Monthly Archives: January 2014

How Does Southern Europe Weather the Storm?

The 2013-14 European winter storm season has been pretty active so far. Early in the season, Windstorm Christian raced across northern Europe, followed by Xaver in early December, and then storms Dirk, Erich, Felix and Anne hit the U.K., Ireland, and northwest France over the Christmas and New Year period.

To date the season has been a great demonstration of how northern Europe is a common target for winter storms. However, this week sees the 5th anniversary of Windstorm Klaus, reminding us that storms can also impact southern Europe, affecting regions not acclimatized to extreme winds and causing severe damage.

What happened when Klaus hit and what have we learned from it?

Can such a storm occur again in the near future and more importantly, can we predict it, or at least estimate how bad it could be?

Windstorm Klaus sprung to life on January 23, 2009 in the central Atlantic, directly in line with southern France. The climate backdrop to this storm was pretty uncharacteristic. The large-scale Icelandic low-pressure system and the Azores high-pressure system were farther south than usual. Also, the North Atlantic Oscillation (NAO) was entering a negative phase.

A positive phase of the NAO creates favorable conditions for strong storms to pass over northern Europe, as Lothar and Anatol did in 1999. But a neutral or negative phase of the NAO can lead to storms that affect southern Europe and this is exactly what happened with Windstorm Klaus.

By midnight on January 24, as Klaus approached land, it had a central pressure of 963 hPa, comparable to Windstorm Lothar. Winds reached severe gale force in the southwest of France, peaking with gusts above 140 km/h at coastal locations near Bordeaux, accompanied by violent seas with wave heights of several meters. Local infrastructure was severely disrupted by fallen trees and electricity pylons.

Over 1.7 million households were without power immediately after the storm and over 60% of maritime pines in the Forêt des Landes were destroyed. Once the damage had been appraised, Klaus was estimated to have caused insured losses of €2.5billion (US$3.4 billion).

Shortly after the event, RMS scientists Dr. Navin Peiris and Dr. Christos Mitas conducted a reconnaissance survey, which helped to enhance our understanding of building vulnerability in this region. They observed frequent non-structural wind damage, such as the uplifting of roof tiles and collapsed chimneys, but also direct wind damages from tree fall, due to the high density of trees in close proximity to properties.

Source: RMS 2009 reconnaissance

Closer examination of the roof damage revealed little evidence of proper fixation, particularly along roof edges, leaving them more vulnerable to wind damage. Another observation was the use of canal-type tiles, which are prone to uplift from the build up of air pressure, caused by strong winds. Also, damage was more frequent in residential properties, compared to commercial or industrial buildings that are generally engineered in line with building codes.

This survey, combined with an assessment of claims data, provided us with an enhanced understanding of regional vulnerability differences. For example, we observed a significantly lower fragility of buildings in the Perpignan area compared to the southwest of France.

Ratio of the modeled and observed losses by postcode using non-regionalized vulnerability functions. Variation supports need for distinct vulnerability regions.

Ratio of the modeled and observed losses by postcode using non-regionalized vulnerability functions. Variation supports need for distinct vulnerability regions.

This information is vital for us to continually develop and inform our models, in order to represent the risk accurately. Due to the inherent uncertainty in the climatic phenomena driving windstorms, it is not possible to forecast exactly when the next strong storm will hit southern Europe. Catastrophe models provide a range of possible events, which can help the insurance industry prepare for the next big event.

The RMS Europe Windstorm Model contains storms comparable to Klaus, including some that impart larger wind intensities and damages. The below image compares two examples of stochastic storms with the actual Klaus wind footprint to illustrate storms that could potentially cause insured losses similar to or higher than Klaus.

Klaus and Stochastics

Currently we are in a close to neutral phase of the NAO, so does that mean a Klaus type storm could occur this winter? No one can answer that question for certain, but a model at least enables us to explore the possible worst-case scenarios and be prepared.

Twenty Years After Northridge: a Seismic Blink, or an Eternity?

January 17, 2014 marks the 20th anniversary of the M6.7 Northridge, California Earthquake, which caused the largest U.S. earthquake insurance loss to date. Similar to many of my RMS Californian colleagues, I recall clearly when the event occurred. I was working as a structural engineer in San Francisco, and we in the earthquake engineering community responded to the call to assess building and infrastructure damage.

When the dust finally settled 18 months later, the full picture of damage and loss emerged, and losses were dramatically higher than when first assessed. Direct economic loss reached $40 billion and insured loss totaled $12.5 billion, the majority from commercial and residential lines of business. It was an industry-changing event for the catastrophe modeling and insurance industry, as well as the broader earthquake engineering community.

Lessons Learned

The Northridge Earthquake overturned a number of previous assumptions on earthquake science and engineering. The earthquake occurred on a previously unrecognized fault—a blind thrust fault—beneath the highly populated San Fernando Valley. This triggered urgent research to identify other blind thrust faults underlying the Los Angeles Basin and the San Fernando Valley, as well as other areas of California.

These hidden faults were incorporated into the 2002 version of the U.S. National Seismic Hazard Maps. The shaking also revealed new information around the vulnerability of certain classes of engineered buildings, in particular unanticipated welding failures at the beam-column connections of steel moment-resisting frame structures (SMRFs). This damage initiated the multi-year and multi-disciplinary SAC steel project, which investigated various repair techniques and new design approaches to minimize damage to SMRFs in future earthquakes.

The earthquake also overturned a number of assumptions on earthquake insurance risk management in California. Up until this point, the standard method of managing the risk was to employ a “probable maximum loss” approach that focused on the accumulation of insured exposure across the whole Los Angeles Basin.

Within the accumulation zone, the dominant risk was assumed to come from commercial structures. However, the majority of insured loss from the Northridge event actually came from residential properties—although residential damage was not extreme, the repairs proved very costly.

When the magnitude of the residential losses prompted insurers to restrict sales of homeowners coverage, the California legislature established the California Earthquake Authority (CEA), a publicly managed, privately financed entity, to provide coverage. In 1996, the CEA opened its doors for business and residential insurers could opt into the program. The majority of the California homeowners insurance market joined the CEA, and the aggregate share (of 70%) has changed little over time. However, the current take-up rate of residential earthquake coverage in California remains historically low—at approximately 10%. The CEA continues to explore ways to increase the take-up rates, as insurance is a crucial element in the recovery of an area and the reduction in federal aid assistance following a major California earthquake.

Risk Reduction Strategies

To commemorate the 20th anniversary, the earthquake engineering community is gathering in Southern California at the Northridge 20 Symposium. Local community leaders, emergency managers, financial insurance industry professionals, and scientists, and others are coming together not only to celebrate the achievements in risk reduction and management over the past 20 years, but also to determine how to become more resilient to earthquake damage moving forward.

There are various strategies around reducing the impacts of the next Southern California earthquake. These include the development and enforcement of seismic design provisions in building codes, widespread adoption of seismic strengthening techniques (e.g., to unreinforced masonry commercial buildings and older wood frame residential structures), and responsible land-use planning informed by accurate earthquake hazard information.

For example, the California Geological Survey (CGS) recently issued preliminary versions of new Alquist-Priolo Earthquake Fault Zone Maps for the Los Angeles region. Within Alquist-Priolo fault zones, construction of certain structures is prohibited to reduce damage and loss of life in a future earthquake. The new maps place the Hollywood Fault in an Alquist-Priolo zone, with major consequences for the potential redevelopment of Hollywood and in particular the value of some key land parcels.

Twenty years have elapsed since the Northridge Earthquake—and the industry has seen 20 years of relatively loss-free earthquake experience in California. In the most seismically active region of the U.S., a gap of 20 years—though a blink in geologic time—is an eternity for some consumers who increasingly question the cost of their earthquake insurance premiums. While the timing, location, and severity of the next major California earthquake are uncertain, the next major California earthquake is a certainty. We must remain vigilant in building a culture of prevention to reduce the impacts from this inevitable event.

2013/2014 Winter Storm Season in Europe

This winter storm season has seen the most severe sequence of storms since 1990.

Another windstorm to affect the U.K., Windstorm Christina, brought heavy rain, strong winds, and large waves to coastal regions of the U.K. causing further flooding on Monday, January 7. More than 300 properties have flooded to date as a result of Windstorm Christina in England and Wales. The worst affected areas appear to be the coastlines of South England and Wales.

In England, sea defenses were breached at Chesil Beach in Portland, Dorset, as a result of high waves, causing evacuations. Flooding has also affected properties in parts of the Somerset Levels, leaving some villages cut off, with roads and buildings damaged. In Salisbury, Wiltshire, residents were also asked to leave their homes.

Since the peak of the 2013/2014 winter storm season started in December, Northern Europe has been bombarded by 17 low-pressure systems. These weather systems have swept in off the North Atlantic and brought strong wind gusts in excess of 90 mph in remote locations and gusts over 70 mph over large parts of the U.K.

Extreme rainfall levels have been observed across the country, particularly in Scotland where one location observed 73.6 mm of rain on Christmas Eve. The significant precipitation contributed to December seeing 154% of the U.K. historical average rainfall for that month and in Scotland it was the wettest December since records began in 1910.

This extreme rainfall has left the ground saturated, with groundwater levels in some regions much higher than usual for this time of year, heightening the risk of flooding should the persistent rainfall continue.

At the start of December 2013, windstorm Xaver raced across northern Europe bringing strong winds and coastal flooding to the east of the U.K., northern Germany and the Netherlands. Flood damage was mitigated by improved coastal defenses, which had been upgraded following the devastating 1953 (U.K. & Netherlands) and 1962 (Hamburg, Germany) floods.

Over the Christmas and New Year period, storms Dirk, Erich and Felix brought more strong winds and precipitation, particularly to the U.K., Ireland and northern France. While the winds from these storms caused only moderate damage and disruption to power and transport networks, persistent precipitation was and remains a significant concern for flooding.

The succession of large and intense low pressure weather systems has been a function of a very strong and stable jet stream, fed by the strong contrast between the warm tropical Atlantic air masses and intensely cold Arctic air over North America.

Although this sequence cannot be directly linked with climate change – the coastal flooding will be more common in the future, while winter rainfall totals are expected to rise.

A number of towns along the west coast of England and Wales are relatively unprotected from sea flooding, and consideration will now need to be given to how they can be defended. As sea level rises it will become increasingly and prohibitively expensive to protect all these communities.