Downscaling of physical risks for climate scenario design

• Main Authors: BIFFIS, Enrico, WANG, Shuai
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: climate change; climate models; climate scenarios; machine learning techniques; Southeast Asia; Asian Studies; Environmental Sciences; Environmental Studies
• Online Access: https://ink.library.smu.edu.sg/skbi/18; https://ink.library.smu.edu.sg/context/skbi/article/1017/viewcontent/Downscaling_of_Physical_Risks_for_Climate_Scenario_Design_White_Paper_Final.pdf
• Institution: Singapore Management University

Southeast Asia is arguably one of the areas most vulnerable to natural disasters due to its dense population, coastal urbanization, and rainfall variability driven by the local monsoon systems. In this report, we focus on the impact of global warming in the region along four climate dimensions: temperature, precipitation, wind speed and coastal surge. The latter represents the surge of water from the ocean in excess of astronomical tides. Our objective is to downscale the outputs of global climate models to temporal and spatial resolutions of interest to market participants wishing to quantify climate risk vulnerability via climate stress testing exercisestruly representative of their exposures at location. Throughout our study, we consider the representative concentration pathway 8.5 until 2050 (referred to as RCP8.5-2050 henceforth), which is widely considered as a high emission scenario and often regarded as capturing the potential failure of the international community to coordinate to implement effective climate risk mitigation policies. As such our focus is broadly consistent with climate scenarios variously referred to as Business As Usual(BAU), No Additional Policy Action (NAPA), or Hot House3 . Our downscaling approach uses machine learning techniques and places emphasis on the entire distribution of climate variables rather than just the evolution of their average/median level along a climate scenario. We are therefore able to discuss the impact of global warming on both the average level of climate variables of interest and the tail of their distributions. This allows us to disentangle systematic shifts in risk profiles from increases in the frequency and severity of extreme events. We summarize some of the main findings of the report in Figures A to C.