Keywords:Extreme Event Attribution, Climate Change
The rainfall extremes and strong winds associated with tropical cyclones lead to significant damages and lost to where they make landfalling. Upward trend in term of financial lost was indicated for the past few decades from the report of major reinsurance firms. Whether the past anthropogenic warming played a significant role in such extreme event and their changes remained very controversial. On one hand, people argue it’s nearly impossible to attribute an individual extreme event to global warming. On the other hand, the increase of heavy rainfall is consistent with the expected effects of climate change on tropical cyclone. To diagnose possible anthropogenic contributions to the odds of heavy rainfall associated with tropical cyclone, we adapt an existing event attribution framework of modeling a ‘world that was’ and comparing it to a modeled ‘world that might have been’ for that same time but for the absence of historical anthropogenic drivers of climate. The analysis was applied to Typhoon Morakot (2009) as an example. There was more than 2000 mm rainfall occurred over southern Taiwan when a category 1 Typhoon Morakot pass through Taiwan in early August 2009. Entire village and hundred of people were buried by massive mudslides induced by record-breaking precipitation. One limitation for applying such approach to high-impact weather system is that it will require models capable of capturing the essential processes lead to the studied extremes. Using a cloud system resolving model that can properly simulate the complicated interactions between tropical cyclone, large-scale background, topography, we first perform the ensemble ‘world that was’ simulations forced by the high resolution ECMWF YOTC analysis. We then re-simulate, having adjusted the analysis to ‘world that might have been conditions’ by removing the regional atmospheric and oceanic forcing due to human influences estimated from the CMIP5 model ensemble mean conditions between all forcing and natural forcing only historical runs. Thus our findings are highly conditional on the driving analysis and adjustments therein, but the setup allows us to elucidate possible contribution of anthropogenic forcing to changes in the likelihood of heavy rainfall associated tropical cyclone.