partner login

BMS News

BMS Atlantic Hurricane Seasonal Forecast

The Official Start to Atlantic Hurricane Season?

In theory, the Atlantic hurricane season started late last week with the naming of subtropical storm Alberto. However, interestingly, prior to 2002, subtropical storms were not given names, even though the National Hurricane Center would issue forecasts and warnings for them. Since 2002, such storms have been given names from the tropical cyclone list (Gustav in 2002 was the first). Taking that into consideration, and since no tropical storm days and no Accumulated Cyclone Energy (ACE) have been recorded yet this year, the season, perhaps, has not quite begun.

Regardless, June 1 is the official start of the season and, therefore, providing a bit more detail in regard to expectations for the 2018 Atlantic hurricane season and its impact on the insurance industry is timely. As I mentioned in the closing of the last BMS Tropical Update, there is no relationship between the date of the first Atlantic named storm formation and overall Atlantic seasonal hurricane activity.

Since 1950, there is no relationship between date of first Atlantic named storm formation and overall Atlantic seasonal hurricane activity. Source: CSU

Over 18 different companies and organizations produce seasonal forecasts for the Atlantic hurricane season and are tracked by the Barcelona Super Computing Center. Just last week, NOAA issued its outlook for the season, with a wide range of a 70% likelihood of 10-16 named storms for the 2018 season. When reviewing the historical record, 55% of all hurricane seasons fall into this range, however, which doesn’t provide much informational value to the insurance industry.

Seasonal hurricane forecasts generally come in one of two methodologies: statistical/dynamical models vs. educated guesses from people interpreting the models that produce analog years. The models can be trained and tested against previous years to provide either uncertainty or confidence. The educated guessing is prone to human bias, rules of thumb and big wins or losses. Although there is a general correlation between the greater number of named storms that occur in the Atlantic Basin and the higher chance of landfall, very few forecasts go beyond producing Basin-wide activity to predicting specific impacts or landfall forecasts. To illustrate this point, even if 16 hurricanes were to develop in a season, but none ever hit land, the hurricane season would be classified as above-average, but one that costs the insurance industry nothing in paid losses. Conversely, if three hurricanes develop and all hit land, it’s a below-average hurricane season that could cost the insurance industry millions. I believe there needs to be some attempt to connect the basin activity with landfalls in order to increase the value of utilizing such forecasts.

Where could storms track this season?

One of the most common climate forcers used for seasonal Atlantic Hurricane activity is the El Niño Southern Oscillation (ENSO) and, this year, there appears to be agreement that a weak El Niño event will take hold during the summer, which historically suppresses hurricane activity in the Main Development Region (MDR) of the Atlantic Basin due to higher wind shear. If the El Niño is weak, it could shift development of named storms further north and into the western part of the Atlantic basin near the U.S. East Coast thus increasing landfall chances. Historically El Niño, has little to do with activity in the Western Caribbean / Gulf of Mexico, and any activity in this region increases the chances of U.S. landfall.

Various models from IRI and NOAA CPC on the SST prediction in the Central Pacific ENSO region 3.4. Notice the ASO forecast suggest a weak El Niño should be developed with SST anomaly at 0.5C.

View from NOAA CPC ENSO analysis of subsurface temperature anomalies across the Central Pacific showing overall anomaly below the surface at 150m.  This warm water should rise to the surface over the next few months building the chances of an El Niño during the season.

The subsurface water temperature anomalies across the Central Pacific are warm, suggesting a developing El Niño similar to a Modoki-like Niño, which also points to higher landfall chances during a hurricane season.

Yet another climate forcer taking center stage in early season hurricane forecasts is the cool Atlantic Sea Surface Temperature (SST). In fact, the current North Atlantic SST anomaly pattern looks to be the opposite of an SST pattern which, historically, would be associated with an active Atlantic hurricane season, and could potentially be the coldest since the summer of 2015. However, in the plot below, the warmest SST relative to the average temperature is observed above 30 degrees north latitude and warms closer to the eastern coast of the U.S. This type of pattern suggests that development of stronger named storms might not occur this season in the MDR, but instead north and closer to the U.S. coastline or in the Bermuda Triangle Region of the Atlantic. It should be noted, however, that the MDR Atlantic SST will ultimately trend warmer; almost every year since 1995, the general SST in the MDR has warmed by August regardless of how cold it was in May.

Left is the current SST anomaly and the right the current SST. Reminder, named storms need at least 26C to maintain development.  Currently these types of temperatures are hard to find and likely not a great depth below the surface.

While acknowledging that global weather patterns can change toward the peak of the hurricane season, we can, at a minimum, surmise what the current named storm tracks might be based on the current position of the Bermuda – Azores high pressure center across the Atlantic Ocean. Named storms will often traverse the Bermuda – Azores, and this might provide some clues about the track of future hurricanes in the Atlantic. In its current state, we can see that storms will steer further west, increasing U.S. landfall chances. But if the high weakens to the east, storms may track off shore.

An enormous blocking atmospheric pressure ridge over the Atlantic Ocean stayed resilient for days, which prevented precipitation systems from propagating east. Source NASA GES DISC

Two other climate forcers to watch this Atlantic hurricane season will be Saharan dust and the Madden Julian Oscillation (MJO). Both of these climate forcers are difficult to predict seasonally, but understanding the phases during the season could help determine when storm development will occur. Saharan dust can be an inhibitor of Atlantic hurricane activity, but often moves away from Africa in waves.

Current level of dust off the coast of Africa. Breaks in this dust need to be watched over the given season. Source:

During the breaks in dust, the potential for the genesis of named storms is likely and, if combined with a positive phase of the MJO, could pose a threat to the insurance industry. In fact, I would speculate that the MJO was recently in a state that enhanced tropical convection, which assisted Alberto’s movement across the southwest Caribbean last week. If the MJO was in a favorable state, then we are just moving out of a positive phase for the Atlantic Basin, the next one not occurring for another 45 days. Therefore, this cycle is something to watch closely during the 2018 season, as both of these climate forcers could be the catalysts for named storm development.

The MJO can also considerably influence hurricanes in the Gulf of Mexico, Caribbean Sea, and tropical Atlantic. More hurricanes tend to occur in MJO phases 2 and 3 than in phases 6 and 7. Differences in major hurricane numbers and hurricane days in the main development region are a factor of 3. Source:

It seems that, based on the information above, the forecasts from various organizations show a reduction in the overall numbers of named storms and hurricanes, thus indicating a less active season. However, even in quieter years like 1989 (Hugo) and 1992 (Andrew), there can be storms that have great impact on the insurance industry. The climate forcers and analog years (1934, 1951, 1996, 2002, 2006, 2009, and 2011) suggest less strong storms over the MDR, but as named storms or tropical waves track out of the MDR the could become stronger as they more reach northern latitudes.  The climate forcers and analog currently indicate named storms could track closer to the U.S. coastline increasing the chances of insured loss.    However, since 1920 there are only 16 years in which two or more hurricane made landfall in one season that have resulted in two or more landfalls the next season.

Given the state of the MJO and the cooler Atlantic SST, I don’t expect a lot of new named storm activity for the month of June. But, during the first part of the July, the MJO could return to a more favorable state over the Atlantic Basin, which will allow for a reassessment of future hurricane landfall possibilities.

Barcelona Supercomputing Center has a nice summary of the various Atlantic Hurricane Forecasts
NOAA Climatology Products – Avg Start Dates, Return Periods, Develop Origin by Date, Strike Density
CSU U.S. Landfalling Hurricane Probability Project

BMS Tropical Update 5/24/2018

It’s that time of year

We have reached the time of year when the insurance industry starts paying attention to potential tropical troubles. In fact, over the last two weeks many media outlets have been focusing on possible development in the Gulf of Mexico. There was a weak area of disturbed weather that brought tropical rains to much of Florida last week, and now there is a new weather disturbance in the western Caribbean that is forecasted to move into the Gulf of Mexico for the upcoming Memorial Day weekend.

How common is a named storm in May?
In each of the last three years, the Atlantic Basin has seen some form of named storm development prior to May 24, so we are bucking the current trend with no named storm development yet in 2018. However, with the official start of the Atlantic hurricane season being June 1, it’s not too far-fetched to expect some type of named storm development by the end of May. In fact, 32 named storms have occurred during the month of May since 1851, with most of them forming in the western Atlantic or Caribbean and just a handful in the Gulf of Mexico. Of these 32 named storms, only 4 went on to become hurricanes, with no hurricanes actually making landfall in the U.S. during the month of May.   The table from the NOAA Hurricane Research Division website counts the number of tropical storms and hurricanes, and it is clear that May storms are somewhat rare but not impossible.

This table shows the total and average number of tropical storms, and those which became hurricanes, by month, for the period 1851-2015. It also shows the monthly total and average number of hurricanes to strike the U.S. since 1851.

Is Alberto brewing for Memorial Day weekend?
Alberto would be the first name of the Atlantic hurricane season if the current weather disturbance off the Yucatan Peninsula gets named. However, regardless of whether this particular system gets a name or not, the threat of heavy rainfall is significant. The latest NOAA Weather Prediction Center 5-day Quantitative Precipitation Forecast suggests another 3 to 5 inches of rainfall for Florida and parts of the southeastern United States. Some parts of Florida have already received up to 5 to 12 inches of rainfall over the period from May 13 to May 25, so the soil is already saturated over much of the area which could raise flooding concerns for the insurance industry.

This product from the National Weather Service is the Flash Flood Guidance. This is an estimate of the amount of rainfall required over a given area in a given duration to cause flash flooding to small streams and rivers.

The current model war between the American (GFS) and the European (ECMWF) weather models currently disagree on how far west or east the disturbance will track as the system moves north into the Gulf of Mexico. The GFS model appears to be moving the system over the Florida Peninsula, with the ECMWF model sending it towards the central Gulf Coast and Louisiana. Based on the overall track record of the GFS, the model of choice at this point would be the ECMWF track because of what appears to be ongoing issues with the GFS model being too progressive with developing tropical systems and tracking them to the east. However, given the broad and disorganized nature of the system at this time, there is a fair amount of uncertainty in its eventual track, but given the NHC current probability of formation, a subtropical or tropical depression is likely to form by late Saturday over the southeastern Gulf of Mexico.

Shown above are the  32 Atlantic tropical cyclones during the month of May since 1851.  Most have formed in the western Atlantic or Caribbean, with very few in the Gulf. Here I have outlined where Alberto will likely form this weekend and make landfall during the long weekend.

Over the busy 2017 Atlantic hurricane season, the landfall of hurricane Nate is often forgotten. Nate made landfall around the mouth of the Mississippi River, impacting the Southeast states and resulting in $108 million in insured loss. Very few model runs develop the current disturbance into a hurricane, so the insured impact should be well below any level of loss experienced from Nate last year. Based on the current forecast, gusty winds, large waves and very wet conditions should occur over much of the central Gulf Coast and Florida over the upcoming Memorial Day weekend.

Regardless of whether this system develops into something more significant, the tropical troubles we’re now seeing are a sign that we’re approaching another hurricane season. However, don’t get too anxious due to an early season of tropical development. History has shown us that it tells us nothing about the season ahead. The 2017 season was close to the median of climatology until the end of August, when the season ended up going near the climatological maximum in terms of Accumulated Cyclone Energy (ACE).

ACE as measured during 2017 vs climatology from @wx_Tiger.  There is no correlation to early season activity and how the season will end up.  2017 is a great example of this.

New Hurricane Products for 2017 Season

Every year there are new tools and products that can help the insurance industry understand named storm risks.  In this write-up, I highlight some of these new tools and products for the 2017 hurricane season, which starts today.   For ideas on the type of activity that is expected this season please see my previous update here:  May Tropical Update issued May 15th.

Advisories Will Be Issued Before a Storm Is Named

The National Hurricane Center (NHC), which decides when a named storm gets a name, will issue advisories for tropical systems before the tropical system has a name. In these cases, these tropical systems will pose a threat of bringing tropical storm-force or hurricane-force winds to land areas within 48 hours. For decades these tropical disturbances have been called ‘Invests’, or areas of investigation, and for the last few years the NHC has been giving Invests forecasts related to the chance of formation within the next five days.

Currently some re/insurance contract language is directly related to named storm activity. However, the advisories for these tropical systems that will likely impact land could now lead to earlier activity in re/insurance contracts where coverage is triggered by storm warnings or watches, as the NHC would previously wait for a storm to be named before issuing such warnings or watches. In most cases, these types of named systems would be in a developing stage just off the U.S. coastline and would highlight not only the likelihood of genesis of a named storm, but the possible strength of winds along the coastline.

Hurricane Humberto in 2007 is a classic example of where the NHC would have likely issued watches and warnings before Humberto was named if it took place in 2017. Humberto strengthened into a hurricane in a 24 hour period. Source: AccuWeather Inc. & NOAA

Storm Surge Watches and Warnings Are Coming 

For the last several years, the NHC has made large improvements to storm surge forecasts from named storms.  In fact, the detailed Potential Storm Surge Flooding Maps now rival the high resolution flood model simulations that are becoming common in the insurance industry by various Cat Modeling vendors.  However, another tool that might help the insurance industry is that the NHC plans to release storm surge watch and warning graphics to provide further guidance on where the greatest threat to life and property from a named storm might be.   The insurance industry is all too familiar with the hazards and damaging storm surge that occur with a threatening named storm and this guidance can help pin point areas likely to be most impacted by storm surge.

New storm surge watch and warning product for 2017 hurricane season. Source: NHC

Earliest Reasonable Arrival of Tropical Storm-Force Winds Will Be Forecast

The NHC has always provided guidance as to the position and timing of the center of a named storm; however, to provide more added value, the NHC will now be directly forecasting when tropical storm-force winds will begin to affect land. This will allow the insurance industry to better understand when winds greater than 39 mph are expected, which should aid in allowing more time for an insured to apply preventative measures to mitigate risk from damaging storms. Above 39 mph, winds can make it difficult and even dangerous to be outside continuing preparations for a tropical storm or hurricane.

The IBHS has some helpful hints to reduce hurricane damage to homes and businesses.

The Cone of Uncertainty Will Be Smaller

Every year the NHC reviews the accuracy of their previous five seasons of hurricane forecasts.  This review suggests the forecasts are getting better, and with that, the average error in the NHC forecasts that make up that famous cone of uncertainty will result in a smaller cone and just maybe more certain hurricane forecasts for the 2017 season.  Track errors have gone down over the last 10 years and forecasts have gotten better.  In fact, since 2007, the size of the cone of uncertainty at 120 hours (or five days) has shrunk by more than 35%.  Since last year, the size of the cone at 120 hours has shrunk by more than 10%.

The shrinking cone of uncertainty. Source Brian McNoldy Univ. of Miami

Hurricane Model Wars

As long as I have been studying meteorology there have been wars among the various forecast models as to which model is the most accurate at forecasting named storm activity.  This war was brought into the public limelight after Hurricane Sandy in 2012, when the American Model [Global Forecast System (GFS)] falsely forecasted Sandy to track offshore, and the European ECMWF model correctly predicted Sandy to make landfall in New Jersey.  Just like the various hurricane products described above, the weather models themselves are under ongoing improvements.  One item to watch this hurricane season is the upgrade to the GFS model on July 12th.  As part of these upgrades, the NOAA asks forecasting divisions like the NHC to run performance tests.   Although the tests might suggest better temperature forecasts or precipitation forecasts for different areas of the globe, one area where the upgraded GFS shows deficient skill is with hurricane forecasting.   The report I have seen from the NHC suggests for the 2014-2016 retrospective runs, in comparison to the 2016 GFS model, the 2017 GFS showed a 9-10% degradation in track forecast skill at 48-72h in the Atlantic Basin.  In terms of intensity forecasting, the 2017 GFS showed degradations in intensity forecast skill at nearly every forecast interval out to 120 hours in the Atlantic Basin.  The new GFS model also indicated less run-to-run consistency.  It likewise showed little overall improvements in TC genesis forecasts for the Atlantic Basin.

Evaluation of the proposed 2017 GFS implementation done by the NHC in February 2017

Another troubling factor is that the GFS model will likely have other fallouts with the regional models, such as the HWRF and GFDL, and the statistical models such as the GFEX.

Here is the quote from the NHC about the upgrade to the GFS model:

“The loss of short- to medium-range TC track and intensity forecast skill for the Atlantic basin in the proposed 2017 GFS is unacceptable to the National Hurricane Center.  We are also concerned about the lack of testing of the downstream impact of the 2017 GFS on the regional hurricane models.  Therefore, we oppose this implementation.”

As the insurance industry watches all of the various hurricane forecast model runs to determine where a hurricane might track, it might be good to put a bit more weight in the ECMWF model this season.  However, the ECMWF will also be upgraded on July 11th and very few people know what these upgrades will do to its hurricane forecasting.

Detail of this model upgrade can be found here:

May Tropical Update

Seasonal Hurricane Forecast Skill

Seasonal hurricane forecasts, with varying lead times, have been produced in the Atlantic basin since 1984 by the late Dr. Gray from Colorado State University (CSU).  Partly as a result of the early success of those forecasts, seasonal named storm forecasts are now a dime a dozen with many different research and operational groups making seasonal hurricane forecasts for tropical basins around the world.  Although there are some very brilliant minds making these forecasts, some might argue there has been little skill improvement over the last several years. The insurance industry is accustomed to using climatology to understand the risk of a U.S. landfalling hurricane.  This climatology also suggests, on average, the more active the overall Atlantic basin hurricane season is, the greater the probability of U.S. hurricane landfall which can be useful guidance to the insurance industry heading into the hurricane season. However, when it comes to seasonal landfall forecasts, really no skill has been demonstrated; this is ultimately what matters for the insurance industry.  In fact, there isn’t a forecast group that has skillfully predicted the current drought in major hurricane strikes over the last 11 years.

CSU Atlantic Basin wide hurricane forecasted and observed hurricanes with missed forecasted number shown.

CSU provides easy access to historical forecasts which allows for simple validation of there seasonal forecasts.  In a simple analysis, the CSU forecast has done fairly well with their April predictions when compared to using historic climatology (6.5 hurricanes per year) .  Fourteen CSU forecasts since 1995 have come within three hurricanes of the seasonal hurricane total in the Atlantic Basin.  On seven occasions, they have only missed the cumulative number by one.  They hit the mark in 2008.  Of course, CSU have posted a few wayward forecasts as well: 2005 (undershot the total by eight hurricanes) and 2012 (fell short by six hurricanes).

CSU Atlantic Basin wide major hurricane forecasted and observed hurricanes with missed forecasted number shown.

Regarding the major hurricane forecasts, CSU has made seven forecasts that failed to fall within two storms of the final tally. This should not obscure the fact that they’ve also posted four perfect forecasts of major hurricane occurrence and 13 others that approximated the seasonal total within one storm.
It is very easy for one to sit back and grade objective season hurricane forecasts like CSU, and there is no doubt a lot of work goes into the refinement of these forecasts each year. I am a firm believer that seasonal forecasts and even landfall forecasts will improve over time. Currently, many of these seasonal forecasts are statistical schemes which will inherently fail some years. There is some promise that a hybrid of statistical and dynamical forecast could be the future of seasonal hurricane forecasting.

The Real Value – Landfall Forecasts

As mentioned, the real value to the insurance industry is in getting an accurate picture of potential landfalling storm activity. CSU issues probabilities for landfalling hurricanes across various regions of the U.S. This year, they are forecasting a 75% probability of a hurricane landfall. The probability of landfall for any one location along the coast is very low and reflects the fact that, in any one season, most U.S. coastal areas will not feel the effects of a hurricane no matter how active the individual season is. And, given the variation of exposure along the coast, providing accurate estimated insured loss ahead of season is very difficult if not impossible.
However, given the current state of the science of seasonal forecasting it is possible to provide guidance to which area of the coastline will see a higher than normal probability of landfall. This year, landfall forecasts may prove to be one of the more challenging years. This is due to the large amount of uncertainty of many of the climate forcers used to predict Atlantic named storm activity. One of the most common climate forcers used is El Niño Southern Oscillation (ENSO) and, this year, many forecasts are uncertain if an El Niño event will take hold during the summer, which historically suppresses hurricane activity. CSU thinks an El Niño will be weak or moderate by the peak of the Atlantic hurricane season based on their April forecast. However, as mentioned a few weeks ago, spring El Niño forecasts can be notoriously misleading, and result in difficult forecasts. (This phenomenon is so familiar that it has its own name – the “spring predictability barrier”.) Over the last few months many of the monthly climate models have backed off of the idea of a strong El Niño during the peak of the Atlantic hurricane season and now have a weak to neutral ENSO condition. If the El Niño is weak, it could result in increased landfall chances, and El Niño historically has little to do with any activity in the Western Caribbean / Gulf of Mexico and any activity in this region increases the chances of U.S. landfall.

Official Climate Prediction Center ENSO probability forecast for Mid-April (Top) and Early May (Bottom) Showing in just a half of month the probability of a El Niño during the hurricane season (Gray Box) went form 70% to below 50%. There is an increasing signal that ENSO will be neutral during the hurricane seasons.

The other challenge this year is that earlier this spring, water in the far North Atlantic and water off the coast of Africa had a relatively cold signal which is potentially indicative of a negative phase of the Atlantic Multi-Decadal Oscillation (AMO).  However, over the last month or so, the Atlantic water has warmed substantially.  This continues the debate as to whether the AMO has switched to a negative phase.  The U.S. government suggests the Atlantic is still in a positive state, whereas the CSU data suggests the AMO has trended negatively over the past few years.  If water continues to warm over the main development region, like it has over the last month, it could allow more storm formation in the heart of the hurricane season.

NOAA SST anomaly (degree C) for 4/10 (left) and 5/11 (right) showing a warming of SST off the west coast of Africa off the East Coast of the US. The cold SST in the North Atlantic have expanded and SST after months of above normal in the Gulf Of Mexico have recently turned colder than normal which should be short lived.

Another factor that has had a large influence on named storm activity over the last few years has been the Saharan Air Layer (SAL).  Named storms interact with the SAL in several ways.  Some named storms get embedded in the SAL their entire life-cycle and often struggle to intensify beyond strong tropical storm strength.  Other named storms can be overrun by the faster moving SAL and are quickly weakened.  The SAL in general can hamper convection, making it hard for named storms to develop.  This year could be similar to the last couple of years which would hamper development of named storm activity in the main development region but, as tropical waves get closer to the U.S. and Caribbean, the SAL decreases, and conditions become more favorable for development outside of the SAL.

SST anomaly for 5/11 showing warmer than normal SST around Hawaii.

Not to forget about Hawaii, which is also a target for named storm activity and, like the U.S. mainland, has gone a long time without a major hurricane impact.   This year, the insurance industry should also keep an eye on this 50th state as, just like the warm sea surface temperature (SST) off the East Coast of the U.S., the SSTs are above normal near Hawaii, and any tropical activity that might come close to the islands could be enhanced by these warmer than normal SSTs.

Forecasting of Atlantic named storm activity is not easy, and CSU has demonstrated variability in skill year over year, with no real improvement in the April seasonal forecasts over the last several years, but overall there is skill improvement over using basic climatology. Although some forecasts are currently calling for below normal activity in the Atlantic basin, climate forcers are pointing toward the ideas that these forecasts will be adjusted upward as we approach June 1 to a more active or normal season. The major climate forcers suggest there could be less development in the main development region of the Atlantic but, as tropical waves move closer to the U.S., they could have a better chance at developing, due to lack of wind shear and lower SAL environment. SSTs near the U.S. coastline are also expected to be warmer than normal which would add fuel to any disturbances that develop. Therefore, parts of the U.S. coastline have a higher probability of a landfalling named storm this season, with the Mid-Atlantic and Northeast seeing the highest chance.
Barcelona Supercomputing Center has a nice summary of the various Atlantic Hurricane Forecasts
NOAA Climatology Products – Avg Start Dates, Return Periods, Develop Origin by Date, Strike Density
CSU U.S. Landfalling Hurricane Probability Project

Peak of 2016 Atlantic Hurricane Season Is Approaching

It’s hard to believe, but the Atlantic hurricane season began 189 days ago when Hurricane Alex formed on January 13 and went on to become the strongest hurricane ever to form in the month of January in Atlantic Basin. The early season continued with three landfalling named storms that formed this June, making it easy to assume that an early season means an active season. But an early start to the season does not necessarily mean that the heart of the hurricane season will be active. A current lull in the basin since June 21 and long-range forecasts suggest there is limited opportunity for development for the reminder of July. However, climatology suggests we are not out of the woods yet. Instead, we are only just approaching September 10: the peak of the season when storm formation becomes much more frequent.

This is the time of year when eyes are trained on the massive cloud clusters that move off the West African coastline. These clusters have been limited so far this season. The only movement off this region is massive plumes of Saharan dust.


Current CIMSS Tropical Cyclone Team imagery that is useful for monitoring the position and movement of dry air masses such as the Saharan Air Layer (SAL) and mid-latitude dry air intrusions. Animations of the imagery are useful for tracking these features and can also help identify the source of the dry and/or dusty air that is indicated in the imagery.

In some cases, these dust plumes have traveled all the way to the Texas Gulf Coast. Known as the Saharan Air Layer (SAL), this dry, dusty air has about half the moisture of the typical tropical atmosphere and can discourage the clouds and tropical convection needed for named storm development. Although we still lack a full understanding of how the SAL affects tropical systems, there are likely several reasons why it limits cloud and tropical convection development, and thereby limits the likelihood of named storm development.

  • Dry air can enhance downdrafts (sinking air), suppressing convection around the system.
    As tropical waves move off Africa, any resulting convection quickly reaches the bottom of the SAL (typically at altitudes between 5,000 and 15,000 feet); then entrains dry air which limits further convection.

Here is the current 850 mb temperature normalized anomalies which shows that areas of heavy dust have warmed the atmosphere at this level 3 – 5 degrees Fahrenheit.

  • Typically atmosphere cools with height, but the SAL absorbs sunlight, which retains warmth and creates an inversion for thousands of miles across the Atlantic basin. This essentially caps the development of the showers and thunderstorms that are needed in tropical cyclone development.
  • Dusty conditions can be enhanced by stronger easterly winds that increase wind shear and tilt or outright displace the convection aloft from low-level circulation, thus limiting convection and tropical cyclone development.
  • The SAL shields sea surfaces from the sun and can keep the sea surface temperatures cooler than normal across the main development region.

Data suggests decadal variability in the SALs that may impact tropical activity (unfortunately, since the insurance industry doesn’t need another multi-year decadal pattern that could influence tropical cyclone development). Records going back to the 1960s and 1970s that were collected by satellites and island stations (using dust as a tracer) show that SAL activity have ebbed and flowed over the years. In the 1980s, some studies point to quiet periods that coincided with a stretch of increased dust outbreaks. In the 1990’s, dust activity decreased and tropical cyclone activity began ramping up.
At this time it is difficult to say if the dry, dusty air will continue into the heart of the Atlantic Hurricane season which effectively peaks around September 10th. We are only one-third of our way through hurricane season, and June and July are not usually good indicators of what is to come: on average, those months account for only 4% of Atlantic major hurricane activity.


— Philip Klotzbach (@philklotzbach) July 20, 2016

And just as a reminder, the 2004 named storm season didn’t have its first named storm form until August 1, but 2004 ultimately became one of the most active seasons on record for U.S. Florida landfalls.

The insurance industry shouldn’t let its guard down. If the dust persists and Cape Verde storms are hampered by the SAL, storms may instead develop in the western Atlantic which provides a higher likelihood of making U.S. landfall. And given the Western Caribbean and Gulf of Mexico is experiencing record warm surface temperatures and heat content, there is plenty of energy for these storms to become powerful hurricanes should they track over such warm waters.


Total oceanic heat content (called the Tropical Cyclone Heat Potential, or TCHP) in kilojoules per square centimeter (kJ/cm^2), for July 15 for the years 2005 – 2016. TCHP was at near-record or record values over much of the Caribbean, Gulf of Mexico, and waters surrounding the Bahamas in July 2016. TCHP in excess of 90 kJ/cm^2 (orange colors) is commonly associated with rapid intensification of hurricanes. Image credit: NOAA/AOML.

In summary, El Niño is gone and La Niña conditions are slowly building in the Central Pacific, and the SAL is currently hampering development in the main development region of the Atlantic. Long range forecast models remain quiet with little development chances. However, the warm ocean heat content and sea surface temperatures near the U.S. coastline would provide plenty of fuel for a strong hurricane if one were to track over these waters.

Already off to a head start – Hurricane season officially begins June 1

After two years of below normal named storm activity in the Atlantic basin most prognostication suggests that the 2016 Atlantic hurricane season will be above average. Many of these forecasts are citing a weakening El Niño and warmer than average seas surface temperatures over much of the Atlantic basin as a reason to expect conditions would slightly favor more named storm activity.

So far the prediction of an active season seems to be holding, with named storm Alex forming in January and most recently tropical storm Bonnie making landfall in South Carolina this past Memorial Day weekend. In fact, 2016 now joins the year 2012 as the only years, since reliable satellite coverage began; to have two named storms form prior to June 1.


Univ. of Miami – Calendar of activity during the Atlantic Hurricane Season

Historically, the average number of named storms to develop in a season is 12, with six reaching hurricane status and three becoming major hurricanes. This year the overall number of named storms will likely be higher due to the lack of El Niño that tends to dampen hurricane formation in the Atlantic by increasing wind shear. In an environment with high wind shear, weak tropical disturbances have a more difficult time reaching into the atmosphere and forming into named storms. A La Niña looks to be rapidly developing and historically this climate forcer has lead to an above normal Atlantic hurricane season, which increase the chances of U.S. landfall named storm activity.


Like most hurricane seasons there are a few wild cards that could factor into more or less storm activity on top of general El Niño/ La Niña influences. One of these factors would be the Atlantic Multidecadal Oscillation (AMO) which has two phases: —a high hurricane activity phase and a low hurricane activity phase. Given the decrease of hurricane activity over the past few years, forecasters are uncertain as to whether the high-activity phase, which began in 1995, has ended. If it has, we could continue to see fewer hurricanes despite the otherwise favorable conditions of La Niña and warm ocean water.

Atlantic mutidecadal oscillation from 1950 to 2015

Atlantic mutidecadal oscillation from 1950 to 2015

Another wildcard could be the Saharan Air Layer (SAL) which is a mass of very dry, dusty air which forms over the Sahara Desert during the late spring, summer, and early fall and usually moves out over the tropical North Atlantic Ocean every 3-5 days. The SAL can have a significant negative impact on tropical cyclone intensity and formation. Its dry air can act to weaken a tropical cyclone by promoting downdrafts around the storm, while its associated strong winds can substantially increase the vertical wind shear in and around the storm environment also making it difficult for storm formation. The SAL can cover an area the size of the continental U.S. and has been tracked as far west as the Caribbean Sea, Central America, and the Gulf of Mexico.

Current position and movement of dry air masses (SAL) (Orange / Red mean dry air)

Lack of Named storm Landfall forecasts

Many of the seasonal outlooks don’t predict how many storms could make landfall, but there has been a dearth of major hurricane landfalls over nearly the past decade, and even more so the lack of any hurricane landfall of any intensity in Florida since Wilma in 2005. In that time, more than 2.5 million people have moved to Florida who might have no experience preparing for or responding to a landfalling hurricane.

To get an idea of possible landfall areas for this up-coming Atlantic hurricane season some seasonal forecasts use current atmospheric and global sea surface temperature patterns to put together analogs years of past storm tracks. These analog years (1988, 1995, 1998, 2007, 2010, 2012) provide ideas of where storms have tracked when past condition were similar to this years conditions. This year it would appear there will be more named storm activity forming closer the U.S. coastline vs storms forming in the middle of the Atlantic Ocean in the main development region. This would yield a higher chance of named storm / hurricane landfall. The analog years point toward more storm activity in the Western Caribbean Sea which could increase the chances of a storm tracking into the Gulf of Mexico as well.


Storm track maps of the current analog years


In summary NOAA’s outlook is in line with those from other organizations, both academic and private. Last month, Colorado State University research scientist Phil Klotzblach issued an outlook for a nearly-average season. Britain’s Met Office predicts a slightly above average season, as does private weather companies like WeatherBELL Analytics and WSI. The London-based weather consortium Tropical Storm Risk is forecasting a season 40 percent more active than the past 10 years, with 17 named storms, nine hurricanes and four major hurricanes. On the lower side would be the algorithm derived by University of Colorado Boulder which suggests only 6-12 named storms this season.

Fading El Niño – What’s Next For Insurance Industry?

A Q1 and El Niño wrap up

With winter and Q1, 2016, behind us, the insurance industry can review the active weather pattern and resulting insured losses. As mentioned in my blog post last fall, Florida experienced lots of weather activity, which is typical during strong El Niño winters such as this past one.  However, overall insured losses, while not historic, haven’t been benign either: Q1 losses ran about 31% above the 10-year average insured loss according to my estimates with Texas and the Southern Gulf states taking the brunt of the insured losses.  Although the remarkable blizzard (January 22–24) resulted in limited insured impacts based on the Property Claim Services (PCS) initial estimate, winter storms increased losses in the West.  In fact, the PCS issued four separate bulletins for the state of California more than any other year during the last 10 winter seasons.  Examples like this highlight the relatively predictable impacts of an El Niño winter.

After virtually tying the record for the strongest El Niño (as defined by a three-month running mean sea-surface temperature anomaly in the so-called Niño 3.4 region of the central and eastern equatorial Pacific Ocean), sea-surface temperatures (SST) are steadily cooling. NOAA’s March El Niño outlook suggests this El Niño may be all but gone by late spring or early summer.  In fact, some climate models and a recent government outlook suggest a shift to its opposite, La Niña by this fall.  As a result, the insurance industry needs to consider the potential for higher losses which are often associated with the La Niña phenomenon.


Looking below the surface in the ocean waters in the Central Pacific, you can also see a trend of colder-than-average water working its way eastward across the International Date Line, eating away at the warmer-than-average equatorial Pacific water from below – another sign of a weakening El Niño.  Source: CPC

The unpredictable El Niño-La Niña relationship

Scatterplot showing the relationship in El Niño / La Niña states from one year to the next, for every year since 1950 in which an El Niño occurred. Each dot represents a pair of “year 1 vs. year 2” El Niño / La Niña states. In general, the stronger the El Niño (higher values on the x-axis), the stronger the subsequent La Niña (lower values on the y-axis). For more details and a larger version of the graphic, see the associated ENSO Blog post

Scatterplot showing the relationship in El Niño / La Niña states from one year to the next, for every year since 1950 in which an El Niño occurred. Each dot represents a pair of “year 1 vs. year 2” El Niño / La Niña states. In general, the stronger the El Niño (higher values on the x-axis), the stronger the subsequent La Niña (lower values on the y-axis). For more details and a larger version of the graphic, see the associated ENSO Blog post

El Niño and La Niña events each typically last for only 9-12 months, and they typically recur every 2-7 years, according to Columbia University’s International Research Institute for Climate and Society.  Flip-flops from a strong El Niño to La Niña are not unusual. For example, the record-setting El Niño of 1997-98 was almost immediately followed by La Niña the following summer, reaching moderate-to-strong intensity before finally ending in Spring, 2001. A similar pattern followed the strong El Niño of 1972-73.  However, neutral conditions followed three other strong El Niño’s that occurred in 1982-83, 1965-66 and 1957-58.

What Does It All Mean for the Insurance Industry?

Expect higher losses! If La Niña develops, historical insurance industry losses suggest worldwide impacts, but likely these impacts won’t be fully felt until 2017 when the La Niña is fully developed.  However, there are some broad trends that have shown up in past weakening El Niño events that could give a hint on what to expect in 2016.

Severe Weather Season

Various climate forcers such as North American snowpack, Pacific Decadal Oscillation and Gulf of Mexico SST can influence North American severe weather, muddling the impacts that El Niño or La Niña might have. Disregarding these factors and only looking at past weakening El Niño events that transitioned to La Niña, the data suggests weather will vary based on location. A moderate-to-strong La Niña tends to promote severe weather across the Southeast (conversely, El Niño promotes severe weather in Florida, the high plains of Texas, and up through Colorado to Minnesota).

Although insurance loss data suggest on average a La Niña year sees about double the insured loss that might occur during El Niño years, the reason for increased losses may have more to do with the location of the losses than the severity of the storms. La Niña years seem to favor stronger storms over the Southeast (Dixi Alley). This area is densely populated, which may lead to more claims of severe weather.  But like with any weather peril, even a quiet year can have an EF-5 roll into a city and cause devastation.  In the end, insured losses are largely based on good luck or bad luck.  Example: if the Joplin tornado of 2011 had shifted a few miles south during that La Niña year, no one would talk about it today.

This severe weather season the key could be the summer heat expected in the central Plains and Great Lakes could mix with the wet spring expected in Texas and Louisiana and increase instability leading to storms in the Northern Plains and Ohio River Valley.

Atlantic Hurricane Season 2016

Much hype accompanied last hurricane season due to El Niño’s tendency to produce stronger wind shear, which tends to tear apart developing or mature tropical cyclones and result in less tropical development. Sure enough, June through October, 2015, Caribbean wind shear was the highest on record since 1979, according to Dr. Phil Klotzbach, tropical scientist at Colorado State University (CSU). The team at CSU has also in the past pointed out that the timing makes a big difference:  If a transition to La Niña happens late in the year, it’s less likely to influence the Atlantic hurricane season.  The forecast team at CSU will issue their first outlook for the 2016 Atlantic season on April 14.

However, the CSU team and other leading researches have pointed out that U.S. hurricane impact rises dramatically in a La Niña or neutral season compared to an El Niño season. On an annualized basis since 1950, major hurricane landfall rates during La Niña years are 20% higher than neutral conditions and almost 280% higher than El Niño rates.


General stats of past La Niña and El Niño Atlantic hurricane seasons and the number of landfalls with adjusted historical total insured losses when accounting for all hurricane and tropical storm impacts for those years.

With El Niño potentially vanishing by the start of the 2016 hurricane season, the chance increases for tropical cyclones surviving to make U.S. landfall. If El Niño was the only factor, that is. I expect the team at CSU to discuss El Niño / La Niña, but also discuss the possible switch to a cool mode of the Atlantic Multidecadal Oscillation (AMO), which might suggest less storm activity in the next decade. This is because the far North Atlantic has been quite cold for about three years, and the SST pattern continues to cool (evolving as it did in the early 1960’s [the last time the AMO switched from a warm phase to cool phase with SST cooling in the North Atlantic and a slow progression of colder anomalies propagating into the tropical Atlantic and with warmer anomalies hanging on closer to the U.S. coastline).

The odds may shift a bit toward a more active Atlantic hurricane season in 2016, but El Niño’s absence doesn’t guarantee that outcome. Since the 2006 season, there have been some very active seasons with very few landfalling hurricanes, and the insurance industry still awaits that major hurricane landfall.


El Niño, La Niña, or the lack of either (known as the neutral phase), is only one large-scale forcing on the atmosphere. Its presence or absence does not definitely determine severe weather or hurricane.  Climate models indicate a La Niña will follow the recent powerful El Niño, and we can look at past weather patterns to speculate future impact on particular insurance portfolios.  Right now the best analog years would be 1988, 1995, 1998, 2007, and 2010 during those years PCS losses averaged 10.6B, but what is more important is using those years to understand where the severe weather and hurricanes occurred to get an ideas of what might occur this year.  For example: U.S. landfalling hurricanes were limited, but in almost all those year the western Caribbean and Gulf of Mexico experienced some named storm activity.


Historical hurricane tracks of the 1988, 1995, 1998, 2007, and 2010 hurricane seasons  Source: NOAA Historical Hurricane Tracks

Climate forcers like El Niño and La Niña can help predict the frequency of overall extreme weather activity, but truthfully, long-term predictions about the number of named storms, location of landfall or the power of other severe weather is impossible. The best way for the insurance industry to prepare is to carefully consider the risks and their potential impact. BMS’ weather risk management module in iVision can help carriers better understand their risk and manage portfolio accumulation in areas prone to hurricanes and severe weather. iVision also has tools to track forecasted hurricanes, including detailed hurricane wind fields. It has several severe variables around severe weather which can be combined with hurricane layers to provide a holistic view of an event and help carriers understand the range of potential loss outcomes from extreme weather events.  Learn more about the Hurricane Risk Management Module and Severe Storm Risk Management Module.

BMS Tropical Update Joaquin 10/02/2015 12PM CDT

All week the discussion around the track of hurricane Joaquin has been about the uncertainty. Today the end game for Joaquin is much more certain as Joaquin has now started its northward movement away from the Bahamas where Joaquin grew to an impressive category 4 hurricane. Joaquin is the first category 4 hurricane to hit the Bahamas in October in 149 years (1866) and should easily cause billions in economic loss for the country.

The latest NHC advisory shows Joaquin gaining a bit of latitude, and most forecast model guidance now show that Joaquin will track northeastward away from the U.S. East Coast. There is only a small probability Joaquin will make U.S. landfall, and the NHC cone of uncertainty no longer touches the U.S. coastline. Nantucket, MA, has the highest probability of impact at 14%, with other northeast cities at a lower percentage. Overall, by Monday, Oct 8, Joaquin should be tracking between the U.S. northeast coastline and Bermuda as a weakening hurricane. It should be noted, however, that there is still a possibility that Joaquin could be captured by the coastal nor’easter moving up the coastline this weekend, which could pull Joaquin into New England. But again, the probability of that scenario is low.

There still remains a high flood threat along the East coast as Joaquin funnels tropical moisture northward. Rainfall totals will continue to increase across much of the East Coast with the potential that some locations could experience more than five inches of rain by the end of the weekend with a large part of South Carolina forecasted to experience 10+ inches of rain.


NOAA quantitative precipitation forecast 3 Day Forecast Rainfall amounts.

Minor coastal flooding is already occurring along the East Coast due to strong onshore winds being sandwiched between Joaquin to the south and strong Canadian high pressure to the north. That flooding will likely worsen over the next few days, regardless of the track of Joaquin.

Joaquin has generated more Accumulated Cyclone Energy (ACE ) in 3 days than all other Atlantic storms during the month of September. Joaquin might have been a bit of a surprise given all the talk that this season is an El Niño year and overall activity should be lower than normal. However, Joaquin did not originate in the deep tropics off the coast of Africa where El Niño has its greatest influence. Instead, El Niño had very little influence on Joaquin because of the location of its origin. Joaquin formed at 27.5 degree north latitude and moved southwestward toward the Bahamas where it was able to feed off the very warm waters of the Bahamas.


The Tropical Cyclone Heat Potential anomalies around are among the highest in the Atlantic basin along the East Coast which will support continued genesis of tropical systems if they form

With 59 days left in the Atlantic hurricane season, we need to watch for more storms like Joaquin that form without El Niño’s calming influence. There is very warm water off the East Coast which will continue to cause headaches for the insurance industry right through the winter. These warm waters can strengthen tropical systems like Joaquin and/or stronger nor’easters like we are seeing this weekend. Historical climatology data suggests typical October tropical cyclone development should occur in the western Caribbean, but given the hostile conditions in the deep tropics due to high wind shear from El Niño, the development will likely be closer to the U.S. Coastline and into the Gulf of Mexico.


Climatological areas to see tropical storm development and the likely tracks of that development

BMS Tropical Update Joaquin 10/01/2015 12PM CDT

  “Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones.” – Donald Rumsfeld


The quote above is often ridiculed, but it’s actually a brilliant description of Joaquin. In fact. the insurance and reinsurance industry are all too familiar with knowns and unknowns. Model uncertainty is part of doing business, and it is common knowledge that all catastrophe models have some degree of unknown error. There has been a push by insurance companies to funnel better location-level data into the catastrophe models in an effort to limit the overall model uncertainty in modeling risk at a location level.

Weather models are similar to catastrophe models in terms of uncertainty: uncertainties exist in every model run; but generally, the more detailed the input, the greater the accuracy of the output. Better input helps limit uncertainties. Already this year we anxiously awaited the arrival of a Tropical Storm Erika as a catastrophic hurricane to impact Florida hurricane, only to watch Erika fizzle a couple hundred miles south of Cuba.

View post on

Meteorologist view after every new forecast model run for Hurricane Joaquin

I have highlighted the overall uncertainty in Joaquin’s forecast track all week. As anticipated, we have already seen the model and official forecast for Joaquin change drastically from east to west to back east.  Most arm chair meteorologist are unaware that behind the scenes increasing amounts of data detail are feeding the weather models to try to get a better picture of the overall atmosphere. Extra weather balloons are being launched at sites all over the Eastern U.S and Caribbean. NOAA and the U.S. Air Force have multiple aircraft sampling the environment in and around Joaquin. All this data is being fed into the weather model to hopefully limit overall uncertainty: just like in a catastrophe model.

Ironically, with more detailed input, the overall forecast for Joaquin has become more certain, but yet the overall track options for Joaquin extend past the National Hurricane Center’s (NHC) cone of uncertainty in this case.  In my opinion, Bermuda, which is not even in the cone of uncertainty, is just as likely as Boston to feel Joaquin’s impact.

In the short term, Joaquin will significantly impact the islands of the Bahamas as Joaquin will fluctuate in intensity as a major hurricane for the next 24 – 36 hours. As mentioned yesterday, after 36 hours the door is wide open for several different track paths. A U.S. landfall is still possible, but the new forecast models (that are being fed all this higher resolution data) suggest a more eastward shift in track away from a U.S. landfall.

I consider the ECMWF model (image below -Right) to be very good and reliable, and it illustrates Joaquin moving out to sea and closer to Bermuda. I expect by Friday we will have a much clearer picture of where Joaquin will track this weekend, and with that, insured impacts can start to be calculated.


American (GFS Model) left and the European (ECMWF Model) right ensemble runs which in total is over 75 separate model runs of possible tracks for hurricane Joaquin.

So until the track is known it’s best to focus on what is known:

  • As Joaquin funnels tropical moisture northward, rainfall totals will continue to increase across much of the East Coast regardless of whether the hurricane hits land. Over a foot of rain could occur in some area which  will produce flash flooding.
  • Minor coastal flooding is already occurring along the East Coast from strong onshore winds being funneled between Joaquin to the south and strong Canadian high pressure to the north. That flooding will likely worsen over the next few days, regardless of the track of Joaquin

BMS Tropical Update Joaquin 9/30/2015 12PM CDT

Hurricane Joaquin rapidly intensified overnight and is now a Category 1 hurricane tracking west toward the Bahamas. As I wrote about yesterday, the forecast uncertainty for Joaquin is extremely high. The National Hurricane Center (NHC) has clearly communicated this uncertainty in their forecast discussions which I have quote below.

“Confidence in the details of the track forecast late in the period remains very low, since the environmental steering currents are complex and not being handled in a consistent manner by the models. Given that a wide range of outcomes is possible, it is too soon to say what impacts, if any, Joaquin will have on the United States.”

Further, the Director of the NHC tweeted:

Although uncertainties exist, I think it is safe to warn about what the known impacts will be “IF” Joaquin approaches or makes landfall along the East Coast.

We do know that hurricane Joaquin is now located 215 miles east-northeast of the Central Bahamas and will continue to strengthen into what will likely be a major hurricane off the East Coast by Saturday. In fact, water temperatures near Joaquin are currently at all-time record warm levels and could, if all other factors align, easily support a Category 5 hurricane. In fact, a high-end Category 3 or 4 is now likely for Joaquin, which would pack sustained winds of more than 120 mph over the Bahamas. However, it should be noted that regardless of the hurricane’s strength over the Bahamas, as hurricanes move northward out of the deep tropics, climatology suggests they tend to weaken and speeds up. How much Joaquin could weaken is still unknown. Therefore it is still premature to estimate a landfall location and insured impacts along the East Coast.

Regardless of storm strength, as an East Coast hurricane, Joaquin will create large waves, and the stronger the hurricane, the larger the waves will be. In this case, a constant on shore flow will cause extensive beach erosion along the coastline and impact coast properties.


GFW wavewatch model with 45-50 foot waves off NC coast, 20 plus all the way to Long island by early Sunday AM

Depending on the final track and if Joaquin makes landfall, a large storm surge will likely accompany Joaquin. In fact, the devastating current possibility that Joaquin could track up the Chesapeake or Delaware Bays can’t be ruled out. This type of storm track has been modeled to produce devastating storm surge for these coastal bay waters, and the already high water levels from rainfall and a near super moon will not help the situation.

In addition to the torrential rainfall currently impacting much of the East Coast, some models are forecasting more rain depending on Joaquin’s forecast track. Some forecast models produce an additional 8 – 10” of rain on top of saturated ground. And with already high river levels, some major river flooding can be expected.


A foot of rain—or more—is possible across much of the East Coast this week as Hurricane Joaquin approaches.

In summary, the uncertainty in the current track forecast cannot be understated, and it is not even represented well by the official track forecast by the NHC. Unfortunately in this situation, the spread in the forecast models is far greater in size than the cone of uncertainty in the official forecast by the NHC. As the image below shows there are still several models including the very good and reliable ECMWF (not shown) that take Joaquin out to sea.  I expect by Friday we will have a much clearer picture of where Joaquin will track this weekend, and with that, insured impacts can start to be calculated.


GFS Ensemble model shows 2 distinct solution clusters for storm tracks Door #1 up the east coast. Door #2 out to sea.