Covering All Bases: Modeling Typhoon and Non-Typhoon Driven Flood in the Philippines

In a country that is used to the regular rhythm of typhoon seasons, 2017 disrupted the pattern and was a surprisingly quiet year in terms of landfalling typhoons in the Philippines. While 26 named storms formed in the western North Pacific basin, equaling the long-term average, all other tropical cyclone statistics fell below the 1981-2010 average. Only 12 of these named storms developed into typhoons, and just four reached a strength of category 3 or above on the Saffir-Simpson scale. No typhoons made landfall in the Philippines during the year for only the fifth time in recorded history.

Tropical cyclones can form in the western North Pacific basin throughout the year, but activity usually peaks between May and October. In 2017, the first named storm, Muifa, developed on April 25, but it took until July 23 for Noru to become the first storm to reach typhoon intensity — the second latest occurrence on record for the formation of the first typhoon of the year. After this sluggish start, tropical cyclone activity in the western North Pacific remained suppressed throughout the remainder of 2017.

Compared to a very active North Atlantic tropical cyclone season in 2017, for only the fifth time since 1950, the total accumulated cyclone energy (ACE) was higher in the North Atlantic than the western North Pacific basin.

Two factors underlined this suppressed tropical cyclone activity in the western North Pacific during 2017; a dominant pattern of anomalous convergent flow aloft over the western North Pacific basin both suppressed the formation of tropical cyclones and prevented them from intensifying into strong typhoons. Second, the El Niño Southern Oscillation (ENSO) began the year in a cool-neutral state, trending towards a weak La Niña into the latter part of the year.

Weaker Storms Deliver Heavy Rainfall

When managing all causes of typhoon and flood related risk, even in a uniquely quiet year for the Philippines, the impacts from weaker storms and non-typhoon flooding can be considerable. 2017 saw several weaker tropical storms that brought heavy rain and flooding, causing significant damage.

Tropical storms that are relatively weak in terms of wind speed often bring sudden and heavy precipitation that can lead to flooding. In addition, the southwest Habagat monsoon from May to October, and northeast Amihan monsoon from November to April can bring heavy rainfall and flooding, with non-typhoon flooding accounting for 20 percent of modeled average annual loss (AAL). The topography and geography of the Philippines means that rivers tend to be short and steep, so that they respond quickly to heavy precipitation, while expansive urban areas with impermeable surfaces, such as Metro Manila, inhibit the ability of water to drain away.

In such tropical storm events, rainfall-induced flooding causes relatively more damage and loss than wind (Figure 2 below). This outcome is reflected by the modeled relative contribution of tropical cyclone flooding to modeled average annual loss (AAL). For storms of category 2 and below, flood contributes much higher modeled AAL for the RMS industry exposure database (IED) in the Philippines than wind.

These types of events can bring heavy rainfall but generally do not produce winds strong enough to cause significant damage to insured property in the Philippines, which tends to be constructed to a reasonable standard to mitigate against damage from both typhoons and earthquakes. Conversely, for category 4 and 5 typhoons, the modeled wind contribution to AAL is higher than flood, as these types of events possess very high strength winds that can cause significant damage over large areas.

This combination of typhoon, non-typhoon storms and monsoon rains means that (re)insurers need a modeling approach that can capture the breadth of the full risk. RMS is releasing the RMS® Philippines Typhoon and Inland Flood Model with typhoon-induced flooding and flood events from monsoon and seasonal rainfall captured by precipitation models, and this allows for spatial correlations in flooding. The flood model also accounts for local flood defenses and urban drainage systems.

Coastal risk is also significant as the Philippines comprises thousands of islands and possesses the fifth longest coastline of any country in the world. Coastal flood risk is assessed using a fully hydrodynamic storm surge model, driven by the wind field of each stochastic typhoon event. The storm surge model covers the main areas at risk in the Philippines, including western Luzon and the Visayas islands, as well as the area surrounding Laguna de Bay.

The model includes a stochastic event set of over 100,000 simulated tropical cyclone events and 82,000 non-typhoon related inland flood events, to capture the risk from typhoon winds and all sources of flooding. The stochastic tropical cyclones represent a distribution of different storm strengths, including events with tropical storm strength winds up to category 5 typhoons. This means that the risk can be analyzed not only for strong typhoons, like Haiyan (Yolanda) or Bopha (Pablo), but also for weaker tropical storms such as those experienced in the Philippines during 2017.

2018 marks the fifth anniversary of Typhoon Haiyan (Yolanda), which on striking the Philippine province of Eastern Samar in November 2013 became the strongest typhoon ever recorded to make landfall. The storm caused devastation across the Central Visayas islands and led to the biggest typhoon loss recorded in the Philippines. This reinforces that although 2017 was a record-equaling quiet year for the Philippines, the potential for intense typhoons to strike the archipelago remains.