partner login

BMS News

BMS Tropical Update 9/8/2017 11 AM CDT

Margin of error very slim now

As tough as this is for me to say, we are now in a window of model error (48 hours at about 50 miles) that suggests we are at the point where south Florida will likely see a catastrophic hurricane landfall. Based on some scenarios, this could be a historic and even game changing event for the Florida insurance industry. However, without yet knowing where Irma’s turn to the north will be, there is still some uncertainty around the exact impacts from the storm. Given how the models are now consistently suggesting a landfall location between Key West and Miami, several catastrophic scenarios are likely depending on the track. The consensus at this time seems to be that the landfall location could in the middle of the Florida Keys by early Sunday morning.

If there is any good news, it’s that as of this morning Irma is now a high-end Category 4 hurricane, likely due to an eye wall replacement cycle overnight. However, I feel this small downward tick in intensity will be short lived because some of the warmest sea surface temperatures in the world are in the southern Bahamas, and Irma is expected to track over this area. In general, I feel that Irma will be a strengthening storm as it heads toward the coastline, rather than weakening like Rita and Katrina did. Also, the overall environment continues to have light wind shear, so at this time it is expected that Irma will be a high-end Category 4 or low-end Category 5 hurricane with winds of around 155 mph as it nears the southern Florida coastline. This is likely near the building code requirements for most residential structures in south Florida. Of course, every building’s construction and the surrounding frictional effects will ultimately determine the damage at any given property.

Florida has some of the best building codes in the U.S. These codes were largely developed from past major hurricanes and will be tested with Irma. Source; IBHS


Summary of Impacts

  • Wind impacts will depend on slight east or west movements in track over the next 48 hours, with local building construction type and surrounding frictional effects ultimately determining damage. At this point, it is expected that the Florida peninsula will see damaging winds.
  • Storm surge is also dependent on track. The current storm surge forecast suggests that the Florida Keys and southwest coast of Florida could see the highest storm surge values of around 9 feet.
  • With Irma being a progressive storm, rainfall is expected to be around 15 inches

Wind Impacts
As suggested a few days ago, Irma will continue to grow in size. Irma is a much larger storm than many past hurricanes that have impacted Florida. In fact, this is a graphic put together by CIMMS/NASA that shows the comparison in size between Category 5 Andrew 1992 and the much larger Category 5 Irma as of yesterday.

Hurricane-force winds are currently extending outward up to 70 miles from the center, with tropical storm-force winds extending outward up to 185 miles. This means virtually the entire Florida peninsula will likely experience damaging winds regardless of track. The exact track, however, will determine where the core of the highest winds in the eye wall will be, about 30 – 40 miles from the center of the storm. Therefore, Irma’s exact track is still critical to understanding the total insured loss.

If we continue to follow the best performing ECMWF model, it puts a landfall location in the lower Florida Keys. This suggests that the winds will be just away from the urban core of the tri-county Miami area, lowering the expected insured loss, but any shift back east would increase this expected loss.

This is last night’s ECMWF model wind swath from Weatherbell. It only shows one scenario of expected winds across south Florida.


This is the BMS iVision Verisk Climate 3-sec wind swath. It is just one possible scenario at this time, but take note of the detailed frictional effect within the model.


Current estimate of power outages. Updates are shown here.

Storm Surge
Like the wind impacts, storm surge impacts are also highly dependent on track. Based on the NHC’s very detailed storm surge forecast, the Florida Keys and parts of southwest Florida will likely see the largest storm surge heights. The current surge value predictions in the Miami area are not likely to throw watercraft off their moorings, which would be critical to insured loss estimates given the high concentration of watercraft along the southeastern coast of Florida. However, with a change in track and if the storm moves up the coast, I expect these east coast storm surge values to increase putting watercraft at risk.  Also, storm surge needs to be watched carefully in the Jacksonville area as Irma tracks northward.

NHC Storm Surge Inundation from the NHC Advisory 37. Latest forecasts of surge found here.

Since Irma will likely be progressing up the Florida coastline and not stalling out, the rainfall amounts should be more typical of a landfalling hurricane. Quantitative precipitation forecasts are currently suggesting upwards of 15 inches of rain.

Latest NWS rainfall forecast.

We have seen in the past that it does not take much rain in the Miami area to cause flooding, due to poor drainage and the built-up urban environment. Other areas to watch are the local channels and possibly the Lake Okeechobee level, which is currently at a low level. However, dikes that hold water in the lake are a known hazard and have grabbed national attention as critical infrastructure that needs to be shored up.

Current water level of Lake Okeechobee. Real-time information can be found here.


Longer Range Forecast  Irma
Just like the impacts to south Florida are dependent on the track shifts of west or east along the Florida Peninsula, the impacts to points north could vary as well. Currently, it appears that a west track is favored up the western side of Florida, which would keep the center of Irma over the Florida peninsula for much of the track and starve it of its energy as it moves northward. The negative aspect of a Florida track is that the entire east coast of Florida would likely see stronger winds, with a decreasing wind field as Irma tracks into Georgia and tropical storm-force winds across South Carolina early next week.

Insured loss
As mentioned in the last few blog posts, small shifts in Irma’s track will ultimately result in major differences in the outcome of insured losses. This is still the case today. Regardless, it is safe to say that Irma will likely be one of the costliest storms to impact the insurance industry. Irma’s track is very similar to some of the realistic disaster scenarios set up by rating agencies. These scenarios often suggest over hundreds of billions of dollars in insured loss. It is likely that the insured losses from this storm will at least be in the tens of billions of dollars, on top of over tens of billions of dollars in losses already from Irma’s strike in the Caribbean.

We will be able to have a better grasp on the exact magnitude of this event by tomorrow, as we will have a better understanding of the turn northward that will determine not only the south Florida impacts, but also the losses that may occur in other east coast states.

Replay of Special BMS Irma Webinar

Due to the special forecast situation and potential impact on the insurance industry, BMS Catastrophe Analytics held a Webinar on September 7th which provided the latest on Irma, its forecasted track and impacts for the insurance industry.

Quick note on Jose
As we know, catastrophe models are built on the concept of clustering. Hurricane Jose was upgraded yesterday afternoon to a major hurricane, heading for the northern Lesser Antilles and hitting the same areas that Irma hit with a peak intensity of 185 mph on Wednesday of this week. Please keep in mind that because of the widespread devastation seen from Irma, the vast majority of structures in these areas are already unsafe. In the long-range, Jose is expected to stall north of the North Caribbean during the middle of next week, with some models then tracking Jose west towards the Bahamas or eastern seaboard of the U.S. I will be watching this closely next week.


BMS Tropical Update 9/7/2017 12 PM CDT

The Critical Turn North

We are now in the critical 72-hour window before Irma’s expected interaction with south Florida.  The model runs overnight have been fairly consistent, but for a 72-hour forecast, there still seems to be a considerable amount of spread in the timing of the northward turn.  I have been blogging about what is causing this northward turn since last Friday, which is a trough of low pressure now positioned over the eastern half of the U.S.  I have also been mentioning that Irma would be breaking all sorts of records, which is certainly the case, as only three other Atlantic hurricanes have logged more time as a Category 5 hurricane than Irma.

What we know about Irma over the next few days

  • Irma will be a major Category 4 hurricane, and perhaps even a Category 5, as it approaches south Florida.
  • The likelihood of a landfall in south Florida is currently 85%.
  • The severity of impacts to Georgia and the Carolinas will ultimately be determined based on whether Irma makes landfall along south Florida or if Irma tracks up Florida’s east coast.
  • There is still a 55% chance that Irma will make landfall along the Florida Keys or points along the southwest coast of Florida.
  • Several models also turn Irma northward before a south Florida landfall which would result in less damage to Florida.
  • Based on my analysis of catastrophic model stochastic events, any Category 4 hurricane making landfall in south Florida will likely create at least a $10 billion insured loss event.

Uncertainty in Irma’s forecast improving

Although there is still a high amount of uncertainty in the short range forecasts, the evening forecast models are converging on the idea that Irma will be extremely close to south Florida by late Saturday night with a possible landfall early Sunday morning.  I have seen only a handful of models that do not bring hurricane-force winds to south Florida.  Most forecast models at this time suggest that winds in excess of 115 mph are possible across the major metropolitan corridor of south Florida.  I expect Irma to grow in size as it approaches south Florida, with hurricane-force winds easily extending 50 miles from the center of the storm.  However, if Irma tracks up the east coast of Florida, the strongest winds would stay on the right side of the storm, similar to what occurred with Matthew.  Irma looks like it could be larger in size than Matthew at this point in time though, so hurricane-force winds would likely have a greater reach inland than Matthew even if Irma tracks up Florida’s eastern coastline.

Depending on whether Irma makes landfall, the impacts beyond Florida could vary.  If Irma tracks up the middle of Florida, it would likely be in a much weakened state as it tracks northward.  If it stays just off the shores of eastern Florida, the hurricane would be in a much stronger state as it tracks up the coast toward Georgia and the Carolinas due to continuously being able to feed off the warm gulf stream waters, and it would likely be a major hurricane at any landfall points north.  Currently, for this model scenario, the border between South Carolina and North Carolina has the highest likelihood of landfall at 60%.

Specific Impacts

Wind Damage

With Irma’s forecast cone of uncertainty shrinking, the BMS iVision Verisk Climate wind swath product is providing its first views of what the expected wind damage across south Florida might look like as Irma makes its turn north.  It should be noted, however, that this is just one scenario of many that is still possible.

Storm surge

At this time it is too early to determine storm surge.  I expect the NHC to run scenarios based on the NHC official forecast beginning Friday.  They have already issued storm surge watches for much of south Florida.  Currently, Irma  is not a worst case scenario for storm surge.   At this point areas along the coastline could expect 5 – 10 feet of storm surge, which is highly dependent on storm movement and location.


At this point it looks like Irma should be relatively progressive, so unlike Harvey, it should not stall out over the Southeast.  However, large amounts of rain in excess of 15” could still fall over Florida as Irma tracks northward over the weekend and into early next week.  These rainfall forecasts will be revised over the next few days.

Irma’s Insured loss

Irma will likely already go down as one of the newest realistic disaster scenarios for the damage it has caused on the Caribbean islands, so I have little doubt the industry will be talking about its impacts to the area for decades to come.  Irma passed 50 miles north of San Juan, Puerto Rico, last night.  At that time, the strongest hurricane-force winds were on the north side of the storm and extended out from the center of the storm at an estimated 50 miles.  This is important because it appears that the northern coastline of Puerto Rico was spared widespread damage as the strongest winds stayed just offshore, although some local damage is being reported across social media.  This serves as a great example of why the storm track in relation to the radius of hurricane-force winds is critical in understanding the impacts we might see as Irma tracks up the East Coast of the U.S.  The catastrophe modeling companies are releasing event sets to help determine these impacts, but it is premature at this time to share the details of these losses aside from saying the losses will reach into the billions.

Over the last several days I have tried to finding analog events to match Irma’s intensity and there are very few.  Although there is less uncertainty in the forecast track today, there are too many different solutions, and how those solutions unfold could have a large impact on the heavily insured south Florida coastline.  Given these sensitivities, a difference of even 20 miles can influence loss amounts by the billions.

The catastrophe modeling companies should start issuing their pre-landfall preliminary set of tracks tomorrow which will give some early estimates, but these estimates will also change.  I think even tomorrow there will still be some uncertainty on the track Irma will take northward, either over Florida or off the coast.  As I mentioned, in a search for Category 4 events that make landfall in south Florida, it is hard to find an event with under $10 billion in insured loss.

BMS Tropical Update – Hurricane IRMA GoToWebinar

Due to popular demand for today’s webinar on Hurricane IRMA (4p ET/3p CT), we have changed the weblink for the webinar.

Please use the link below to register for the Webinar.

Andy Siffert, VP/Meteorologist will be providing an update on Irma’s forecasted track and intensity.

Additionally, Julie Serakos, EVP and head of Cat Analytics will provide a modeling update and Kris Westall, COO and Urban Friesz, VP will provide an update on claims and accounting services.

Please join us for this important update.

BMS Tropical Update 9/6/2017 1 PM CDT

One For The Record Books

It appears that the global models have a common case of the flip-flops over the last 48 hours, which is typical with a complex forecast situation. We are also seeing some of the history books on hurricane intensity being rewritten. Based on information received from Phil Klotzbach at Colorado State University, Irma has now had a maximum wind of 180 mph+ for the past 24 hours, which is an Atlantic Basin record in the satellite era. This is just one of the many records Irma is breaking, and unfortunately it’s not done yet as it is still a Category 5 hurricane.

Caribbean Islands take a direct hit

Irma crossed several Caribbean islands overnight and is now moving in a slight west-northwest direction at 16 mph, placing it very near the U.S and British Virgin Islands. The damage I have been seeing across social media on many of these islands is heavy. This is to be expected from a Category 5 hurricane, as even some of the strongest buildings can sustain damage.

Below is a once in a lifetime sunrise as the northern Caribbean islands of Angulla, Saint Martin, Sint Maarten and St. Barts saw daylight break in the eye of the most powerful hurricane to impact these islands in recorded history.


Model Flip Flop and Irma’s Future Track
As we continue to watch Hurricane Irma strengthen and shatter the record books on intensity, the question of where Irma will go this weekend is still hanging in the air.  It’s all about the turn north, but when and where will that occur?
As I have been saying since Friday, a potential turn north all has to do with a weakness in the Bermuda high that will develop as a result of the passing of an upper level trough of low pressure on the East Coast. This turn north is very critical in determining Irma’s future track and any resulting impact to the U.S. insurance industry. This past weekend, we saw a lot of model guidance suggesting the possibility of Irma tracking up Florida’s east coast. On Monday and earlier Tuesday, the general model guidance was suggesting that Irma might track closer to the Florida Keys. As of last night and this morning, the model guidance is back to suggesting that Irma could track up Florida’s east coast. This is a classic case of model flip-flop in a very complex weather pattern, and the west and east coasts of Florida are still very much on the table as potential tracks for Irma. However, as I talked about yesterday, if you forecast with the trend, the trend is currently moving towards a track up the east coast of Florida.

Current Water Vapor Image detailing the global weather systems at play causing the model flip flop

The image above is the latest water vapor satellite image of North America. In this image, note the area labeled key polar shortwave. This is the shortwave that needs to be watched very carefully. This disturbance is forecasted to dive rapidly into the Mississippi River Valley within the next three days. The timing of this shortwave may cause it to interact with Irma as the storm moves into the Bahamas.
The concern here is that if this shortwave is stronger than some of the models suggest, than the upper level low will be able to form more quickly. If the upper level low forms faster in the Tennessee River Valley than expected, Irma will be forced to turn northward sooner over the Bahamas, and thus would become a threat for the Carolinas and possibly even Virginia. The possibility of Irma escaping out into the Atlantic Ocean with no landfall is still quite unlikely due to the position of the Bermuda high.
The location of this key polar shortwave is what I think is causing some of the model flip flop because that polar shortwave is in an area of poor sampling.  Therefore the global model guidance used to forecast Irma’s track has a high degree of error, and also why there is so much volatility in the models from run to run. The shortwave is also why the threat of Irma tracking east of Florida to locations further north to North Carolina remains on the table today. However, as I highlighted in yesterday’s post, a track into the Florida Keys and up the west coast of Florida is still a possibility – it’s just not a favorite model solution today.  As shown below, even the NHC official forecast for days four and five are seeing large adjustments, and these forecasts have errors of 175 and 225 statute miles, respectively.

With so many models, it is important to keep track of which model is performing the best. Currently, the ECMWF has the lowest track error at days 4 and 5.


This is the current ECMWF and its ensemble of probability. Currently, Florida is at the highest risk of seeing Irma landfall. However, many model members of the ensemble also track Irma into the Florida Keys and up the East Coast.


Irma’s future intensity
While the models have a case of the flip-flops in terms of the forecasted track, what is remarkable is that there is very little disagreement on Irma’s future intensity. Since last Friday when I first started writing about Irma, it was well forecasted to become a major Category 4 or 5 hurricane impacting the northern Leeward Islands. These forecasts have been remarkably consistent in an era when forecasting intensity is typically the biggest challenge.

At this time it looks like Irma will track far enough away from the mountainous Hispaniola, which would have a significant impact on the circulation of the storm. Because of this and the warm sea surface temperatures along Irma’s track, the models have been forecasting a major hurricane to impact the southern Bahamas and south Florida for several days now and they are not backing away from this prediction.  At this point it is a very safe bet, depending on track, that Irma will be a major hurricane (likely a Category 4 or stronger) near Florida by early Sunday morning. If Irma does not make landfall in Florida, the areas along the coasts of Georgia and the Carolinas should also be prepared for a major hurricane landfall.


Current water average water temperature along Irma’s forecasted path. 29.3C is 84.7F


New Analog Events and Insured Loss

I have been trying to find analog events for Irma, but there are very few events of this intensity that have tracked up the east coast of Florida. The events I provided yesterday would still fit in the model solutions today. However, David 1979 could be used as another possible Florida analog impact, but it was a weaker Category 2 hurricane as it tracked up Florida’s east coast. Due to the current intensity, any historical loss comparison can’t really be used at this time because Irma is expected to be a much stronger storm than anything I could find that has tracked up the east coast of Florida.

With the forecast uncertainty around the potential turn north, it is too early to provide detail on what potential losses could occur. However, just taking a glance at some of the disaster scenarios from catastrophe models, any hit at Category 4 or greater to southern Florida would be at least a $15 billion loss. Some of these scenarios only go up depending on the landfall location, such as a $100 billion loss or even greater, from a direct Miami hit. The ranges of potential losses are great. Matthew, however, showed us that if a storm can track just offshore, it can spare large levels of loss.

Special BMS Webinar Tomorrow

BMS Tropical Update – Hurricane IRMA GoToWebinar

Due to popular demand for today’s webinar on Hurricane IRMA (4p ET/3p CT), we have changed the weblink for the webinar.

Please use the link below to register for the Webinar.

Andy Siffert, VP/Meteorologist will be providing an update on Irma’s forecasted track and intensity.

Additionally, Julie Serakos, EVP and head of Cat Analytics will provide a modeling update and Kris Westall, COO and Urban Friesz, VP will provide an update on claims and accounting services.

Please join us for this important update.

BMS Tropical Update 9/5/2017 12 PM

Irma Heading Into The History Books

I mentioned in Friday’s BMS Tropical Update that Irma would become a major hurricane, possibly a Category 5, near the Leeward Islands. As of this morning, Irma can be found 180 miles east of Antigua and moving at 14 mph, officially a Category 5 hurricane. In terms of wind, Irma is the strongest hurricane in the Atlantic Basin since Felix 2007, and in terms of pressure, the strongest since Igor 2010. Regardless of history and what records might be broken by Irma over the next few days, the fact is that it’s a large and dangerous hurricane not seen in the Atlantic Basin in a long time, and it likely won’t weaken below a major hurricane this week as it tracks closer to the U.S. coastline.

In the short term, Irma will remain a major hurricane as it impacts the northern Leeward Islands later tonight and into tomorrow. These small islands will likely suffer a significant amount of damage as it has been a very long time since a hurricane of this magnitude has impacted these islands (e.g., David 1979 and Sept 1928).

Only 2 Category 5 in NOAA’s best track database w/in 200 nm of the Northern Leeward Islands: David (’79) and Sep. 1928. Both were Category 5 after center passed Leewards Islands.

Irma will track into the southern Bahamas toward the end of this week. There is a chance Irma could weaken slightly due to land interaction with the northern Caribbean islands, but the water in Irma’s path has plenty of energy, which will limit any weakening in the coming days.

The tropical cyclone heat potential (TCHP), is defined as a measure of the integrated vertical temperature from the sea surface to the depth of the 26°C (78.8°F) isotherm. The warmer the more fuel for hurricanes.

Meteorological Rule of Thumb

In my last update, I mentioned that Irma’s ultimate track could be influenced by a trough of low pressure that is now moving into the Upper Midwest and will be hanging around the East Coast for the next several days. Below is a great illustration I found by Philippe Papin, a PhD student studying atmospheric sciences at the University of Albany.

As Philippe mentions, Irma will be moving along the southern edge of the Bermuda high pressure over the next several days. At the same time, a deep trough of low pressure will be moving to the East Coast. This steering flow around the high and the deep trough will begin to turn Irma to the northwest later this week. However, the combination of the trough of low pressure, which gets elongated along the East Coast later this week, and the strength of the Bermuda high likely won’t allow for Irma to turn north before it gets to the Florida coastline. However, what will likely happen is that it will allow for an abrupt northward turn in Irma’s track later this weekend as the storm is near south Florida, when Irma is on the southwestern periphery of the Bermuda high. This abrupt change in track northward means Irma could significantly slow its forward motion as it turns northward late this weekend into early next week. This increases the chance of major inland flooding due to its slower forward motion.

All of this meteorological assumption is based on an understanding of large scale atmospheric motion 120 hours from now, which is five days. Although five day forecasts are improving, small details in the large scale motion can have a big impact on where Irma ultimately makes landfall. In fact, the NHC track errors on a four and five day forecast are 175 and 225 statute miles, respectively.

According to the University of Albany, which keeps track of model forecast errors, the NHC’s five day forecast error is currently running about 200 miles.

Current mean absolute error for Irma forecast tracks from various weather models.

Keep your eye on the ECMWF model, which currently has a very low track error, but overall it’s bound to errors as well. Just look at how the 10-day ECMWF ensemble forecast (an ensemble is a forecast of 52 similar forecasts with different settings) has already evolved for Irma since Friday.

This is the ECMWF Ensemble Forecast from last Friday of all the various track scenarios for Irma.

Irma has been consistently defying the modeling of a northward turn up the East Coast, which was the most common model ensemble solution last Friday. Today fewer ensemble forecasts track Irma in between Bermuda and the East Coast.

This is last nights ECMWF Ensemble forecast which show a much more westward track.

In fact, as I talked about the recurving typhoon rule in my Friday post, I was thinking of another rule that is taught in meteorology school – don’t forecast against the trend. In this case, the trend has been for Irma to track further westward, meaning the chances of Irma ending up in the eastern Gulf of Mexico early next week are much higher today than even just a few days ago. Regardless, Irma will likely have an impact on south Florida along its forecasted path. Based on the latest forecast, there is an 80% chance that Florida will see a major landfalling hurricane.

Initial Insurance Analog Events For Irma

The probability is high because Irma will be a very large hurricane this weekend, with a large eye and hurricane force winds that could extend out from the center for up to 50 miles. With the southern part of Florida only being separated by just over 100 miles from east to west, it seems very likely that some part of southern Florida will be exposed to hurricane force winds.
At this point, the insurance industry should be preparing for another U.S. landfalling hurricane. To help understand the potential impacts at this early stage, a few analogs come to mind. Donna 1960 could be a good analog to the forecasted track of Irma over the next few days. If Donna occurred today, the Florida wind and surge insured losses could amount to $13.6 billion according to one catastrophe modeling firm. Potentially a more extreme analog might be the 1935 Labor Day Hurricane, which today would cause an estimated insured loss from wind and surge for Florida at around $40.7 billion.

There is still a great deal of uncertainty in the forecast for later this weekend, and there are many different scenarios that could play out. One thing for certain is that there has been a large population growth in Florida with a 27% increase in the number of homes since 2000, according to the U.S Census. The exposure is significant, and if Irma arrives with major hurricane force winds, the current building codes over much of the state will likely be challenged.

BMS Tropical Update 9/1/2017 12 PM

As mentioned in the last BMS Tropical Update, it’s a bit early to determine where hurricane Irma will be heading 10 to 15 days from now, which is when it could be eyeing a potential U.S. coastline impact. However, as much of the U.S. insurance industry heads into a long weekend, here is what my gut thinks will happen with Irma over the next 10 days.

When forecasting in the long range, it is important to look at what is occurring worldwide. There is a very good chance that Irma’s future track will be influenced by what is going on in the West Pacific. Currently there is a typhoon named Sanvu southeast of Japan. This typhoon is expected to recurve into the westerlies over the next few days.

Current location and track of Typhoon Sanvu in the West Pacific Ocean

There is a general rule of thumb in meteorology that when this occurs, it could cause a trough of low pressure to move towards the U.S. East Coast between 6 to 10 days later. This means that if the typhoon recurves as forecasted over the next few days, a trough of low pressure could potentially be on the East Coast sometime between September 7th and 11th. In fact, some global models are picking up on this trough of low pressure for next week. Some even produce the first frost of the season for parts of the Upper Midwest, so it will have some punch.
Around September 7th and 8th, Irma will be a major hurricane and will likely be a Category 4 or perhaps even a Category 5 over or near the Leeward Islands. This will result in three potential scenarios for later next week, based on the forecasted trough of low pressure and Irma’s latitude at that time:

  1. Irma could be at a high enough latitude that it gets pulled north by this trough of low pressure, similar to Gret’s path the second week of August, and recurve in between the U.S. and Bermuda.
  2. Irma could be at a low enough latitude that it misses getting pulled into the trough of low pressure and tracks south of Puerto Rico, perhaps into the Gulf of Mexico.
  3. If Irma is near the Northern Leeward Islands, it could get pulled up by the trough of low pressure, but miss the full connection, and head towards Florida and the East Coast with a close landfall threat later the following week.

Right now this might be the best graphic I have found on the future of Irma. Its based on the ECMWF Model, but it provides a good long range outlook at this time of the three opitons above. Source The Weather Channel


Based on the current forecast of Irma being a major hurricane near the Leeward Island next week here is a history of all the major hurricane over this forecasted area and where they tracked. It follows The Weather Channel guidance nicely, but keep in mind this is history and not a forecast.

At the start of the season, I thought the East Coast was going to be the biggest threat of landfalling storms in the U.S., and I don’t see a reason why that should change. Based on the forecast and history, there is currently a 60% chance of Irma impacting the U.S as at least a tropical storm, and  a 30% chance of seeing a major hurricane impact at this time.

As a reminder, I still feel that the MJO will make it more difficult for new storm formation between September 10th and 25th, so the peak of the season could be quiet for new development.

BMS Tropical Update 8/30/2017 12 PM

Harvey Help
Over the week I have shared resources that can help the insurance industry understand the impact from Hurricane Harvey. Now that Harvey is moving slowly inland, with smaller creeks receding and damage assessments beginning, I would like to summarize and share more resources that will help the insurance industry understand Harvey’s impact.

As previously mentioned, BMS clients and prospects have access to several tools that may help understand the various impacts of Hurricane Harvey via the BMS iVision Geospatial Mapping Solution. These products are available across various geographic regions, including the Gulf Coast, with storm attributes that include hurricane and severe weather perils that often accompany a hurricane. Users can intersect these attributes with policies at risk to get an idea of exposure and apply damage ratios as a first guess at potential losses.

• Hurricane past and forecasted track, sustained winds and wind gusts
• Hail size, probability and duration
• Severe Thunderstorm winds, radar max and rainfall

BMS iVision Verisk Climate Total Rainfall Map from between 8/23/2017 – 8/29/2017


BMS iVision Verisk Climate Hurricane 3-sec Wind Gust Swath


BMS iVision Verisk Climate Hurricane 1-min Sustained Wind Swath


BMS iVision Verisk Climate Wind Gust Duration (hr) of 50+ mph

Of course it is possible to add other data attributes to iVision as needed and that has been done for specific users. Outside of BMS iVision, there are several great sources that have appeared to help understand Harvey’s impacts.

Harvey Wind Damage Resources
The NOAA Remote Sensing Division has provided high resolution airborne imagery to support NOAA national security and emergency response requirements. These images have been combined into a larger mosaic and tiled for distribution. The approximate ground sample distance (GSD) for each pixel is 50 cm / zoom level 18.

ESRI Harvey Disaster Response Mapping with various resources with Social Media Layers in which your login for these would be required.

Harvey Flood Resources
Detailed FEMA Flood Zone Mapping of the Houston Area

U-Flood is a crowd sourced effort to map inundated roads in Houston. Flooded (or cleared) streets may be reported by zooming into the map and clicking/tapping them.

The interactive map of evacuation zones for the Brazos and San Bernard Rivers is below.

Addicks and Barker Potential Flood Maps

FM Global Flood Mapping

Houston Harvey maps from the New York Times

Background of urban development in the Houston Area with some flood maps

Other Tropical Troubles
As I mentioned in Monday’s post, a new tropical wave has been named (Irma) in the Main Development Region of the Atlantic and will track toward the U.S. coastline. Irma is still 10 days away, but given the pattern so far, the chances of a U.S. tropical threat late next week are high, but there is still a lot of uncertainty in the pattern as a major cool down for the east coast next week will likely play a role in Irma’s future track. I will provide new updates on this threat as it develops.

BMS Tropical Update 8/28/2017 12 PM

Harvey the Good and Bad News

If there is a bit of good news, it might be that Harvey is pulling in dry air which is giving some reprieve to the heavy rainfall near the center of Harvey.  The bad news is that Harvey as forecasted to moving out over the warm water of the Gulf of Mexico which could further enhance convection and increasing the flow of moisture once again over the next few days.

Harvey Current Status and the End Game

Harvey is currently a tropical storm near Port O’Connor, which is about 50 miles from where Harvey initially made landfall on Friday evening as a Category 4 hurricane. Harvey, which is only moving at 3 mph, will slowly head back over the warm Gulf of Mexico. It is at this point that it will likely slowly start its northeastward movement towards a second landfall area between Houston and the Texas / Louisiana border.  During this time Harvey will continue to tap moisture from the Gulf of Mexico, but at this point there is very little model support for Harvey to regain hurricane strength.  As it moves to the northeast the extensive rain shield will also move to the northeast into Louisiana, which will increase the flood potential for much of the state over the next few days.  By Friday this week Harvey will be extra tropical system over Arkansas and the named Harvey will be retired into the history books only to be referenced back to much like Andrew, Katrina and Sandy are today.



The flood situation continues to unfold.  I honestly can’t keep track of all the records that have been broken, there are so many. So far as far as I can tell the 48″ rainfall record has not been broken yet, but 35″ rainfall totals have been reported which verify the forecasts made early last week.   Some of the most important records are the forecasted river crest in and around the Houston area.  For example, Buffalo Bayou that flows into downtown Houston, has already seen record flooding, in some places by 10 ft.  The gauge at Point Village suggest it will be at record flood stage for at least five days which is likely due to the  Army Crop releasing water from the giant flood retention areas of George Bush Park (Barker Reservoir) and Addicks reservoir and dam just north of the Katy Freeway (1-10).  Both of which are Army Corps project to protect Buffalo Bayou and downtown Houston from flooding.

The Army Corp has told some residents their homes could be flooded for up to two months.


Insured losses

It is still too early to get a complete view of what the insured impacts will be from Harvey as the flood situation continues to develop. With the wind and surge loss estimates ranging from $1 to $3 Billion in insured losses the flood losses will highly depend on what the commercial and energy sectors sustain from the flooding.  There could likely be high content and BI losses within these sectors. The general view is personal line home owner risk is largely not covered and falls on the NFIP. However, if past events are used severe flooding can take its toll on to the auto segment of the industry.  A bit unknown is some private-sector insurers have started to sell stand-alone flood coverage to homeowners, but it is unclear what the market penetration is at this point.

Economically Harvey will likely be one of the largest natural disasters in U.S History. There is talk that it could rival Katrina and given that Houston is the 4th largest city in the U.S. there is clearly more economic potential to be damaged.

Maybe the best initial conservative estimate of Harvey on economic loss might be from disaster economist Kevin M. Simmons, Austin Collage. 

Other Tropical Troubles

Currently Potential Tropical Cyclone #10 is off the Georgia/South Carolina border and has a 90% chance of becoming a tropical depression or named storm Irma over the next 48 hours.  The overall impact to the insurance industry should be minimal as the Outer Banks would be the only area to experience tropical storm force conditions as it races out over the north Atlantic this week.

The next area of tropical trouble is off Africa this weekend and is currently south of the Cape Verde Islands.  This system is looking healthy and unlike many of the other African waves the dry dusty air currently is not a factor.   I expect that a depression could form later this week.

The long-range models take this system across the Atlantic over the next 10 days. The general consensus is that it would follow a track similar to Garet and approach the East Coast of the U.S.  However, as stated many times this season as tropical systems get closer to the U.S. overall conditions get better for strengthening.

Hurricane Risk from now until September 11th. Again after this tropical wave I think the overall activity shuts down for most of September.


BMS Tropical Update 8/27/2017 12 PM

Bad Situation Getting Worse

Another day that I am overwhelmed with data and the impacts from Harvey and unfortunately today the data is suggesting the rainfall amounts are likely higher than what was predicted to fall over the seven-day period starting this past Thursday.  I will once again try to focus on insurance industry impacts from Harvey which has now turned to major flooding in the Houston area and will likely be one of the costliest flood events in U.S History.

Harvey Position and Forecast

Harvey, now a tropical storm with winds of 40 mph, is currently centered near Yorktown, TX or about 70 miles east southeast of San Antonio, TX and moving at 1 mph to the southeast.  Harvey is still not forecasted to move out of Texas until later next week, therefore days of rain are still expected on top of what has already fallen.  The general consensus is that Harvey will weaken to a tropical depression and move slowly back towards the coastline and work its way up the Texas coastline toward the Texas Louisiana border.  Some models suggest Harvey could move back over the Gulf of Mexico where it could regain some strength as it moves toward the border, but the wind damage threat is mostly over for Harvey and any wind damage would come from tornadoes or isolated thunderstorm.

Because Harvey will continue to draw in warm Gulf of Mexico moisture, its rain shield will continue move to the northeast along its track.  The heaviest rain will move into Louisiana as Harvey moves to the Northeast over the next several days.

Radar Image showing that moisture trail feeding heavy rainfall into the Houston area.


I only briefly touched on the tornado threat leading up to Harvey’s landfall.  So far Harvey has spawned 20 confirmed tornadoes and some of these tornadoes have cause insured loss and more tornadoes can be expected as Harvey continues to impact the area over the next few days.

Areas of Tornado Yesterday are represented by Tornado Warnings issued by the NWS

Flood & Rain

Since Monday the forecast have called for over 20” of rain to fall from Harvey as it stalls out along the Texas coastline.  These forecasts continue to show that another 26” of rain could still fall over the area over the next 7 days.   This means that Harvey will likely go down as one of the heaviest rainfall events form a Named Storm in history as total rainfall reaches over 50”. This is very impressive considering Texas is no stranger to very heavy rainfall events. Just last year the Tax Day flooding event in Houston area caused widespread flooding, dumping as much as 23.5″ in 14.5 hours in Pattison, TX and causing insured losses of $584 million dollars according to PCS.

Tropical Storm Amelia in 1978 still holds the record for most rainfall over the state from a named storm as it dumped 48” over a five-day period and Harvey could just surpass this event before the end of the week.

Tropical storm Claudette 1979 still remains the twenty-four-hour rainfall record for any location in the United States with 42” of rain falling.     

More recently Allison 2001 caused massive flooding in the Houston area.

The rain continues to fall from Harvey, below is the most recent rainfall totals as viewed in BMS iVision in its severe weather module which has rainfall maps over the last seven days.



Radar estimated storm total for Harvey so far and the areas of major river flooding.  Basically any body of water is likely to see record flooding.Below is a Free Flood mapping tool created by FM Global that can help determine areas of flood hazard. This is not a flood map for Harvey, but it’s a safe bet any areas shown as flood areas are or will be flooded from Harvey

Insured Loss

It is now easy to understand that Harvey will easily reach well over $10 billion USD in economic loss.  The winds and surge losses are still being assessed and given the footprint of affected areas, it is going to take weeks until full assessments are gathered to understand the totality of the damage.   However, from what we know now the general consensus seem the wind related losses will likely be in excess of $2 Billion in insured loss with a much higher economic loss. If you combine this with what historical flood events have done just to the Houston area in the past (Allison 2001 $3.4B adjusted – $10B economic) insured losses are likely in excess of $5 billion dollars.   In fact, for more perspective the catastrophic flooding that occurred in Louisiana last year (August 11 – 15) caused $1 Billion in insured loss according to PCS.   That flooding occurred over an area with about 850K people.  Houston has 7 times more population being the fourth largest city in the U.S. and the rainfall totals will be higher than what occurred in Louisiana last year.

In all the figures above I have not even accounted for NIFP losses which will just add to the expected losses.  As indicated earlier in the week flood coverage is highest along the Texas coastline, however, according to NIFP many areas along the coastline still fall below 20%  coverage.

Clearly this could likely be one of the largest NFIP flood payout in recorded history which will require a major relief bill.  It will likely have major impacts to the flood reform that will be discussed in Washington over the next several months.

Storm off the East Coast

I expect the storm off the Southeast Coast Invest 92L  to develop into a named storm (Irma) but this will not be a threat to the U.S. Coastline.

There is a medium chance of more storm development off the coast of Africa later this week.


BMS Tropical Update 8/26/2017 12 pm CDT

What a Week

What a week meteorologically that shows the opposite ends of how great and devastating nature can be.  The week started with a rare eclipse that crossed the U.S. and ended with a tropical system that went from a depression to a Category 4 in 48 hours.

With daylight we are starting to see the initial impacts from a major hurricane landfall that made landfall at 10 pm CDT last night .  Given that this is the first major hurricane in the information mobile age there is just a wealth of information to take in that is almost overwhelming.  In fact, since this is the first major category 4 to make landfall since Hurricane Charley in 2004 it will likely provide very valuable data to the insurance industry, which has a lack of datasets for such events.  From mobile weather stations and surge devices being deployed ahead of landfall to mobile radar and drone aircraft, this will no doubt be one of the best documented major hurricane landfalls that has ever occurred.

However, I can’t stress enough that the event is not over. It is just starting, and as forecasted, the storm has slowed and is still expected to stall out over the next week and is expected to dump feet of rain along the Texas Coastline.

Harvey Landfall Details

Hurricane Harvey is now only the sixth Category 4+ landfall in Texas weather history (since 1851), and the 2nd in the last 100 years.  Below are the top ten strongest hurricanes at landfall and Harvey is the strongest since Charley in 2004, but likely won’t make it into the top ten in terms of wind and will be 14th in terms of pressure.

List of the strongest hurricane by wind speed to make landfall in U.S. Harvey will not make this top ten list.

In fact, what is more remarkable is from 1926-1969 (44 years) 14 Category 4 U.S. landfalls occurred, but since 1970-2017 (46+ years) only 4 Cat 4+ landfalls have occurred.  This is a decrease greater than 70%.  Statistically the 12 years without a major hurricane landfall will be very hard to replicate, but let’s hope this does not start a trend in the opposite direction.

As pointed out the building standards in Texas are not as good as other states, but the fact is even the best built homes are not designed to withstand category 4 winds. The heaviest wind damage from Harvey appears to be in Rockport, Fulton  area.  Thankfully Harvey’s highest winds likely occurred over Matagorda Island State Park and Aransas National Wildlife Refuge, where there are very few structures or population. This will no doubt limit wind damage form Harvey.

Area of landfall that shows overall luck played a roll in limited damage as a state park likely took on the strongest winds. Any other island like this along the U.S. coastline would be full of homes.


If Harvey would have tracked just 20 miles further south it would be a completely different story in Corpus Christi, which has sustained damage to buildings, but not like the Rockport area.

It would be rare for a surface observation to verify Harvey category 4 winds, but category 3 winds have been verified.  In fact, at 8:48pm CDT, the C-MAN station at Port Aransas, TX sustained 10 meters winds of 114mph with gusts to 131mph. Some other notable wind speed gusts were 110 mph at Copano Bay, 108 mph in Rockport and 102 mph in Aransas Wildlife Refuge, but as mobile station data is collected and communication is restored, more data will be collected providing even more high wind readings.  Another interesting observation is that the pressure gradient was very steep, with an air pressure of 994mb in Sinton, TX and 943mb or so in Rockport, TX  which is a difference of 51mb over 30 miles. This is why the winds were so strong but also indicates that the radius of maximum winds was quite narrow which will also limit damage over a large area.  However, it is known there are many high value homes in the area as lots of folks from Austin and San Antonio like going to the beach on weekends.


BMS iVision Max Wind Gust over the last 24 hours.

BMS iVision Max 1 Min Sustained Wind Speed showing the swath of Harvey and track. Notice the limited inland impacts from high winds.

The exact central pressure has yet to be verified as there are several storm chasers that recorded a pressure of 938 mb.  One notable chaser “icyclone” who likely holds the world record for a human to experience so many hurricanes worldwide said on twitter that this was “one of the worst I’ve been in”.  His calibrated pressure reading was 940.8 mb at 10:31 pm.  This could be very important as once again these readings could influence payout in the Cat Bond market, in particular the newly revised  IBRD / FONDEN 2017  Mexico’s Fund for Natural Disasters Bond that was revised to cover points in North Texas this go around, however, at this early stage it is unclear if the center crossed the box needed with a pressure of 930 mb.

Insured Loss

As mentioned above there are several cat bonds at risk of payout.  However, it is still too early to determine what the exact insured loss will be as the event is still unfolding with days of heavy rain yet to occur.  And with heavy rain still to fall in Harris County this will likely determine the overall event impact with a high percentages of flood-prone properties there.  It has been suggested that the LiveCat market was predicting a payout attachement point of $10B U.S. dollars, however Harvey strengthened right up until landfall and will stall out making modeling of this event very difficult.  The modeling that was provided yesterday will likely better account for the strengthening at landfall and new loss estimate will likely be revised upwards because of this.

Storm Surge

At this point in time the storm surge has not been as high as predicted as Harvey just did not have the overall size to cause devastating surge similar to Hurricane Ike, this will also limit some insured losses.

Various Storm Surge Readings from the landfall area.

Flooding Rains

As mentioned all week, flooding rain will be a major problem with Harvey stalling out.   The latest storm summary from the national weather service suggests rainfall totals are already approaching 15″ with several other spots closing in on 10″.


This is the latest 7 day forecast rainfall total.


There is no change in Harvey’s forecast.   Harvey will slowly spin in the same spot for the next several days.   The option of a second landfall later next week are slowly dwindling off the table, but it can’t be ruled out at this moment as pointed out by the ECMWF model below

BMS Tropical Update 8/25/2017 12 pm CDT

There is now a plethora of information regarding Harvey’s impact but in this update I will attempt to focus on the specific impact to the insurance industry. For exact weather detail, with the most up to date information, it is always best to follow the watches, warning, bulletins and advisories from the National Hurricane Center and the local National Weather Service offices.

Harvey‘s Hazard Affecting Land
• Still very much a serious multi-day situation.
• Hurricane conditions to start later today with landfall between 12 am and 2 am local time.
• High-end Category 3 winds of 115 – 129 mph are expected.
• Current forecast suggests relatively narrow areas of maximum winds 60 miles across are expected at landfall.
• 6 to 12 feet of storm surge will be generated from Padre Island National Seashore to Sargent, TX.

Historic rainfall is expected. 7” of rain could extend into other parts of Texas and the lower Mississippi Valley. The Texas coastal area is expected to get at least 20” or rain and some places could see 35” of rain over the next 7 days.

In summary, Harvey combines the worse attributes of nasty recent Texas named storms: the storm surge from Ike in 2008, strong winds from Bret in 1999 and the record rainfall form Allison in 2001 – all in one storm.

Synopsis of Harvey’s Track and Intensity

As suggested all week Harvey would likely keep intensifying until landfall and is still expected to be a high-end major category 3 hurricane at landfall, which now can be narrowed down to a landfall area between Corpus Christi and Port O’Connor, TX.
This morning a hurricane hunter aircraft dropped a probe in front of Harvey’s path and measured the sea surface temperature at 80 degrees, not at the surface, but at 226 feet deep! Which means Harvey will continue to travel over very warm Gulf waters until landfall. Some model are still suggesting a category 4 hurricane so this can’t be ruled out before landfall, but becoming unlikely. The most recent forecasts suggest a category 3 hurricane with winds 111 mph – 129 mph. One factor to watch is if frictional land effects cause the storm to tighten before landfall due to the cyclonically curved coastline of Texas. One thing is for certain Harvey is going to be a slow-moving storm, meandering over central Texas days after landfall which will exacerbate the inland flood impacts. There are still a handful of models that take Harvey back out over the Gulf of Mexico early next week and have it making a second landfall near the Texas / Louisiana boarder as a tropical storm or even a weak hurricane. Overall, over the next few days Harvey is expected to meander northeast.

Storm Surge Impacts

This is the Advisory 20 Hurricane Harvey Storm Surge Potential. Storm surge is a very detailed hazard and these maps provide a detailed view of expected water level at a street level.

Recently CoreLogic provided an estimate of Texas storm surge risk. However, it is important to note CoreLogic is not forecasting impacts direct from Harvey, rather they have calculated the total homes at risk and reconstruction cost value of key regions along the coastline. In the key landfall locations that could experience category 3 storm surge just under 40,000 risks are exposed.

Table from CoreLogic Harvey Storm Surge Report

Images from social media already show water at property level so flood losses will be very high for areas between Corpus Christi to Matagorda, TX which is expected to see as much as 12 feet of storm surge. Remember Ike and the Bolivar Peninsula?

Wind Impacts
BMS clients have access to high resolution Verisk Climate wind swaths that provide 1 min and 3-second wind speed and durations for not only the forecasted part of Harvey, but also what was experienced. Currently, a 60 mile wide swath of 110 mph to 129 mph 3-second wind gusts are forecasted, with a few pockets of 130 to 149 mph 3-second wind gusts.

It should also be noted that tornadoes can be expected inland which is often an overlooked hazard form landfalling hurricanes.
With AIR-Worldwide estimating that 80% of the residential construction is wood frame and given these very strong winds that are forecasted, it is very important to look at the building codes for Texas. According to the IBHS DisasterSaftey publication, Texas has a relatively poor building assessment compared to other coastal states, but TWIA requires policyholders to meet 2006 IRC standards, so coastal standards are likely higher because of this.

NWS expected Rainfall Amount over next 7 days

Rainfall & Flood
For several days now the forecasts have suggested several days of heavy rain as Harvey stalls out. The official National Weather Service forecast calls for at least 20” or rain over the next 7 days with isolated 35” possible in some coastal areas.Over 10” could occur stretching from San Antonio, Austin, Collage Station, and into Western Louisiana. I expect wind driven rain to be a major factor over much of the Texas coastline which will also be a factor for increased insured losses.

To get a great understanding of the areas that might be prone to inland flooding, I suggest checking out FM Global’s Flood Map which shows displays high (100-year) and moderate (500-year) hazard flood zones via a 90 meter x 90 meter grid.

FM Global inland Flood Map

Power Outage
The following is a great source for live prediction of power outages from Hurricane Harvey. I expect some coastal locations to be without power for several weeks.

Insured losses
It is way too early to estimate insured losses that could be caused by Harvey. However, it is safe to say this will be a multi-billion dollar event. Historical analog storms, like Carla 1961 and Celia of 1970, have losses in the range of $5bn to $20bn but neither of those storms stalled out for days producing the rain that is forecasted to fall with Harvey. This brings me to the last point. The catastrophe models likely have very few events that will match Harvey’s impact. In fact, using just one of the leading modeling company’s stochastic track sets, I could only find seven events out of over 45 thousand that even come close to having a similar track that is forecasted for Harvey, and not one of those seven have the correct event parameters as to what is expected from Harvey. So at this time there is significant uncertainty and little confidence in any early insured loss estimates.

BMS selected events from one of the Cat Models. Shows only 7 events with a similar track to what Harvey is forecated over the next 5 days. None of the events match the intensity.



Nature is beautiful, but its destructive fury is not. Be safe. It’s going to be a tough few days.  Now check out this satellite loop below.