Numerous studies have employed climate models to directly simulate the formation and intensification of tropical cyclones. Since the early work of Manabe et al. (1970), the ability of climate models to generate lows that resemble tropical cyclones has developed considerably. Currently, numerous groups worldwide are developing a capability to perform these types of simulations. A recent review is given in Walsh (2008).
Climate models have varying abilities to simulate tropical cyclone characteristics. In general, though, they usually do not simulate numbers that are very close to observed formation rates. One difficulty, as detailed in Walsh et al. (2007), is that the storm detection schemes used to determine the rate of formation within the models are often tuned to the observed formation rate, thus making it impossible to determine the actual ability of the model to generate tropical cyclones in the current amount.
The situation is even worse for intensities, with climate models having an inadequate simulation of the observed cyclone intensity distribution, mostly simulating tropical cyclones that are considerably weaker than observed. Thus simulating observed intensity trends as part of a model-based attribution study would be problematic.
Nevertheless, in general agreement with the earlier predictions of MPI theory, direct simulations of the effect of global warming on tropical cyclones suggest intensity increases of 5-10% by some time after 2050 (Knutson and Tuleya 1999; Walsh and Ryan 2000; Knutson et al. 2001; Knutson and Tuleya 2004).
Thus at present direct simulation as a tool for detection and attribution studies is in its infancy. Recently, though, Knutson et al. (2007) showed that a modeling system could reproduce the observed trend in Atlantic tropical cyclone numbers over the period 1980-2005. Emanuel et al. (2008) use a downscaling methodology employing a synthetic track generator to produce climatologies of tropical cyclones from climate model output. Similar modeling systems have the potential to elucidate the causes of the increase in numbers in the Atlantic by performing attribution experiments that change aspects of the simulation and examine their effects on simulated formation rates.
Was this article helpful?