Skip to main content

Flood Modelling in Southeast Asia

Climate|Willis Research Network
Climate Risk and Resilience

By Nalan Senol Cabi | May 4, 2021

Join the WRN conference on May 18th - 20th, 2021 to tune in the flood panel session where we will discuss changing landscape of flood risk with panellists from academia, government and the industry.

Indonesia is prone to various natural disasters. Although earthquake, volcano and tsunami are the most dangerous extreme events, and potentially cause the largest insurance losses, severe flooding also gives extreme as well as regular attritional losses, meaning that it causes frequent small to medium flood impacts throughout an annual insurance contract, which accumulate to a significant annual total damage. Willis has been providing the market with flood related analytical services for the region since 2010, having identified a need for an updated and comprehensive view of Indonesia flood risk. As this work has progressed, the latest offering to the market from Willis Re is a brand-new countrywide model, developed in collaboration with internationally recognised experts via the Willis Research Network (WRN) partners at the National University of Singapore (NUS). In line with the latest scientific developments, the project team consisting of academic experts and internal experts, created a fully probabilistic, high-resolution model, which provides a robust view of risk and is validated against significant events.

Modelling Extremes in Southeast Asia: Flood Risk

Flood hazard affects all territories in Southeast Asia. There were devastating floods in Thailand a decade ago in 2011, China flood in 2016 and the catastrophic floods in Jakarta, Indonesia in 2007.

More than a dozen of major floods have impacted Indonesia since the beginning of 20211. In January this year, 40 people were killed in flash floods in the Indonesian town of Sumedang in West Java.

Landslides and flash floods are common in Indonesia during its rainy season. Seasonal downpours spark widespread destruction each year2. About 125 million Indonesians, nearly half of the country’s population, live in areas at risk of landslides and flash floods, according to the country’s disaster agency3.

Willis Re has been working with the NUS Tropical Marine Science Institute (TMSI) since shortly after the devastating floods in Jakarta, Indonesia in 2007. NUS, one of the longest continuous academic partners of the WRN, has significant experience in developing hydrological and hydraulic models for evaluating flood risk in South East Asia, particularly in Indonesia.

The economic centre of Indonesia, Jakarta, is one of the fastest-sinking capitals in the world where some areas are already below sea level4 and was severely impacted during the 2007 floods. Many of our clients suffered large losses from the 2007 event which was unmodelled at the time. Significant insured losses in 2007 came from industrial risks in the very low-lying northern area of Jakarta. The response of many companies was to move their large manufacturing and processing facilities out of the main city and into Industrial estates in neighbouring areas, with increased levels of protection and lower natural flood risk. However, this was a slow and complex process. Insurance companies, therefore, had a need to assess their risk in Jakarta, and then subsequently in newly developing industrial areas elsewhere in Indonesia.

In September 2020, Willis Re launched a new Indonesia Flood model to provide a comprehensive and robust view of risk for the region. Developed jointly with the WRN partner, NUS, the probabilistic model provides a detailed, countrywide assessment of the flood risk (both fluvial and pluvial); this includes the industrial estates in West Java and captures the impact of differing levels of flood protection in those industrial sites. The model is calibrated on an up-to-date proprietary Indonesia-wide Industrial Exposure Database, using insurance and reinsurance market loss claims.

Willis Re’s Executive Director, Head of Catastrophe Analytics in Singapore, Richard Sanders says “the new flood model released last year has been helping with client risk analysis, filling the much-needed gap in the market.”

Changes in Precipitation Patterns in Southeast Asia

With more than 17,000 islands, including Sumatra, Java, Sulawesi, and parts of Borneo and New Guinea, Indonesia is strategically located between the Indian Ocean and the Pacific Ocean, that could induce extreme wet or dry seasons related to the El Niňo-Southern Oscillation (ENSO), an atmosphere and ocean-related climate phenomena that could trigger flooding, drought and forest fires5.

In 2019, Indonesia experienced its worst drought since 2015. Virtually the entire country – 92 per cent – impacted by drought due to the El Niño phase of ENSO at the end of 20186. The country’s National Meteorological Service, the Badan Meteorologi, Klimatologi, dan Geofisika (BKMG), said conditions were being made more extreme by the strongest Indian Ocean Dipole – another driver of climate variability that is represented by the difference in sea temperatures across the ocean – in a century, which can cause drier weather in south-east Asia and Australia7, 8.

Dr Shie-Yui Liong, consultant at NUS TMSI, mentions that understanding the spatial and temporal variability of rainfall has always been a great challenge, and the impacts of climate change further complicate this issue9. The swing between the excessive precipitation and prolonged dry spells, emphasizes the urgency for better understanding of the precipitation patterns.

In their recent study, TMSI scientists evaluated NASA’s Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) dataset over Southeast Asia for present and future climates. According to their findings the future climate projections indicate that average surface temperatures over Southeast Asia are likely to increase by more than 3.5 °C by the end of the century. As for precipitation, both the mean and extreme rainfall are likely to increase but the biases in the historical simulations could contribute to larger uncertainties in the estimates of rainfall projections10.

The NUS TMSI specializes in climate research and has extensive experience, particularly in high-resolution dynamical downscaling of global climate models at a range of spatial scales – from regional (10-20 km) to local (urban) (400 m – 2 km), that are applicable for a suite of impact studies. By utilizing a multi-disciplinary combination of climate science and flood expertise, TMSI and WRN are exploring the influences of climate change on extreme flood and drought that could form the basis of scenarios and future event sets.

Dynamical downscaling of multiple climate models for multiple emission scenarios in Southeast Asia domain gives a range of projected changes in precipitation and drought indices. Together with gridded observation data, projection outputs provide the much-needed insight on drought and flood risk quantification for plausible future climates. As Willis Re’s Head of Analytics, Asia Pacific region, Tai Hui Yen says, “this will help insurers develop innovative solutions with relevant contingency planning that can better prepare and protect businesses and communities, increasing resilience to catastrophic flooding”.

In a changing world, being equipped with functional risk solutions to assess and quantify catastrophe risk is paramount. The WRN bridges the gap between academia and the industry and provides the scientific basis for understanding the strengths and weaknesses of industry tools. By working with our academic partners, Willis Re apply the latest scientific knowledge on developing new solutions to support risk management decisions that influence capital management within the insurance lifecycle.













Head of Flood Risks Research

Related content tags, list of links Climate Research Network Insurance Climate Change
Contact Us