partner login

BMS News

California Wildfires October 2019

What Is The Concern For The Insurance Industry?
Like clockwork, it’s wildfire season again in California, and it appears to have the insurance industry on edge for good reason. The last two years have been extremely destructive, with insured losses reaching over $30B and, according to Milliman, wiping out about twice the combined underwriting profits for homeowners insurance for the past 26 years. Applications like Zillow make it easy to understand exposures within fire parameters, and some of the properties’ high insured values are likely eye-opening, undoubtedly worrying the insurance industry.

BMS iVison shows the USGS GeoMac Current Fire Perimeter and can show in-force risks. Other free tools like Zillow allow for cross-reference to understand risk.

However, I want to first say that, so far, the fires of the last month have caused nowhere near the level of destruction to the industry that was experienced in 2017 and 2018. Headlines like the one below are simply causing unnecessary panic. There is a huge difference between the number of structures exposed versus structures that burn.

Overall statistics from the last two California wildfire seasons suggest this year is not even in the same ballpark in terms of loss and structures destroyed.

In fact, of the current wildfires reported by CAL FIRE, only 242 structures have been destroyed, as compared to the thousands that occurred in 2017 and 2018. I guarantee that this number will climb as the assessment continues. A large fire that wipes out thousands of homes could, of course, still occur this season. But, in my opinion, the industry needs to take a deep breath and trust that the last week has been a buildup to something less extreme at the start of November. Unfortunately, the true end to the fire season is a long time from now, and I don’t see any rain in the forecast for at least the next 10 days. Lastly, we can’t underestimate the amazing work that firefighting crews are doing to try and contain these fires before any more serious damage occurs. For example, the Misty Fire near Oroville, CA was quickly attended to yesterday with air and ground support minutes after it started. These efforts limit insured loss.

New Normal of Forced Power Outages
Although the insured loss level has been relatively low, there are many issues for the insurance industry to consider. The power outages have been massive. In fact, PowerOutage.US confirmed that this is likely the largest forced outage in U.S. history, with close to 3 million people without power at one point. At a time when we are more dependent on electricity than ever – ATMs can’t operate, restaurants are closed and businesses try to limp along on backup generators – there is no doubt that these unprecedented outages are causing huge economic losses. I have seen reports that the economic costs are anywhere from $400M up to $3B, with the caveat that economic losses are complex to understand. Another important aspect of determining insurance loss is that California policies may include a “Loss of Use” provision, which would cover for hotel stays during evacuations. How many insureds are utilizing such coverage is unknown at this time, however. Regardless, this might mean an uptick in such coverage going forward. There is clearly business interruption that could also lead to increased losses.

The power outages could also cause an even greater insured loss because they complicate evacuation. Without lights to find the things they need, along with street lights and stoplights to guide them out of the evacuation zone, evacuees are helplessly gathering their belongings by flashlight and trying to leave via streets that are at a higher risk for accidents due to no traffic lights. One accident can often shutter some of the only exits out of these towns, creating the potential for an even larger loss of life from fast-moving wildfires. First responders are also challenged by the power outages.

Another issue with a lack of power is increased crime. Without video cameras or street lights to hinder criminal activity, the opportunity for insured losses due to crime is escalated. Further, a lack of power can lead to loss if towns can’t pump water. Clearly, this creates problems when trying to fight fires, but it also causes dangerous situations for businesses and residents not having access to water. Evacuation plans and supplies are critical to businesses and residents in locations that have a high earthquake likelihood in addition to wildfire potential.

Is No Trash Pick-Up Next?
Much of the focus has been on the problems endured by utility companies, but other industries are facing huge liability risks as well. A few weeks ago in southern California, a garbage truck caught fire and the driver dumped the burning trash on the side of the road (protocol to do so to save the truck). Minutes later, powerful winds blew the flames across a hillside and into a mobile home community, killing two people and destroying dozens of homes. This is another example of unforeseen risk that may be addressed by limiting trash pick-up during extreme conditions.

Evacuation Fatigue
The vast majority of evacuees will not be affected by the fires and can return to their homes without consequence. Taking a lesson from hurricane evacuation as we mark the seventh anniversary of Hurricane Sandy 2012, however, is that evacuation fatigue is real. During Sandy, many people didn’t leave because, in the prior year during Hurricane Irene 2011, calls for evacuation were ordered, but the storm weakened with limited impact to the Northeast. Some of the evacuees chose not to leave the next year only to find them themselves suffering from the impact of Sandy. I am sure evacuation fatigue is already setting in for many California residents, and unfortunately, this could cause more problems in the future as people wait longer to leave, causing major traffic jams. The fallout could be deadly.

Season Is Not Done Yet
In summary, even though it currently appears that the insured loss will be limited, the overall risk will remain high until, hopefully, a weather system around the middle of November will bring some much-needed precipitation to the west coast of the U.S. The early December fires of 2017 that totaled over $2.5B in insured losses serve as a reminder that the California wildfire season is not yet close to being done. With any luck, however, the recent events will have been the peak of the season’s activity. The insurance industry and California residents are on edge, and the forced power outages could get worse – even with massive outages, two of the major fires were caused by utility lines that were not de-energized. These outages are having large economic impacts and, to some degree, will be felt by the insurance industry going forward. On a positive note, however, they could also provide an opportunity for the development of new products to help mitigate the loss of power without physical loss to a structure.

BMS Tropical Update Oct 17th 10:00 am CDT

With each passing day in October, the chance of a U.S. landfalling named storm decreases.  In fact, since 1900, only 16 named storms have made landfall in the U.S. after October 17, but that also means that a large loss to the insurance industry is still possible.  In fact, the late fall can have some devastating storms, such as Hurricane Sandy.  While Sandy was not technically a hurricane at landfall, it still caused close to $20 Billion of insured loss.  Some storms can form rather rapidly at this time of year.  It took Hurricane Michael (October 7-10, 2018) just three days to go from a tropical depression to a Category 5 hurricane.

Historically, the majority of landfalls occurring late in the season are along the Gulf Coast, with the overwhelming majority of these being along the Florida Gulf Coast.  Some of these landfalling storms have been significant, such as Hurricane Wilma in 2005, but none of the long-range forecast models (out to October 27th) currently suggest any type of tropical trouble like Wilma, which originated from an African wave.  Tropical development from African waves become rare once the season passes the middle of October, as the Intertropical Convergence Zone (ITCZ) shifts south towards the equator and overall wind shear begins to increase across the Atlantic Basin.  This is why Tropical Depression Fifteen quickly lost its organized structure earlier this week and failed to intensify into a tropical storm off the African coastline.

In the late season, tropical troubles often arise from what is called a Central American gyre, and this is currently what the National Hurricane Center (NHC) is watching off the coast of Mexico.  This Central American gyre is a broad area of low pressure over parts of southeastern Mexico and Central America that extends into the Bay of Campeche and the adjacent east Pacific Ocean, and is typical of the late spring and early fall hemispheric weather patterns.  In fact, roughly 50% of Central American gyres have a tropical cyclone associated with them, according to Philippe Papin, a research scientist at the U.S. Naval Research Laboratory. 

As you can see from the tweet above, these Central American gyres tend to form smaller areas of low pressure, and sometimes they form on the eastern side of the gyre and rotate counterclockwise around the larger circulation.  These pieces of energy can detach from the overall gyre circulation and become independent tropical systems. 

For the last several days now, the forecast models have consistently shown that a piece of the Central American gyre will move into the western Gulf of Mexico over the next 72 hours.  If this occurs, the NHC could designate a tropical or subtropical depression or storm once it has a complete, counterclockwise surface circulation with organized thunderstorms nearby.  If it becomes a tropical/subtropical storm, it would be called Nestor.  The NHC is currently forecasting a 90% chance of tropical cyclone development.

The forecast models suggest that the system will quickly track toward the northern Gulf Coast by Saturday, and then may track inland to bring much-needed rain to areas of the southeast.  After all, the last thing the insurance industry needs is a surprise wildfire event across the southeast this fall like the one that occurred in Gatlinburg, TN in 2016.  Also as a reminder, the area where Hurricane Michael made landfall had a large amount of tree fall which will increase the wildfire risk for several years.

The flash drought continues to impact the southeastern U.S. with both above-normal temperatures and below-normal precipitation.  This will drastically change with the forecast of a wet weekend as tropical moisture makes its way across the region.

As the tropical system takes shape, it will stay weak as increasing upper-level winds over the northern Gulf of Mexico should produce enough wind shear to prevent anything of substantial strength to develop.  It’s safe to say that the insurance industry shouldn’t see a repeat of a rapidly strengthening storm like Michael last year. Currently the intensity, track, closely resembles Tropical Storm Josephine in 1996 which was a PCS loss event.  It formed in the west-central Gulf, tracked toward the northeast, and made landfall in Florida’s Big Bend area which is especially sensitive to storm surge.

Above is the model forecasted guidance from the America GFS ensemble model and the various intensity forecast models.  U.S. landfall impacts are expected to occur on Saturday, October 19 around 48 hours from now. Source

Overall, the long-range forecast continues to show a North American weather pattern that is not conducive for tropical named storm development or U.S. landfalls, as the polar jet begins to take over the weather pattern across North America and increases wind shear over much of the Tropical Atlantic Basin.  However, let’s not forget that the season continues until November 30. 

Since 1966:  83% of Atlantic seasons have had one or more named storms; 64% have had one or more hurricanes; and 21% have had one or more major (Category 3+) hurricanes; form after October 15, but the chance of U.S. landfall decreases with every passing day.

BMS Tropical Update Sept 13th 10:00 am CDT

No Rest for the Weary

The Atlantic hurricane season is far from over.  In fact, we are just a few days removed from the peak of the season.  As mentioned in the September 6th BMS Insight, I anticipated the insurance industry would have a solid seven days of rest, but this was in regard to new tropical development.  The insurance industry is grappling with the outcome of Dorian and the under-reported major Category 4 Typhoon Faxai that had a direct hit on Tokyo this past weekend.  The NHC has now classified potential tropical cyclone number nine (PTC9) in the central Bahamas, and the long-range forecast still suggests the end of September and beginning of October could have several named storms develop in the Atlantic Basin – there is no rest for the weary.

Basin Memory

If you have been keeping track of the tropical troubles so far this season, there is clearly a target on areas around eastern Florida and the Bahamas.  Dorian is not the only storm to impact these locations;  Hurricane Chantal’s origins started in this area, along with Tropical Storm Erin.  Tropical Depression 3 also made a brief appearance there as well.

The 2019 Atlantic hurricane season so far.

This is a great example of weather memory or, more commonly known to the catastrophe modeling industry, clustering.  It is a fairly typical occurrence when the atmosphere repeats a particular type of weather pattern.  We see this with floods, severe convective storms, winter storms and with tropical weather events.  There are many reasons why the atmosphere has clustering.  Different modes of atmospheric variability such as the Atlantic Meridional Mode (AMM), El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO) and Madden-Julian Oscillation (MJO) have been shown to affect Atlantic hurricane activity through changes in atmospheric steering currents and vertical wind shear.

In fact, a good understanding of clustering in the Atlantic Basin is found in the paper by Kossin et al. (2010) which shows four distinct clusters of storm activity that are common across the Atlantic Basin. The best recent example of this might be the 2004 Atlantic hurricane season with Frances and Jeanne taking similar tracks and making landfall along the east coast of Florida, basically in the same general area and 21 days apart.  There are many historical examples of clustering, such as Hurricanes Connie and Diane in 1955, which hit the Carolinas just six days apart.  Because of this, some of the catastrophe modeling companies simulate clustering into their event catalogs allowing for a probability of multiple events within a season to impact similar areas of the U.S coastline.

Future Humberto?

For the last several days, there has been a tropical disturbance labeled PTC9 over the central Bahamas which is slowly becoming better organized.  The National Hurricane Center gives it an 80% chance of becoming a tropical depression or storm in the next 48 hours and a 90% chance in the next five days.  Humberto is the next name assigned for the 2019 Atlantic hurricane season.

As mentioned with Dorian, the weather forecast models have a hard time modeling ragged, sloppy, and weak tropical storms.  The current modeling suggests almost every scenario imaginable.  For example, the American Global Forecast System (GFS) keeps the system weak as a tropical depression or tropical storm, which could result in more insurance claims than what occurred with Dorian due to heavy rainfall and locally strong gusty winds reaching the U.S.  As a counterpoint, the European Centre for Medium-Range Weather (ECMWF) indicates PTC9 will develop faster and stronger and holds it off the southeast coast of the U.S.  Given the track record, I would have to buy into the ECMWF model, which makes sense if the system can stay over the warm Gulf Stream between Florida and the Bahamas. It will likely strengthen and mirror the track Dorian followed a few weeks ago.  Similar to Dorian, if PTC9 develops into a hurricane, the strongest winds will remain offshore, but this will depend on its track up the Florida coastline.  The majority of ECMWF ensemble guidance is keen on keeping PTC9 far enough offshore to limit any impact of wind, with the long-range forecast suggesting the system will take a right hook and move out to sea early next week.  However, Florida, Georgia, the Carolinas, and even Bermuda still need to pay close attention to potential damage as, currently, the system is weak and leading to higher model uncertainty.

Above is the ensemble model probability from the GFS and ECMWF  showing two different scenarios.  The GFS is less aggressive and tracks the potential named storm into Florida.  The ECMWF is more aggressive and keeps the potential named storm East of the Florida coastline.   Source: Univerity of Albany Brian Tang

Future Tropical Troubles

PTC9 might not be the only activity into early October as there is a signal that the MJO will aid in a broad-scale rising motion in the upper atmosphere, which will promote named storm development across the Atlantic Basin.  Currently, there is large-scale sinking motion across the main development region which has likely contributed to the lack of activity in this area of the Atlantic.  Multiple waves have come off of Africa, but have failed to develop into a storm since Dorian.  The MJO is a large overturning circulation that propagates east across the globe about every 30 days.  Think of it as a wave, the leading edge of which is “suppressed” and causes sinking air.  In the backside of the wave, the air is more prone to rising, which aids in named storm development.

The areas of brown and green are areas of rising and sinking air.  Notice what happens in the forecast of the MJO – a lot of green across the Atlantic.  That indicates rising air, which can aid in named storm development. Source: NOAA

The first concern for the insurance industry is PTC9, but later next week there could be other tropical troubles starting to take shape from the Gulf of Mexico to the African coastline.

BMS Tropical Update Sept 6th 12 pm CDT

To think, two weeks ago today the NHC first circled a tropical wave 1,400 miles east of the Windward Islands. This tropical wave would become Dorian and result in one of the strongest storms ever recorded globally.  The impact on the Carolinas has largely been limited to coastal areas, and the impact on Atlantic Canada is currently unknown. But, it is nearly certain that the named storm Dorian will never be used again given the historic catastrophic damage it caused to the northern Bahamas earlier this week. 

There are so many things to think about and learn from Dorian, but, to me, the most amazing fact is that one of the strongest Category 5 storms to ever occur in the Atlantic Basin was just 110 miles east of a population of 4.6 million people that were largely allowed to stay put. This speaks volumes to the confidence that exists in the short-term weather forecasting of named storms.  Forecasting can always be improved, but, overall, once Dorian approached the Bahamas, the forecast was fairly accurate in terms of indicating that Dorian would track along the entire southeast coastline.

In the end, Dorian, with all of its possible forecasted landfall locations along the U.S. coastline, finally did make landfall on the last possible piece of land – Cape Hatteras at 8:35 am EDT this morning. I hope that the insurance industry does not soon forget about this storm because of its minimal landfall appearance.  Dorian clearly showed atmospheric conditions indicating the potential for landfall as a 185 mph hurricane with a stall over the greater Miami area, likely creating an insurance industry-changing event.  The industry may want to use Dorian as a stress test by shifting the track ever-so-slightly to the U.S. coastline and modeling scenarios of 60+ inch rainfalls across the state.  Such scenarios would likely be eye-popping when considering what that amount of rain would do to the Lake Okeechobee flood control system, which is already in dire need of upgrades.  These are extreme examples, but Dorian showed that they could happen. The question is will the industry, which is largely adopting private flood insurance across Florida, act on this or dismiss the potential devastation.

While Dorian is still impacting the outer banks of North Carolina this afternoon, conditions will quickly improve today as Dorian starts to race across the North Atlantic with high winds occurring across Martha’s Vineyard, Nantucket and Cape Code before Dorian impacts much of Nova Scotia as a fast-moving Category 1 hurricane. Thus, the insurance industry could still expect some insured loss to occur in Canada. 

As Dorian paralleled the coastline yesterday, power outages increased as winds picked up speed – nearly 190,000 customers were without power in South Carolina last evening.  While the number of power outages has since decreased in South Carolina, they are on the rise in North Carolina, with over 230,000 customers without power as of this morning.

BMS iVision allows clients to run risks to better understand the various impacts of Dorian. This is the three-second wind speed gust in mph from Verisk Weather Solutions. Dorian will now be added to our Property Claims Service historical event archive which will allow clients to stress test portfolios and understand the impact of Dorian for future years of risk management.
Updated listing of peak wind gust reports from Hurricane Dorian across Georgia, South Carolina, North Carolina and Virginia received as of 12:30 pm EDT Friday afternoon. Source: NOAA

Some of the worst damage from Dorian might have come from the 25 tornadoes that impacted the region over the last two days.  

While wind and storm surge have been the big storylines with Dorian, some of the rainfall estimates have been impressive.  Wilmington, NC picked up 8.58 inches of rain, which is more rain than that city experiences during the entire month of September on average.

What is Next?

Beyond Dorian, the next area of interest in the Atlantic is Invest 94L near the Cape Verde Islands. While conditions are marginal for development and there is not much model support, the few models that developed 94L indicate a tropical storm next week.  We will see if this pans out, but I still think the MDR will be lackluster for storm development and much of the named storm activity will be in the North Atlantic and closer to North America, which has clearly been the case so far this season.  Regardless, it looks like the insurance industry has a solid seven days to rest, with the next phase of excitement coming toward the end of this month into early October when large-scale upward motion moves back over the Atlantic Basin, thus helping with new tropical cyclone genesis.

Early signs indicate tropical development over the Atlantic Main Development Region in the coming days, but the probability is low. Clearly, the Basin has storm memory this season, as the pathway for potential tropical cyclones to threaten the U.S is still strong. Source: Michael Ventrice WSI

BMS Tropical Update Sept 5th 11 am CDT

After nine days and well over 10,000 words written in these BMS Tropical Updates, it appears that the end is in sight, but not before the biggest impacts from Dorian are felt by the insurance industry.  Interestingly, though, it seems that interest on social media and other mainstream media regarding some of the impacts of Dorian may have waned because the hurricane is not as powerful as it once was and not threatening the larger population of Florida.  This seems to be the classic psychological burnout of media hype and letdown, while actually downplaying the impact of the real event for the U.S. 

Dorian is still a back up to a Category 3 hurricane that could very well still make landfall along the North Carolina coastline. If it stays at Category 3 strength it could be the first Category 3 hurricane to make landfall in North Carolina since Fran in 1996, which coincidentally made landfall on Cape Fear, North Carolina 23 years ago on September 6th.

Wind Impacts

Dorian is currently located just 50 miles east-southeast of Charleston, South Carolina, and is moving to the north-northeast at only eight mph.  The effects of the storm are being felt along the South and North Carolina coastlines.  The strongest wind gust I have observed as of 5:00 A.M., CDT, was 68 mph at Charleston Airport.  Tropical storm-force winds have been observed along the South Carolina coastline, resulting in about 240,000 reported power outages with that number likely to grow over the next 24 hours.  Dorian is still expected to track right along the coastline and could make landfall near Bald Head Island or Cape Lookout, North Carolina.  Although the strongest winds are on the right side of the storm, and will remain offshore, there is little chance that hurricane conditions will not be experienced along the coast given how close Dorian should track to the coastline.  The worst impacts are likely to occur on the Outer Banks of North Carolina.

BMS iVision allows clients to run risks to better understand the various impacts of Dorian. This is the three-second wind speed gust in mph from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves. Using the three-second gust is often the best wind speed variable to understand damage because three seconds of high winds are often all it takes to cause damage.

Tornado Impacts

Accompanying the high wind gusts embedded in some of the outer band thunderstorms could be tornadoes.  Although often weak and very localized, these tornadic winds can still cause unexpected damage compared to normal tropical-storm-force winds only.  Tornadoes often form in a hurricane’s northeast quadrant and today this quadrant will be over North Carolina.

The tornado threat is substantial with predecessor bands now onshore in eastern North Carolina.
NOAA has a 15% Probability of a tornado within 25 miles of a point.

Storm Surge

Although the moon phase has now moved away from the new super moon phase on August 30, the tidal cycles are still important in relation to the height of water inundation along the coastline.  Last night, cities like Jacksonville and Charleston may have caught a break when the highest storm surge was at low tide.  However, the forecast data for Myrtle Beach, SC does suggest that water could reach a height of 10.3 feet, which would put it at the third-highest tide in history for the city.  Other coastal cities such as Wilmington, NC may also feel similar effects as strong onshore winds push water up along the coast as Dorian slowly tracks along the coastline.  The storm surge will move into the Outer Banks tomorrow, where some of the islands are historically susceptible to water overtaking them and even creating new inlets, cutting off the islands from the only road access.  For an up-to-date view of detailed weather impacts please check out the NHC storm surge inundation forecast.

Myrtle Beach, SC Tide height forecast. Source: NOAA


As expected with a slow-moving major hurricane, rainfall and flooding will also become an issue for coastal areas.  Forecasts call for in excess of 10-15 inches to fall along the coastline.  Combined with storm surge, this will likely create areas of significant flooding or flash-flooding.

Source: NOAA

Analogs Impacts

Although I mentioned tomorrow is the twenty-third anniversary of major Hurricane Fran making landfall in North Carolina, I don’t think Fran is a good analog for Dorian.  Alternatively, Arthur in 2014 continues to be a good analog event, even if, as noted in an earlier post, it remains the only Category 2 hurricane to make U.S. landfall and not result in an insured loss, according to Property Claim Services (this is unlikely to be case with Dorian).  In other earlier blog posts I mentioned the North Carolina impacts from Floyd in 1999 and Matthew in 2016 make a good analog package.  Matthew resulted in $1.2B of insured loss to the same areas, and today Floyd would result in about $2.6B of insured loss, which seem to be a safe range of expected losses at this time for Dorian.

Other tropical troubles

We are a few days removed from the peak of hurricane season, but the Atlantic basin is still very active with several areas of tropical trouble.  The insurance industry will need to focus its attention on impacts from Dorian as there is nothing that is a threat to the industry over the next two weeks.  But, the end of this month could provide some more excitement.

BMS Tropical Update Sept 4th 10 am CDT

Florida Impacts

For the last several days, I have been suggesting Matthew in 2016 would be one of the better analog events to help understand Dorian’s potential impacts.  Now 90 miles east northeast of Dayton Beach, Florida with winds of 105 mph, Dorian is on par with the strength of Matthew, although Matthew was much closer to the coastline at 35 miles.

This morning the power outage map indicates only 15,000 customers are without power across Florida.  Power outages are a good reference to potential damage – during Matthew, for example, power outages across Florida reached over 1,000,000 customers.  The power outages are a function of high wind speeds and, thus far, they have only been of tropical storm force along the coastline.  The highest wind speed I observed this morning was 60 mph at Cape Canaveral.  During Matthew, several hurricane-force wind gusts were reported, including a 100 mph gust at the same location.  It appears that the overall Florida insurance loss from Dorian will likely be less than Matthew, 2016 which was around $1B across the state.

BMS iVision allows clients to run risks to better understand the various effects of Dorian. This is the three-second wind speed gust in mph from Verisk Weather Solutions. The wind swath product maintains the historical swath, not just the forecasted wind speed of the hurricane, so clients can get a better understanding of the wind impact from the full event.

There are signs that Dorian is tapping into the warm waters of the Gulf Stream and may regain some of the intensity that it lost to the upwelling in recent days.  The expansion of the storm’s wind field is notably impacting Florida in terms of high waves and a storm surge of 1 – 2 feet.  Since Dorian is a large storm, it takes exponentially more energy to ramp up wind speeds in the storm’s current eyewall than when it was much smaller.  At this time, I don’t expect Dorian to get much stronger as it tracks along the coastline over the next 48 hours.

Carolina Impacts

Overnight the model guidance has continued to suggest that Dorian will track very close to the coastline of South and North Carolina, with the newest ECMWF forecast indicating that the center of Dorian will track over Bald Head Island and Cape Lookout, NC as a Category 1 hurricane.  The NHC currently does not have Dorian making a U.S. landfall, but a landfall on Cape Romain, SC, Bald Head Island, NC or, especially, Cape Lookout, NC can’t be ruled out.  There is a 100% chance that hurricane conditions will be experienced along the coast given how close Dorian will track to the coastline.  The worst impact may be on the southern Outer Banks of North Carolina.

BMS iVision allows clients to run risks to better understand the various impacts of Dorian. This is the three-second wind speed gust in mph from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves. Using the three-second gust is often the best wind speed variable to understand damage because three seconds of high winds are often all it takes to cause damage.

When analyzing wind impact, it’s important to understand the design of typical structures across the Carolinas.  The following is a great online interactive map.  It should be noted that older structures would have slightly different wind speed thresholds depending on the era, but generally, the trend would be similar to what is shown below.  An assessment by the Insurance Institute for Business & Home Safety (IBHS) suggests that the residential building codes and enforcement level are not as highly ranked as Florida.  However, South Carolina is closer to Florida, while North Carolina was still behind on adapting the latest national building codes as of 2018.

Given how close Dorian is expected to track to the coastline, it will likely result in the coastal plain experiencing 5 to 10 inches of tropical storm force wind-driven rain.  The rainfall amounts will taper inland, but much of the eastern portion of central North and South Carolina could experience 1 to 4 inches of rain.

The coastal section of the Carolinas will experience the highest rainfall totals. However, widespread flooding is not expected with Dorian’s increasing forward motion. There shouldn’t be a repeat of the flooding that was experienced across the region a year ago from Florence. Source: NOAA
Given Dorian’s size as it moves parallel to the coastline, it should be expected that large waves and storm surge will cause coastal erosion. Source NHC
Current Forecasted Storm Surge Advisory Heights. Source NOAA

Analog Events for the Carolinas

I mentioned yesterday that there are many variables to consider when evaluating the potential outcomes of Dorian and one, in particular, may stand out.  When Category 2 Hurricane Arthur tracked over the Outer Banks of North Carolina in 2014, it became the only Category 2 hurricane to make U.S. landfall and not result in an insured loss, according to Property Claim Services.  So, there is a chance that the resiliency of structures and the nature of Dorian’s structure as it passes along the coast of the Carolinas may also result in minimal loss.  In addition to Arthur, hurricanes Ophelia in 2005Irene in 1999Gladys in 1968 and Donna in 1960 are all good analog events that could be used to help understand the impacts from Dorian over the next 48 hours.  The range of insurance industry loss from these events ranges from $250M to $2B for the Carolinas.

There are encouraging signs that the overall insured impacts from Dorian will be minimal across Florida, but there is still uncertainty around the potential impact to points north in the Carolinas, which will likely experience areas of insured loss.

BMS Tropical Update Sept 3rd 10 am CDT

Forever changed

After pummeling the Bahamas for the last 48 hours, the images now surfacing verify that the northern islands have been devastated and may take years to recover. Hopefully, the Caribbean Catastrophe Risk Insurance Facility (CCRIF) will help with rapid infusion of cash to assist with immediate post-disaster priorities, but the damage at this point is extensive. The Red Cross suggests in its early estimates that over 13,000 homes have been destroyed, which does not likely take into account any damage that is occurring on the island of Grand Bahamas. In fact, ICEYE, which is a tech company using satellite radar to map flooding, provided this image yesterday of the storm surge on Grand Bahamas and Freeport, which shows much of the island underwater.

Source: ICEYE

Once again, it’s remarkable that a major hurricane has been stationary just east of where 4.6 million people live along the South Florida coastline. Further, the eye of the hurricane, which could have brought relatively calm conditions to Freeport, Bahamas, was outside that city by just 20 miles and, instead, Freeport has experienced the onslaught of northwesterly 100 mph winds for over 36 hours.

It is already clear that Hurricane Dorian will have its place in history achieving several meteorological records and, once again, Philip Klotzbach has been keeping track of the notable facts.

Impact on Florida

Today Dorian will begin its long-awaited movement to the north, with much riding on the exact timing/magnitude of the turn. Intense hurricanes like Dorian wobble back and forth, so what might look like a turn towards the north may actually be this kind of fluctuation. These wobbles could also be eyewall replacement cycles, which have occurred over the last 24 hours, expanding the wind field of Dorian, as expected. Currently, the radius of maximum winds is 25 miles from the center as opposed to yesterday when it was only 8 miles. It seems that overnight Dorian’s satellite appearance has degraded somewhat as it churns up cooler water from below the ocean surface while it remains stationary just north of Grand Bahama Island. The eye remains well-defined, but it’s not quite as warm nor as symmetrical as it was when it first moved into the Bahamas 48 hours ago. Additionally, the upper-level outflow on the western side of the storm is becoming restricted as the storm begins to encounter some westerly wind shear. Thus, Dorian will likely stay at a weaker Category 3 hurricane over the next 24 hours as it tracks along the Florida coastline roughly 75 miles offshore. This should result in the majority of the hurricane-force winds staying offshore as well, since these winds will only extend about 40 – 50 miles from the center of the storm. So, it seems that the overall impacts will be similar to Matthew in 2016 or less so. For more details, please refer to yesterday’s BMS Tropical Update.

BMS iVision allows clients to run risks to better understand the various effects of Dorian. This is the three-second wind speed gust in MPH from Verisk Weather Solutions. I don’t expect much adjustment to this wind swath at this time. Currently, it shows that the damaging three-second winds will stay offshore.

Impact on the Carolinas

The water vapor satellite image below shows the trough starting to enter the Florida panhandle that will finally lift Dorian off the Bahamas. The trough contains shear and dry air that will aid in weakening the storm further as time progresses. However, Dorian is still a Category 2 hurricane so a lot of weakening must occur for it to become a minimal event to the Carolinas.

The overnight model guidance seems to trend for a higher impact on South Carolina and North Carolina as the models continue to have difficulty understanding the strength of the Bermuda High, so it is still possible that a Category 1 or 2 hurricane could make landfall in the Carolinas. Most of the model ensemble guidance suggests a track with the center of Dorian just offshore with no landfall, but the western eyewall could pass over parts of the coastline such as Cape Fear or Cape Lookout. Currently, there is a 90% probability that the coastline of the Carolinas could see hurricane-force winds. Therefore, the overall effects of wind could be higher along the Carolinas than what is observed in Florida today and tomorrow.

BMS iVision allows clients to run risks to better understand the various impacts of Dorian. This is the three-second wind speed gust in MPH from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves. Using the three-second gust is often the best wind speed variable to use to understand damage because three seconds of high winds are often all it takes to cause damage

While there are many variables to consider when evaluating the potential outcomes of Dorian, one in particular may stand out. In 2014, when Category 2 Hurricane Arthur tracked over the outer banks of North Carolina, it became the only Category 2 hurricane thus far to make U.S. landfall and not result in an insured loss, according to Property Claim Services. So there is a chance that the resiliency and the nature of Dorian’s structure as it passes along the coast of the Carolinas may also result in minimal loss, but this notion will all depend on the storm’s track, just as it has along the Florida coastline. In addition to Arthur, hurricanes Ophelia in 2005, Irene in 1999, Gladys in 1968 and Donna in 1960 are all good analog events that could be used to help understand impacts from Dorian over the next 24 hours.

At this time it is too early to determine the severity of inland flooding and storm surge, but in a faster-moving and recurving storm, the impact of the storm is typically less than if the hurricane were to approach at a perpendicular angle to the coastline like Florence did last year. Tomorrow we will have a much better idea of these impacts.

BMS Tropical Update Sept 2nd 10 am CDT

Miles Still Matter

As you read this insight this morning, think about the following fact:  a Category 5 Hurricane Dorian, with winds of 165 mph, is stalled just 120 miles east of a population of around 4.6 million people.  If you have ever had doubt that a storm capable of producing $200 billion in insured loss can’t happen, this is a prime example that it can, as Mother Nature has no issue creating such a strong storm near the southern Florida coastline.

Similar to Hurricanes Matthew in 2016 and Irma in 2017, the insurance concentration of southeast Florida appears to have once again escaped catastrophic damage.  However, that has not been the case for the Bahamas as the devastation will likely continue for the next 24 hours as Dorian is stalled over the area.  

Dorian is setting all sorts of records.  There are too many to mention here, but if you are interested, have a look at the twitter feed of Philip Klotzbach from Colorado State University where he lists such records and other facts about this storm.  One of the most impressive facts about Dorian is that its winds are the strongest (185 mph) this far north (26.6 degrees N) in the Atlantic ever recorded.

As forecasted over the last several days, Dorian has all but stopped its westward motion over Grand Bahama Island.  Very little movement is expected today before Dorian starts to move north-northwest.   The sharpness of this turn north remains a huge question and will ultimately decide how much of Florida and the southeast coast of the U.S. experience hurricane conditions.

Currently, Dorian’s maximum speed winds are eight miles north-northeast of the center of the storm.  But with hurricane-force winds only extending out from the center of the storm on its southwest side to about 27 miles, Dorian continues to be a small storm.  As Dorian turns north it is expected that a few eye-wall replacement cycles could occur.  That means that the overall maximum wind would come down a bit, but the wind field will expand in size; but by how much is unknown.  The strongest winds of an Atlantic hurricane are generally on its right (east) side; the winds on the left (west) side of a hurricane are almost always weaker and have less extent from the center.  Therefore, as mentioned over the last few tropical updates, the distance from the coastline that Dorian tracks will matter.

BMS iVision not only has the proprietary Verisk Weather Solutions wind swath views, it also takes in the NHC forecast information in which clients can run scenario views. Above, is a measure of the 5:00 A.M. NHC advisory and the distance the center of Dorian will track from the coastline. Its closest approach looks to be about 37 miles from Cape Canaveral. Expecting Dorian to grow in size if a scenario view is done to represent the extent of the hurricane-force winds on the left side of the track (40 miles) we can see only a small section of coastline experience hurricane-force winds under the 5:00 A.M. NHC forecast advisory scenario. This could change over the next few days but clients can perform such an analysis with every advisory.

A Break Down Of Dorian Impacts


BMS iVision allows clients to run risks to better understand the various impacts of Dorian. This is the three-second wind speed gust in MPH from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves. Using three-second gust is often the best wind speed variable to use to understand damage because 3-seconds of high winds are often all it takes to cause damage.

Once Dorian starts it north-northwest movement up along the Florida coastline its forward speed will likely continue to be slow, which could increase insured loss as prolonged high wind speed can also increase damage. However, hopefully homes along the coast are also built to higher standards.


The U.S. has had an above-normal tornado year and I expect some tornadoes to occur with Dorian to add to this total count.  The area where tornadoes often form in a hurricane is in its northeast quadrant, but it does not mean they can’t form on the west side of a hurricane. How severe the tornado threat is will again depend how close Dorian tracks to the coastline.

Storm Surge

The east coast of Florida is, for the most part, not as susceptible to storm surge as the west coast of Florida.  However, there are many inland waterways along the east coast that will have an increased risk of storm surge, but since Dorian is not forecasted to push inland perpendicular to the Florida coastline the risk is mitigated to some extent.  The NHC has now started issuing high-resolution storm surge flooding inundation maps that represent the storm surge flooding over the next few days.  The current forecast shows the most significant flooding concern in these waterways resulting from the very high waves that are expected over the next several days as the storm produces intense swells in excess of 30 feet as it moves north and then northeast.  These waves will likely also produce dangerous surf along ocean-facing beaches, beach erosion, and potential coastal flooding given the extremely high tides this weekend.

Detailed storm surge maps can be found at this link from the NHC.


Dorian is expected to produce very heavy rainfall, with some locations across the Bahamas seeing more than three feet of rain due to the stalling nature of the storm.  Across the U.S., some locations could see six or more inches of rain where Dorian tracks closest to land.  Given the uncertainty in the track, it is still unknown which areas would receive higher amounts of rain.  If Dorian tracks very close to shore or makes landfall, more significant rain could occur across the Florida peninsula and along the Southeast of the country.  However, if Dorian remains further offshore, most of the heaviest rain would remain away from land.

Source: NOAA

Long Range Forecast

There is much more certainty in the long-range forecast now.  Dorian will likely not make U.S. landfall and should track along Georgia, South Carolina, and North Carolina coastline.  By Wednesday, the storm will be approaching the latitude of Jacksonville, Florida and will probably begin turning to the northeast in response to an advancing mid-latitude trough.  This turn will likely bring the storm very close to Georgia, South Carolina, and North Carolina.  On Friday morning, the storm will be departing the area close to North Carolina and heading to the northeast as it transitions into a powerful extratropical cyclone in the North Atlantic. The storm may pass close to the outer banks of North Carolina and up to Cape Cod, Massachusetts, but it’s far too early to tell what the extent of those impacts might be until we see how close the storm will track to the U.S. coast.

BMS Tropical Update Sept 1st 10 am CDT

Miles Will Matter

Hurricane Dorian is a marvel of nature.  Below is a one-minute satellite image of Dorian as the sun rose this morning.  Dorian has maintained a circular and symmetrical eyewall throughout the day and night, and reconnaissance aircraft data has continuously supported a high-end Category 4 storm with slight fluctuations in intensity. In the latest advisory, Dorian has achieved Category 5 status and could stay at this strength for the next 24 hours since ocean heat content is higher over the Bahamas.  Without deeper water to upwell cold water, there aren’t many factors that could weaken Dorian over the next day besides its slow motion and an increase in wind shear, which won’t occur for several days when Dorian turns north midweek.  Unfortunately, Dorian will likely have a devastating impact on the northern Bahamas over the next 36 hours, which could change the landscape on some of the islands for decades.

This is the Brian Tang named storm guidance tool as pointed out in Thursday’s BMS Insight. This shows that the overall observation (red x) is nicely aligned with the National Hurricane Center (NHC) Official Forecast (OFCL) track. The newest NHC OFCL forecasts are darker blue in color and the older forecasts are lighter blue.

Dorian’s motion, for the most part, has closely followed the last several NHC forecast tracks, though it is still consistently traveling incrementally faster than expected. While that sounds inconsequential, slight differences in forward motion over the next two days are hugely critical to what the effect will be to Florida if the storm tracks farther west than expected, versus if the storm tracks more slowly and stays east, which would result in only modest impacts to the insurance industry. The NHC forecast drops Dorian’s forward speed to about 3-4 mph between late tonight through Tuesday morning so, again, this will be critical to watch. The NHC forecasts have been closely aligned with the American Global Forecast System (GFS) and European Center for Medium Range Weather Forecasts (ECMWF) ensemble guidance, which has shown run-to-run stability, as opposed to some of the deterministic models which seem to have different solutions each forecast cycle.  The impact has become a game of miles, since truly dangerous conditions will extend 50 – 75 miles from the center if the storm stays at its current size.  At this time, it does not look like the storm will become much larger in size, like Sandy or Floyd, but this could change with a few eyewall replacement cycles, which largely have not occurred with Dorian thus far.

This is the ECMWF ensembles at 60 hours. Close to 1/2 of the members have Dorian’s low center within 100 miles of the FL coast and about 1/3 are within 60 miles. The average track error at 60 hours is about 100 miles, so it is important not to exclude hurricane impacts to Florida yet. Source:

Longer Range Forecast

There is very little change in the long-range forecast due to the uncertainty in the prior 60 hours. Exactly how close the eye makes it to the coast of Georgia, South Carolina or North Carolina is still something we don’t know, and likely won’t until Monday afternoon or Tuesday. By Wednesday, the storm will be approaching Jacksonville, Florida’s latitude and will begin turning to the northeast in response to an advancing mid-latitude trough. This will bring the storm either over or very close to Georgia, South Carolina, and North Carolina.  On Friday morning, the storm will be departing North Carolina and heading to the northeast while transitioning into a powerful extratropical cyclone in the North Atlantic. The storm may pass close enough to Cape Cod to bring rain and wind to parts of New England, but it’s far too early to tell what the extent of those impacts might be.


The Florida east coast from roughly West Palm Beach to Jacksonville remains in the NHC’s three-day cone of uncertainty for potential center locations. This cone is based on typical NHC track forecast errors over the past five years and is intentionally designed so that about one-third of hurricane positions end up outside the cone. Dorian’s track errors thus far have been anything but typical, so the storm is being watched very closely.  Based on the models, at this time it seems there are three major scenarios: (1) a 15% chance of actual landfall, probably somewhere north of Vero Beach followed by a turn north up the peninsula; (2) a 30% chance of Dorian passing within 50 miles of the Florida coast and bringing significant coastal wind and rain; and (3) a 55% chance of a turn far enough offshore that impact will be relatively minimal.  Of course, these odds are fluid and will change tomorrow once we see what the forward motion of the storm is doing over the Bahamas.


As stated several times, the wind impacts along the coastline will all depend on the track.  Maybe it will be similar to Hurricane Matthew in 2016, as pointed out yesterday, which caused about $1B in insured loss to Florida.  However, if the storm grows in size or tracks closer to the coastline and results in much stronger winds, the insured losses could easily grow as they did with Hurricane Floyd in 1999.  It is far too early to understand what the impact might be to Georgia, South Carolina or North Carolina, but the same principles apply.  If the storm tracks closer to the coastline or makes landfall, surely the insured loss will be greater than if the storm tracks farther from the coast. 

BMS iVision allows clients to run risks to better understand the various impacts from Dorian. This is the 3-second wind speed gust in MPH from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves.
Above is the maximum gust during the forecast period of the ECMWF model. The wind speeds here are in knots. To get a mile per hour value, multiply the speed value in the map by 1.151. Source


Similar to wind, the impact of rain will depend on the track of the hurricane. Currently, the freshwater flooding is not a concern if the storm does not make landfall, and the rain that does fall should be manageable if the storm stays 60+ miles off the Florida coastline.  The biggest flooding concern in the current forecast are the significant wave heights that are expected over the next several days with the storm producing intense swells in excess of 30 feet as it moves north and then northeast.  These waves will likely produce dangerous surf along ocean-facing beaches, beach erosion, and potential coastal flooding given the extremely high tides this weekend.

The rainfall forecast is highly dependent on the forecasted track of Dorian Source: NOAA

Final Thoughts

In meteorology forecasting class, I was taught to not verify a forecast with a forecast. There are encouraging signs that the overall insured impacts from Dorian will be minimal, but there is still a lot of uncertainty to the potential impact to Florida or to points north in the Carolinas.

BMS Tropical Update August 31st 9 am CDT

As expected, Dorian is currently a healthy Category 4 hurricane about 445 miles east of West Palm Beach, Florida.  I would expect Dorain to track toward the northern Bahamas today and it would not surprise me if for a brief period it reaches the status of a Category 5 hurricane.  However, starting tonight, Dorian should start to slow its forward motion which is currently at 12 mph.  There is high confidence that Dorian should be near the northern parts of Abaco Island in the Bahamas by Sunday afternoon.  This, however, is where the forecast uncertainty starts to build.  Because we don’t know how much Dorian will slow down, we don’t know exactly where it will be on Monday.  This is critical because on Monday night, the subtropical ridge that has been steering the storm westward is forecasted to weaken due to the arrival of some upper-level energy from the polar jet stream in the mid-latitudes which might allow Dorian to begin a northward turn into that gap of the upper-level ridge that has been keeping Dorian from moving northward. 

The ECMWF 500 millibars pressure chart. Orange areas are areas of higher-than-normal pressure. Blue areas are lower-than-normal pressure. As shown, a gap opens in the subtropical ridge allowing Dorian to get pulled northward.

Since last Wednesday, I have been talking about the overall forecast uncertainty with Dorian and early on mentioned the angle at which Dorian impacts the Florida coastline will make a large difference in how it affects the state, including the effects of storm surge and coastline winds.  The angle of high winds can create different outcomes, which is why the angle of impact is important.  In the last few track forecasts, the impacts of some of the tracks are less perpendicular to the coastline as the storm is expected to turn northward before a potential landfall which is quite different than the forecast just 36 hours ago.

Since Wednesday, I also highlighted many things that can help the insurance industry limit the effects from the uncertainty in the forecast.  One of the things I failed to mention was in Meteorology forecasting classes one of the lessons they teach is, “the trend is your friend.”  Over the last 36 hours or so the long range forecast guidance has been trending to a slower moving storm which is resulting in forecast guidance that Dorian might not make a Florida landfall and may actually turn to the north and track along the Florida coastline due to the weakness in the upper-level ridge as described above.  Yesterday I mentioned that the slow-movement of named storms often precedes erratic behavior or sharp turns, so what this trend suggests today is that Dorian may stay offshore, similar to Hurricane Matthew in 2016.  In this case, “the trend is your friend” as this overall scenario would drastically limit insured losses.  


There is no shortage of hurricanes that have bypassed Florida.  These types of bypassing storms have had varying impacts.  One might remember that Hurricane Floyd in 1999 caused substantial loss along the east coast of the U.S. and in today’s dollars resulted in close to $15–$20 billion in insured loss.  In contrast, Hurricane Matthew in 2016 was another very powerful category 3-4 hurricane as it tracked up the coastline but only caused about $1 billion in insured loss to Florida.  Both of these storms were clearly different with different impacts. Maybe the coastline has become more resilient since 1999 which might explain part of the loss difference between events. Both Floyd and Matthew were also impactful storms to Georgia, South Carolina and North Carolina, and this potential seems to be the overall risk focus for Dorian in the longer range as the models are locking onto different scenarios than a Florida landfall.  But the short term risk is not over for Florida as there is still a probability of coastal sections seeing hurricane-force winds.

Wind Impacts

BMS iVision allows clients to run risks to better understand various impacts from Dorian. This is the 3-sec wind speed gust in MPH from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves.

As discussed above, many forecast models have backed away from a Florida landfall.  This change is the current outlook that is forecasted by Verisk Weather Solutions, which shows minimal impacts to the Florida coastline.  Using BMS iVision users can also select historical storm tracks to understand expected loss.  Below I show the BMS iVision Verisk Weather Solutions verification for Hurricane Matthew in 2016 which may also represent an idea of what type of winds to expect along the coastline as another potential forecast scenario for Dorian.  It should be noted, however, that depending on the overall storm size, slight shifts in the forecasted track could result in a big difference in what types of wind speeds are experienced along the coastline of Florida which might be the difference in Dorian following Floyd in 1999 or Matthew in 2016.

BMS iVision allows clients to run risks to better understand various impacts from Dorian. This is the 3-sec wind speed gust in MPH from Verisk Weather Solutions. There will be some adjustment to this wind swath as the forecast evolves.

Flood Impacts

The good news is with Dorian shifting its forecasted track to the east of Florida the trend has moved the heaviest rain eastward as well.  This means that freshwater flooding may not be as bad, but the coastal section could still see rain and the compounding impacts of storm surge with waves and coastal erosion. The current forecast expects significant wave heights for the next several days with the storm producing intense swells in excess of 30 feet as it moves north and then northeast.  These waves will likely produce dangerous surf along ocean-facing beaches, beach erosion, and potential coastal flooding given the extremely high tides this weekend.

The rainfall forecast is highly dependent on the forecasted track of Dorian: Source: NOAA

Long Range Ideas Impacts for Dorian

Exactly how close the center of Dorian makes it to the Florida coast is still something we don’t know, and likely won’t until Monday.  By Wednesday, the storm is expected to still be offshore near Jacksonville, Florida.  The current expectation is that it will begin turning to the northeast in response to an approaching mid-latitude trough.  This will bring the storm either over, or very close to, Georgia, South Carolina and North Carolina.  Again, the exact track is still uncertain but the general idea is becoming much clearer.  On Thursday, the storm will probably be departing North Carolina and heading off to the northeast as it transitions into a powerful extratropical cyclone.  The storm may pass close enough to Cape Cod to bring rain and wind to parts of New England, but it’s far too early to tell what the extent of those impacts might be.