity of the cyclone predictions to both model resolution and the initial analyses. The predicted variables of most interest are the central pressure, maximum sustained near-surface wind speeds, extent of storm-force winds, and the horizontal and vertical structure of the storm. Increased detail in the initial state is provided mainly by the assimilation into the archived global operational analyses of high-resolution satellite-derived data, including QuikSCAT scatterometer winds and sea-surface temperatures. The combination of increased horizontal and vertical model resolution, and improved initial model states, was found to produce numerical forecasts with significantly more accurate wind speeds than those obtained from the coarser resolution operational models, which also did not have the benefits of all the additional data. Finally, areas of future research are outlined, including coupling the HIRES atmospheric model with ocean and wave models, to improve forecasts of the sea state, including wind wave heights, swell and storm surges. Copyright (C) 2004 Royal Meteorological Society.