Thursday, January 26, 2017

Global Hail—Australia

Storms that produce large hailstones (greater than 2cm in diameter) occur often in Australia, most commonly in New South Wales and the Australian Capital Territory.
Losses from severe convective storms (tornadoes, hail, straight-line winds, and lightning) make up about 50% of all loss ­producing natural hazard events that occur in Australia. Hence, severe convective storm is the number one peril in national catastrophe (natcat) business in Australia.
In 1999, Australia experienced one of the worst hailstorms in the world. The storm single-handedly damaged 24,000 homes and 70,000 automobiles, creating an insured loss of $1.04 billion (U.S. dollars). The hailstones in the April 4, 1999 hailstorm were one of the largest ever recorded measuring 3.5 inches in diameter.
In 2011, another catastrophic hailstorm hit Melbourne resulting in extensive property damage. Within a year of the 2011 storm, Australia endured subsequent natural disasters, and the damage wrought by this series of events spurred double-digit premium rises by some of the largest Australian insurers.
In 2014, a super cell storm of fierce winds and hail caused over $1.1 billion in damage to cars and homes alone. One year later, the Insurance Council of Australia reported that insurance losses from claims following 2015 spring hailstorms totaled about $1.208 billion with almost 120,000 insurance claims.
Similarly, a series of severe hailstorms devastated Australian farms in November 2016, resulting in the filing of over 28,000 insurance claims, totaling $193 million.
Most Australian insurers provide coverage for weather-related damage if it falls into one of two categories—thunderstorms (hail, flash flooding, wind, lightning) or land gales (gale-force winds that occur over land). Despite the fact that coverage is generally afforded for hail-caused damage, Australia has experienced an unprecedented rise in fraudulent insurance claims.
To address these issues, the Insurance Council of Australia created The Insurance Fraud Bureau of Australia to drive down the impact of insurance fraud.  Moreover, the Insurance Council of Australia encourages policyholders to design and construct buildings to be resilient to the hazards present in the environment as minimum building standards in Australia are generally insufficient to protect against severe convective storms. 

Tuesday, January 24, 2017

What’s in a (Storm) Name?

Since 2012, The Weather Channel (“TWC”) has been naming winter storms according to its own internal procedures—or, as some would see it, internal whims. TWC’s criteria for naming winter storms was first implemented for the 2013-2014 winter season, and is based on “the population and area that is forecast to be impacted by winter weather based on thresholds set by the National Weather Service for winter weather warnings”; however, TWC’s three-person storm-naming committee may override these criteria when, for example, the event is particularly historic or unusual. According to TWC’s announcement for the 2016-2017 season, the names will be “used in alphabetical order to identify winter storms that meet naming criteria.” The exciting list of storm names, including Pluto, Quid, Reggie and Stella, has been contributed to “once again” by “the Bozeman, Montana High School Latin class.”
TWC’s practice is controversial. On one hand, as opined by the editors of the Bulletin of the American Meteorological Society (which TWC linked to its 2016-2017 storm-names page), naming snowstorms is helpful for the efficient dissemination of information to those who might otherwise be unaware of a potential storm, particularly in the era of Twitter hashtags. Indeed, using names as a way to “avoid confusion and streamline communications” is a reason the National Hurricane Center of the National Oceanic and Atmospheric Administration (NOAA) began naming tropical storms some years ago. On the other hand, TWC’s storm-naming practices have come under attack—or, at least, eye-rolling—from those both inside and outside of commercial meteorology. When TWC first decided to officially take on the yoke of naming winter storms, rivals took to the internet to opine that TWC’s practice was for commercial purposes.
Another problem with TWC’s unilateral practice of storm-naming is that, under its naming-rubric, just about any grade of storm could be a “named storm,” since its test links the storm to populations, and also takes advantage of the fact that the National Weather Service’s decision to issue a storm warning varies from region to region. As TWC itself explains, “a [National Weather Service] winter storm warning is issued for Atlanta when 2 inches of snow is expected in 24 hours, but it takes 9 inches to trigger a winter storm warning in Burlington, Vermont.”  Correct: storm warnings are issued by the NWS from local offices, based on “local criteria.”
The extreme looseness of TWC’s formula was evident in its most recently named Winter Storm Kori, which brought rain and snow to the west coast and mostly rain and snow to the northeast. Thus, snow isn’t even a criterion for a TWC winter storm—thanks to TWC’s flexible formula and, one would presume, global warming. Likewise, recent Winter Storm Helena travelled coast to coast, with snow resulting in some regions, sleet, rain and/or ice in others.
Certainly, the practice of naming tropical storms has had a long history. Currently, however, the practice follows strict parameters based on wind speeds. “Tropical storms” that are named by the National Hurricane Center are those that reach speeds of at least 17 m/s (past 33 m/s and they are classified as hurricanes). Most commentators, including those in the insurance industry, would say that the verifiable, measurable classification of these storms makes them valuable benchmarks. In other words, if the storm is named, it has reached a certain level of power.
Still, like it or not, the practice of naming various types of storm events—under varying rubrics and schemes—continues to expand, including expansion to government entities. In 2015, for example, the UK’s National Meteorological Service and the Irish Meteorological Service began naming storms, with the stated purpose of aiding communication to the public. The UK’s criteria are based on its National Severe Weather Warnings alerts service; storm systems may be named “on the basis of impacts from wind but also include the impacts of rain and snow.”
The expansion in storm naming is a phenomenon worth thinking about from an insurance perspective. Will the practice of naming storms eventually cause homeowners to stop paying attention to the weather, the result of “crying wolf syndrome” on the part of the weather vendors and government entities? Or, will naming storms cause homeowners to be on heightened alert for storm-related home damage, a reaction that could result in an increase in both legitimate and illegitimate claims? 
Naming storms could also lead to confusion over what a “named storm” means in the policy context. Typically, property policies use the term “named storm” to refer to storm systems that have been named by the National Hurricane Center or the National Weather Service. In the event of a “named storm,” a named storm deducible may apply, oftentimes a percent of the total insured value of the covered property. A named storm deducible is usually a grade below a hurricane deductible (also named by NWS/NHS) in terms of scope, and a grade above a wind or hail deductible.  To the extent a policy extends windstorm coverage (where it would otherwise be excluded), the policy definition of a windstorm may be based on official “named storm” criteria as well.
While typical policy language is clear that named storms are those that are officially named by NHC or NWS, it is nevertheless a bit confusing that an inland “named” winter storm, that only involves wind and rain, is not actually a true “named storm.” The trend in naming anything and everything might particularly confuse policyholders in the southern US, who are subject to tropical storms and yet are also exposed to “winter storms” that are not always snowy but are oftentimes wind-driven rain events. It may be important to recognize this potential confusion and address the issue with policyholders at the onset of any claim that may arise from one of these TWC “named storms.”
Posted by Megan Shutte

Wednesday, January 18, 2017

Can You Predict Earthquakes? Well, Sort Of

Followers of the CAT-Law Navigator know that our blog posts often address recent earthquakes around the world. As catastrophes go, a massive earthquake is about as bad as it gets so it’s a natural topic for us. And if earthquakes are of interest to you, we recommend you check out Temblor. Temblor refers to itself as “a tech company providing a personal, immediate and credible source of seismic risk understanding and solutions for everyone.” We find Temblor to be a great source of information on earthquakes and earthquake-related news and analysis.

With all of the work the folks at Temblor and others are doing to better understand earthquakes and seismic risks worldwide, can we now predict earthquakes like meteorologists predict storms? Unfortunately, no, but Temblor’s website does reflect a few interesting “predictions” related to earthquake activity. First, you can use the website to find all earthquakes worldwide in the last 30 days, or 7 days, or 24 hours. And you can tailor your search based on the magnitude you are interested in (greater than 5.0, greater than 6.0 or greater than 7.0). And guess what? In the last 24 hours alone, there have been seven earthquakes at or above a magnitude 5.0 (four of which were in Central Italy – and Temblor has closely followed the uptick in seismic activity in that region of late). So while we can’t predict or forecast the next earthquake, it is apparent from Temblor’s data that quakes are happening every day.

You can also use your location on the Temblor site and review your current seismic risk on a scale of 1 to 100, and get a sense of the likely magnitude of property damage to your home in the next 30 years. Temblor even tries to forecast the likelihood of you receiving an insurance payout after an earthquake(!). But as Temblor notes, “we are not” predicting earthquakes; instead, Temblor is forecasting your risk “based on the long term behavior of faults, using the best public scientific data available.”

Temblor does, however, identify one type of earthquake that absolutely, positively CAN be predicted. On certain Sundays in September, October, November, December, and sometimes January, there will be small magnitude earthquakes in Seattle. How does Temblor know this? It all started on January 8, 2011, when Marshawn “Beast Mode” Lynch electrified a stadium full of Seattle Seahawks football fans with a powerful 67-yard touchdown run. As Temblor describes it, “the combination of fans jumping, stands reverberating, and ground vibrating was enough to set off a local seismometer, which registered the shaking as a M=2 earthquake. While not substantial, this truly was a man-made earthquake.” In the years since that wild run by Lynch, seismometers have been installed in and around the stadium to track the seismic activity on game days. And when the Seahawks are rolling, you can predict with absolute certainty that there will be a (very small) earthquake.

Posted by Dan Millea

Wednesday, January 4, 2017

Super-Typhoon Nock-Ten: Late-Season Abnormality or Harbinger of Larger Changes?

An unusual December super-typhoon made several landfalls in the Philippines on Christmas Day 2016, bringing heavy rainfall, fierce wind gusts, and dangerous flooding to the island nation.  While the number of fatalities associated with Super-Typhoon Nock-Ten—or Nina, as the storm has become to be known in the Philippines—remains relatively low, the physical damages are currently estimated to be in excess of USD 100 million. 

Treacherous typhoons are not an uncommon threat to the Philippines, which are located in the Western Pacific Ocean.  In fact, in 2016, the region was ravaged by several powerful super-typhoons, equal to Category-5 hurricanes in the Atlantic basin.  For example, super-typhoon Haima—or, Lawin, as it is known in the Philippines— reached maximum sustained wind speeds in excess of 160 mph.  Haima made landfall in the Philippines on October 19, 2016, causing damages of approximately USD 150 million. 

However, what distinguished Super-Typhoon Nock-Ten is the speed of its formation, its subsequent rapid intensification, and, in particular, its late timing.  In fact, because of these factors Nock-Ten has been awarded the status of “strongest tropical cyclone to be recorded worldwide on Christmas Day in over half a century.” 

The record-breaking storm developed extremely rapidly, reaching maximum sustained winds of 160 mph within a short timeframe of five days.  On December 20, 2016, meteorologists detected a tropical disturbance near Yap in the Western Pacific Ocean, over 1,100 miles east of the Philippines.  By December 23, 2016, the tropical disturbance had intensified to a typhoon, with maximum sustained winds of 115 mph, the equivalent of Category-4 hurricane in the Atlantic basin.  Benefiting from ideal conditions along its path with a low vertical wind shear and warm ocean waters, Nock-Ten then swiftly transformed into a super-typhoon by Christmas Day, with wind speeds equivalent to a Category-5 hurricane. 

Even though tropical cyclones can form in the Western Pacific region all year round, it is unusual for a typhoon of this strength to form late in December.  Typhoons, like hurricanes, require ideal conditions, including surface water temperatures of at least 26.5 degrees Celsius, to form and strengthen.  With cooler sea surface temperatures in the Northern Hemisphere during the winter months, tropical cyclones in the Atlantic and Western Pacific become less likely.  In fact, according to the government weather agency of the Philippines, only seven typhoons struck the country on Christmas Day in the past 65 years

Similarly, in the Atlantic basin, hurricane formation generally peaks in August and September, before dropping sharply in October and November as water temperatures fall.  However, late-season hurricanes do occur occasionally, with devastating effects.  For instance, Hurricane Wilma formed on October 16, 2005 and made landfall in Southern Florida on October 24, 2005.  Over sixty people died in the United States alone in accidents related to this powerful Category-5 cyclone, and damage to property exceeded USD 20 billion. 

Unusual late-season tropical cyclones, such as Super-Typhoon Nock-Ten or Hurricane Wilma, may pose serious concerns for the insurance industry.  Catastrophe models, which are nowadays regularly used in the industry to assess the probability of the occurrence of catastrophic events, depend, inter alia, on historic data.  Where the data is of poor quality or unavailable, it may be difficult to generate reliable models to accurately predict late-season tropical cyclones.  Certain developing regions or countries that are prone to tropical cyclones may not have the tools to adequately record and analyze the data associated with late-season tropical cyclones.

In addition, catastrophe models depend on the correct analysis of the data that is fed into them.  For example, with regard to late-season cyclones, the question may arise to what extent global warming and rising water temperatures contribute to their formation, if at all?  Do other factors, such as, for instance, changes in the atmosphere, explain the strength of recent late-season cyclones?  That is to say, is Super-Typhoon Nock-Ten the proverbial “odd one out” or is it the harbinger of larger changes? 

Lastly, late-season tropical cyclones may impact certain lines of business in the insurance industry more than others.  For example, the impending threat of Hurricane Wilma in late October 2005 forced the PGA Tour to cancel the Miccosukee Championship in Miami for safety reasons.  Several other sporting events were eventually postponed or rescheduled, causing quite a headache to event organizers and attendees alike.  Likewise, Super-Typhoon Nock-Ten impacted Christmas celebrations throughout the Philippines and left thousands of travelers stranded

In sum, the more data on late-season cyclones is recorded and analyzed, the better the insurance industry is equipped to understand these phenomena and their causes and to brace for the future.