Using Forensic Meteorology to Verify Hurricane-Related Insurance Claims
2017 was a record-setting year for global insured (and un-insured) losses due to natural disasters, driven in no small part by the costliest hurricane season in U.S. history. Most of the estimated $200 billion in U.S. damages resulted from the wraths of hurricanes Harvey, Irma, and Marie during August and September 2017; and while nature required only a few weeks to wreak such havoc, assessing, tallying, and ultimately rebuilding from the damage will take far longer.
Although assessing hurricane-related claims may seem fairly straightforward to those unfamiliar with the process – “Hey, the building was either hit by a hurricane or it wasn’t, am-I-right?” – meteorologists and experienced claims adjusters know that often isn’t the case, especially as one moves farther from the eye of the storm and therefore farther from the most extreme and obvious impacts.
It is with these more ambiguous claims that forensic meteorology offers valuable insight – reconstructing conditions at the loss location to identify and quantify hurricane-related hazards.
Almost without exception, damage due to tropical cyclones (hurricanes and tropical storms) can be attributed to high winds, extreme rainfall, and/or storm surge. The magnitude of these impacts at any given location depends on numerous factors beyond simply the closest approach of the eye of the storm. The strength, size, and speed of the tropical cyclone; coastal topography; orientation of the coast relative to the storm’s wind field; distance from the coast; and the relative location of landfall all play determining roles.
Failure to understand and account for these factors can lead to under- or over-estimation of impacts, and ultimately to poor coverage decisions.
What follows is a discussion of each major impact (wind, rain, and storm surge) and the data and analytical tools available to reconstruct conditions at a given location. The following discussion specifically addresses tropical cyclone-related hazards, but many of the same analytical methods can be applied to other types of weather events, including severe thunderstorms.
In the 45 years since the Saffir-Simpson scale was introduced, hurricane wind speeds have become nearly synonymous with hurricane intensity and damage potential. A Category 1 storm with sustained surface winds of 74-95 mph is described as producing “some damage,” mainly to roofs, fences, trees, and telephone poles; while a Category 5 storm with sustained winds above 157 mph is described as producing “catastrophic damage,” completely destroying a large percentage of framed structures and leaving the impacted area uninhabitable for weeks to months.
While there is value to the Saffir-Simpson scale – increasing winds do indeed produce increasing damage – the maximum sustained surface wind speed near the eye of a hurricane does not capture the whole story of that storm’s destructive potential.
Two storms with similar maximum wind speeds can produce vastly different amounts of damage – the size of the storm, and therefore the size of its wind field, as well as the forward speed of the storm greatly influence its destructive potential, as Category 3 hurricanes Ivan ($18.8 billion in damages) and Dennis ($2.5 billion in damages) demonstrated when they impacted the same areas along the Gulf Coast just 10 months apart in 2004 and 2005. Despite similar maximum wind speeds, Ivan – a larger storm – produced seven and a half times more damage. In other words, size matters with hurricanes.
Recognizing that maximum wind speed isn’t the best measure of a hurricane’s destructive potential, the tropical meteorology community has been developing new, more comprehensive indices. Such indices include the Cyclone Damage Potential (CDP) index as well as Integrated Kinetic Energy, both of which account for a tropical cyclone’s size as well as its maximum sustained winds.
Although these metrics have not yet seen widespread use outside the meteorology community, expansion to the emergency management community and ultimately to the media and the general public is likely in coming years as these groups seek a more accurate understanding of – and more effective way to communicate – hurricane-related dangers.
Winds within a hurricane – both sustained winds and gusts – are generally strongest in and near the eye wall, with wind speeds decreasing with distance from the center of the storm. The smaller the hurricane, the more quickly winds decrease with distance.
For example, Hurricane Andrew was a very compact storm whose sustained hurricane-force winds extended outward only 30-45 miles, while Hurricane Irma – a much larger storm – had hurricane-force winds extending outward nearly 100 miles. Despite similar maximum wind speeds near the eye of the storm, if a property were 50 miles from the center of Hurricane Irma, it would have experienced significantly higher wind speeds than a property 50 miles from the center of Hurricane Andrew.
Tornadoes can and most often do occur far removed from the center of a hurricane, in the outer bands of discrete thunderstorms that rake counterclockwise away from the eye of the storm. The vast majority of these tornadoes develop in the right-front quadrant of the hurricane, relative to its direction of motion. During Hurricane Irma, 23 tornadoes were identified across the state of Florida. All occurred in areas impacted by Irma’s right-front quadrant.
Tornadoes that form in association with hurricanes are generally short-lived and relatively weak, occurring most often near the coast where wind shear is strongest as the discrete thunderstorms move ashore. However, tornadoes in tropical cyclones can occur further inland: more than two dozen tornadoes were reported in and around the Houston metro area during Hurricane Harvey, some nearly 80 miles from the coast.
If significant wind damage is observed at a property located 50-200 miles from the center of a tropical cyclone, tornadic impact should be investigated as a possible explanation.
Straight-line wind speeds are assessed through both in-situ (weather station) observations as well as post-storm damage surveys. For locations near a weather radar site, radar velocity data of near-surface winds can also be extrapolated to estimate surface winds (this method is less useful for locations far from the radar, where extrapolation becomes less accurate).
Hurricane wind gusts can and often do cause even the most robust weather station anemometers to stop working. During Hurricane Irma, many of the weather stations located at Miami-area airports stopped reporting wind speeds during the peak of the storm.
Therefore, to obtain the most comprehensive and rigorous reconstruction of wind conditions, all three data sources should be used. Weather station data provide on-the-ground observations, while damage surveys assess the aftermath to determine the wind speeds necessary to produce the observed damage. In areas removed from weather stations or population centers, weather radar offers data on winds aloft and insight into winds near the surface.
Tornadoes, which are small-scale phenomena that rarely happen to impact a weather station, are assessed through damage surveys and radar data. The storm cells that produce tornadoes in tropical cyclones are generally shallow, so radar data is most useful for locations within 60-70 miles of the radar site. Beyond that distance, the radar beam may overshoot the top of the storm.
As Hurricane Harvey demonstrated to devastating effect in August of last year, rainfall-induced flooding can be at least as destructive as extreme winds during tropical cyclones. Harvey is estimated to have caused more than $100 billion in damages, primarily due to widespread flooding in and around the Houston metropolitan area – far more damage than was wrought by the storm’s winds.
Although the more-than-four-feet of rain that Harvey squeezed out of the skies over Texas were truly unprecedented, most tropical cyclones produce intense, heavy rainfall that can lead to flooding both during and after the storm, often far inland of the landfall location. Over the last 30 years, such inland flooding has been responsible for more than a quarter of the deaths associated with tropical cyclones in the United States.
In-situ rain gauge data provide measured rainfall totals at specific locations. In urban areas rain gauges may be less than 5 miles apart, while in rural areas, they can be 50 or more miles apart. Radar data can fill in the gaps between rain gauges by providing rainfall estimates anywhere within the radar’s coverage area.
Radars estimate rainfall using an algorithm that associates radar reflectivity values with rainfall rates. The exact relationship between reflectivity and rainfall rate can differ between storm events; so to determine whether the radar rainfall estimates are accurate for a given event, in-situ measurements should be compared to the radar estimate of rainfall at those locations. Confidence in the radar estimate increases when the radar-estimated rainfall total is similar to what was measured at a rain gauge at the same location.
Rain gauges and weather radar can tell us how much it rained but not whether that rainfall actually caused flooding. To determine the locations and extent of flooding, we turn to storm reports and damage surveys. The National Weather Service collects storm reports for all tropical cyclones, and the US Geological Survey conducts damage surveys for most major flooding events. For prolonged, widespread flooding events like Hurricane Harvey, satellite data can also reveal which areas were impacted.
Along the coast, storm surge is often the greatest threat to life and property when a hurricane strikes. As with hurricane winds, the most severe storm surge typically accompanies the right-front quadrant of the storm, where the storm’s wind field and forward motion act to push water onshore. But unlike hurricane winds, which tend to be similar in nearby areas and gradually decrease with distance from the eye, storm surge can vary greatly even within the span of a few miles.
The reason for this variability is that storm surge is a complex phenomenon dependent on characteristics of both the hurricane and the local landscape.
Among the local variables that impact storm surge are the width and slope of the continental shelf as well as the geometry of the coast, bays and estuaries. All else being equal, storm surge along a coast fronted by a wide, shallow continental shelf – as exists along much of the Gulf Coast – will be significantly greater than along a coast where the continental shelf drops away quickly.
Among the storm-specific variables that impact storm surge are hurricane intensity, size, forward speed, and angle of approach to the coastline. The slightest change to any one of these characteristics can significantly increase or decrease storm surge potential. Despite popular belief, it is the winds of the hurricane that generate the vast majority of its storm surge – less than 5% of the surge is due to the low-pressure effect of the storm “sucking up” the ocean toward its center.
In addition to storm surge – formally defined as the abnormal rise of water generated by a storm – one must also consider the local astronomical tide. Storm surge rides atop the normal tidal flow, creating a combined “storm tide” that ultimately determines the depth and extent of inundation. A storm that strikes at high tide can result in inundation several feet deeper than if that same storm strikes at low tide.
A final factor that influences damage along the immediate coast is wave action. Neither the storm surge nor the storm tide take into account the effects of large, wind-drive waves that batter the coast during a hurricane. Wave action can significantly increase the impact of storm tide along the immediate coast, overtopping seawalls and sandbags where the storm tide alone would not. Often, the height of waves riding atop the storm tide can exceed the height of the storm tide itself.
As a recent example, during Hurricane Irma, wave wash marks in the lower Florida Keys were observed 10 – 15 feet above the storm tide. Water has a weight of 1,700 pounds per cubic yard, so prolonged pounding by large waves can cause substantial structural damage.
Ahead of a hurricane, the United States Geological Survey (USGS) typically deploys a network of temporary storm-tide sensors along the immediate coast in the projected path of the storm. These sensors record the depth of the storm tide throughout the event, and comparison of the maximum storm tide at nearby sensors provides insight into the range of water levels experienced along a particular stretch of coastline.
In those areas not covered by the storm tide sensor network, the USGS often performs surveys of visible high-water marks in the immediate aftermath of the storm. High-water marks are created when small, light debris carried along the top of the water is deposited on vertical surfaces like walls and doorways, and as with storm tide sensors, they provide data about the maximum water height at a given location.
The National Weather Service also performs post-storm damage surveys that include findings regarding storm surge, maximum inundation, and wave height, if available.
The Bottom Line
Tropical cyclones tend to be well-documented extreme weather events. Data from weather stations and radar sites generally become available within a few days of the event, while post-tropical cyclone reports and summary storm reports require weeks to months of processing before they are released. Forensic meteorological analysis of site-specific storm impacts can therefore begin almost immediately and can be updated as new data becomes available.
If you are involved in a weather-related investigation and would like to discuss how a forensic meteorologist could support that work, please reach out to Blue Skies Meteorological Services for a free, no-obligation consultation.
In forensic meteorology, we are often asked whether a given weather event was “foreseeable” – in other words, whether those impacted by or involved in the event should have or could have seen it coming.
Foreseeability generally has three components when considering weather events – climatology (what is “normal”), forecast (what is expected) and communication (whether those making decisions have access to crucial information).
In many ways, climatological expectation is about timeline. Over a long enough timeline, even extreme weather events can be expected to occur.
That may sound counterintuitive, so let’s take a familiar example. If you flip a coin 20 times, what are the odds that you’ll get a run of 15 tails in a row? It doesn’t take a statistician to tell you those odds are pretty low.
But what if you flip the coin 200 times? 200 thousand times? As the number of total coin flips grows so too do the odds of seeing a run of 15 tails in a row, for no other reason than that there are more opportunities for it to occur. While a run of 15 tails would be extremely unusual within a trial of only 20 flips, it would normal and expected within a trial of 200 thousand flips. In other words, if you flip a coin 200 thousand times, you should expect to see a run of 15 tails somewhere within that trial. An event that is unusual within a short trial is actually expected within a longer one.
The same is true of extreme weather events. Take the “100-year flood”. What a meteorologist or hydrologist means when they say a “100-year flood” is an event that has a 1-in-100 chance of occurring in any given year. Over a long period of time, we would therefore expect to see such a flood occur, on average, about once every hundred years.
What “100-year flood” doesn’t mean is that a given location is “due” for such an event if the last one occurred more than 100 years ago or that it is “safe” if the last one occurred less than 100 years ago. Every year, the odds are the same: 1-in-100.
The expectation of a given event (from a statistical, climatological perspective) does not change based on past weather. A “100-year flood” is just as likely to occur next year as it was last year.
Go back to the coin flip example: If you have a correctly balanced coin, each flip carries 1-in-2 odds of showing tails. Even if you flip 10 tails in a row, the 11th flip carries the exact same odds of coming up tails: 1-in-2. It may feel like you’re “due” for heads after 10 tails in a row, but that’s psychology speaking not statistics.
What does change after an extreme weather event is knowledge among the affected population. Once an extreme event occurs, people know 1) that it can happen, and 2) what the impacts are. So while the occurrence of one extreme event does not shift the climatological expectation for when a similar event will recur (e.g. suffering through a “100-year flood” this year doesn’t mean you’re safe next year), it can impact the foreseeability of a similar event. After all, people know that it can happen because they’ve already seen it happen.
The fact that a given weather event is within climatological expectations doesn’t necessarily make it foreseeable for the affected population. The weather-related impacts that population could reasonably have anticipated depend on a number of factors, including the weather forecast and impacts statements, as well as how that information was disseminated to the public and/or decision makers.
Weather forecasts, especially for major events, are generally quite good and have improved significantly over the past several decades. Anyone not living under a rock is unlikely to get surprised by a hurricane or a heat wave. Even the warning lead-time for tornadoes is up to an average of 13 minutes – plenty of time to take shelter if you get the warning. (We’ll return to that “if” in a moment.)
That said, sometimes meteorologists get it wrong. Small changes to atmospheric conditions can have major impacts on how a weather event evolves (that proverbial butterfly flapping its wings in Brazil). When there is significant uncertainty in a forecast, meteorologists try to express it. However, scientific uncertainty is a nuanced thing, and often the uncertainty statements are the first thing to get stripped for the headlines. So while meteorologists and decision makers with direct access to them may understand the forecast uncertainty and its implications, the general public may not.
Remember when Hurricane Irma was definitely going to wipe Miami off the map? Neither do I, because at no point during Irma’s evolution was an impact on the southeast FL coast a sure and immanent thing. But “Miami About to be Obliterated!” makes a far more clickable headline than “landfall near Miami looking possible, but storm could still swing further west or east”.
Which brings us to the final crucial piece in foreseeability: information access.
Even if a given weather event is within the climatological norms; even if the forecast is close to perfect, with uncertainty and potential impacts clearly and understandably communicated – even then a weather event can be “unforeseeable”.
If the people impacted by the weather event do not have access to or should not be reasonably expected to seek out forecast information, the event may not have been foreseeable. For example, every few years, hikers are injured or killed in the desert southwest when flash floods turn dusty arroyos into raging rivers.
That last question is arguably more difficult and is where forensic meteorology largely bows out of the discussion.
A forensic meteorologist can tell you whether a given event was climatologically unusual, whether the forecasts leading up to the event were accurate, and how that forecast information was disseminated.
But whether those who were impacted should have ultimately “seen it coming” is a more complex issue.
What if different forecasts disagreed? If a truly extreme event was forecasted, is it reasonable for people to think the forecast was exaggerated? What is reasonable ignorance or incredulity in this age of information-overload and often hyperbole? If someone has access to forecast information, should they be expected to actively seek it out? Should they be expected to act on it?
These are often the types of questions that arise in weather-related investigations and disputes. What starts out as a series of relatively straightforward meteorological questions ultimately winds into a complex web of human psychology and societal and cultural expectations.
The meteorological science is a crucial component of such investigations, but developing a rigorous and comprehensive understanding of the situation requires cross-disciplinary collaboration and communication. The weather is where the questions start, but not necessarily where they end.
If you are involved in a weather-related investigation and would like to discuss how a forensic meteorologist could support that work, please reach out to Blue Skies Meteorological Services for a free, no-obligation consultation.
A bolt from the blue. A rogue wave.
As anyone who has lived on planet Earth for more than a few seasons can attest, nature is full of surprises.
Lightning usually strikes near the core of a thunderstorm, but occasionally it will strike more than 20 miles away, arcing across an otherwise peaceful sky. Ocean swells on a placid sea tend to be of a similar height, but rarely, a rogue wave many times the average height can appear suddenly, damaging or devouring any ship unlucky enough to be in its path.
Bolts from the blue and freak waves are just two examples of variability and chaos in nature. Here chaos describes a sensitive dependence on initial conditions – the proverbial butterfly flapping its wings in Africa that leads to a hurricane over the Bahamas. If the surface temperature had been just slightly cooler, that bolt from the blue would have sliced the air directly under the storm. If winds over the ocean had been just a tad bit different, that rogue wave would never have formed.
It is due to this chaos that atmospheric scientists so often speak in terms of probabilities and distributions (as frustrating as that can be to folks who just want to know whether a rain shower will or will not do the work of watering their lawn this weekend – I promise if we knew, we’d tell you).
The chaotic nature of the weather leads to variability both between events and within a single event. Thunderstorms are a perfect example: no two thunderstorms are alike, and each storm changes from moment to moment.
To describe such variability, we typically speak of averages and deviations from that average. For thunderstorms, we can describe the average peak thunderstorm wind for an area – say 40 mph for summer afternoon thunderstorms in north-central Florida. However, peak winds in exceptionally strong thunderstorms have been recorded at over 70 mph in this region, while winds in weaker storms may not exceed 25 mph. And within a storm that produces a 70 mph gust, winds are not sustained at that speed throughout the event. The sustained wind speed within such a storm averaged over a 2-minute period may only be 45 mph.
Building codes typically ensure that structures and infrastructure are built to withstand expected conditions for a given area – roofs in snowy areas must be able to withstand a higher load than in areas without snow; structures in hurricane-prone regions must be able to withstand higher winds; structures in seismically active areas must be able to withstand earthquakes. It’s generally not average conditions that cause damage – it’s the exceptions, those events that vary greatly from what is normal and expected.
In the case of thunderstorm wind damage, it’s the gusts. How gusty winds are during a given event depends on a number of factors, but mostly on the surrounding terrain: the rougher the terrain, the gustier the winds. Winds are far gustier in the center of a city, surrounded by skyscrapers and densely packed buildings of many sizes than in a smooth, open farm field.
To illustrate, given a weather station measurement of the 1-minute average wind speed, we would expect peak 3-second wind gusts in the heart of a major city to be nearly 250% higher than that average speed, while peak wind gusts over open farmland would be less than 50% higher.
The orientation of landscape features to the wind direction also impacts wind speed and gustiness. For coastal areas, this is particularly noticeable. Onshore winds – those that have blown unimpeded over a lake or ocean – tend to be stronger and steadier, all else being equal, than winds that have blown over a rough landscape (say, across a large metropolitan area). Similarly, winds blowing parallel to city streets tend to be felt more strongly than those blowing perpendicular.
When evaluating weather station data to determine the peak wind speed associated with a given event, we must consider the location of the weather station. Is it a rural site or an urban site? What is the terrain like surrounding the station? Are there certain directions from which the wind would be more impeded or would be blowing over rougher terrain?
How does the weather station site compare to the site at which damage was reported? Is the surrounding terrain similar? Was damage reported at an elevated location – for example, the roof of a high-rise building, where winds tend to be stronger? Differences in both roughness and height must be considered.
Also of obvious importance in the evaluation of winds associated with tropical cyclones and thunderstorms is estimation of the relative strength of the system when it impacted the weather station versus the damage site. If the damage site was directly impacted by the core of a severe thunderstorm while the nearest weather station experienced only a glancing blow, we would expect the winds experienced at the damage site to be stronger than those recorded at the weather station.
Evaluation of these considerations – location, orientation, height, and roughness – allows forensic meteorologists to estimate peak wind speeds from measured average wind speeds and to evaluate the extent to which a given weather station is an accurate proxy for the site at which damage was reported. The weather station data is just the beginning.
“The devil is in the details.” Nowhere is this truer than in the beautifully chaotic weather, where details determine outcomes.
If you or a client experienced wind damage and need an estimate of the wind speed associated with that weather event, call or email Blue Skies Meteorological Services for a free consultation.
Weather radar works by emitting microwave radiation into the sky and then listening for the signal that’s reflected back. It’s a meteorological game of Marco Polo.
All sorts of targets reflect the microwaves – raindrops, snowflakes, hailstones, bats, airplanes, and even swarms of insects. How well a given target reflects microwaves depends on its composition, size, and shape. For instance, liquid water is a better reflector of radar energy than ice.
When a meteorologist looks at a radar display, she’s seeing the reflected signal from all those targets in a given slice of sky. The radar doesn’t “know” which piece of reflected energy came from a bird and which piece came from the hailstone that moments later cracked your car windshield. The radar simply aggregates the reflected signal. It’s up to the meteorologist to interpret the results.
Until just a few years ago, the National Weather Service’s network of weather radars collected information about only two quantities: the reflected energy from a given section of sky (reflectivity) and the velocity of the targets within that section (mean radial velocity and spectrum width). In complex meteorological situations like winter weather events or severe storms, these two pieces of information provide only an incomplete picture of the type of precipitation that’s falling. When you’re just looking at reflectivity and velocity data, for instance, it can be difficult to tell the difference between hail and heavy rain. Yet on the ground, knowing the difference can be critical.
Enter dual-polarization radar technology. If you’ve ever owned polarized sunglasses, you’re already familiar with the principle of polarization. The short-n-sweet version is that electromagnetic waves (like radio waves emitted by radar or visible light waves emitted by the sun) can be oriented along a certain axis.
Tilt your head from side to side while wearing polarized sunglasses, and you’ll notice that the image you see changes – the color of the sky darkens and lightens, glare off the pavement appears and disappears. As you tilt your head, you’re actually changing the polarization of the light that’s being let through your sunglasses, and that gives you additional information about the world around you.
The same is true with weather radar. Conventional radar sends out radio pulses polarized only in the horizontal direction, so the reflected signal carries only 1-dimensional information. Dual-polarization (or “dual-pol”) radar, on the other hand, sends out both horizontally polarized pulses and vertically polarized pulses, so the reflected signal carries 2-dimensional data.
This may seem rather trivial until you consider that precipitation types have characteristic shapes. Small raindrops are spherical, while big raindrops flatten out like a Frisbee. Hailstones are roughly spherical when they’re dry but can become oblong as their outer layers melt. The two-dimensional data provides invaluable insight into what types of precipitation are present within a storm.
Here in Florida, we don’t have to worry too much about winter weather, but hail is another matter. In the lightning capital of the United States, thunderstorms are part of the scenery for much of the year, and most thunderstorms, if they are strong enough and reach high enough into the atmosphere, produce hail.
But that hail doesn’t always reach the ground. In warm, moist atmospheres, hail melts as it falls toward the ground. If the hail starts out small or if the freezing level is high in the atmosphere, hail can melt completely before reaching the ground. Dual-pol radar data can reveal whether a storm is producing hail aloft, and by examining radar data at different heights within the storm, meteorologists can determine whether and how much that hail is melting before it reaches the surface (and people’s cars and houses).
Dual-pol radar adds three more tools to the meteorologist’s kit. Each of these tools provides unique information about the size, shape, and mixture of precipitation types within a storm.
Correlation Coefficient (CC)
Correlation coefficient measures how similarly the returned horizontal and vertical pulses are behaving. It’s like looking at the world under a strobe light. From one flash to the next, how much does the image change? When the targets within a given region are of the same shape and type (for example, all medium-sized raindrops), one pulse will look much like the next, and the correlation coefficient will be high. If, on the other hand, precipitation types are mixed (like rain and hail swirling together), correlation coefficient values will be lower. Generally, the larger the hail, the lower the correlation coefficient.
Differential Reflectivity (ZDR)
Differential reflectivity compares the reflectivity values returned in the horizontal and vertical directions, like comparing how much the image through your polarized sunglasses changes as you tilt your head. Targets that are wider than they are tall (like large raindrops) have higher differential reflectivity – they reflect more horizontally polarized energy than vertically polarized energy. Hailstones, on the other hand, are more spherical and tend to tumble as they fall, reflecting roughly equal amounts of horizontally and vertically polarized energy. Hail typically has low to near-zero ZDR values.
Specific Differential Phase (KDP)
Specific differential phase is a bit more complicated than correlation coefficient and differential reflectivity. Physically, KDP measures the phase shift of the returned horizontal and vertical signals. In practice, this means that specific differential phase responds to both the shape and the density of liquid water targets. Frozen precipitation, like dry hail and snow, do not contribute to KDP – KDP “ignores” frozen precipitation and sees only liquid precipitation. Specific differential phase is therefore useful for determining rainfall rate.
As part of the dual-polarization upgrade, National Weather Service weather radars now incorporate an algorithm that estimates precipitation type from the dual-pol variables discussed above. Numerous automated hail report websites use the National Weather Service algorithm or a custom one to identify regions of hail. While such algorithms provide a useful first-pass to identify regions within a storm where hail is likely being produced aloft, they do not provide information about whether that hail is reaching the ground and at what size.
When Blue Skies Meteorological Services investigates the presence of hail for a forensic meteorology case, we don’t just run an algorithm and depend on the radar to “know” what was happening in the storm and to assume what was happening on the ground. We examine official storm reports, severe weather warnings and advisories, the atmospheric profile, and dual-polarization radar data at multiple heights and throughout the lifetime of the storm to reconstruct a comprehensive picture of the weather situation – both high in the storm and on the ground, where it matters.
Typically, forensic meteorology is applied to weather events a few years old at most – Did damaging hail really strike that commercial facility in April of last year? Was it a tornado or a microburst that ripped off roofs and uprooted trees last week? Did a lightning strike start that house fire a few months back?
Occasionally, though, forensic meteorologists look decades or even centuries into the past. Such has been the case with Tropical Cyclone (TC) Mahina, which struck Bathurst Bay, Australia, on 5 March 1899 as a Category 5 storm. Since the publication of a research paper by H.E. Whittingham in 1958 titled “The Bathurst Bay Hurricane and associated storm surge” reported that TC Mahina produced a storm surge of 13 meters (or over 42 feet), that storm has generally been credited with the largest ever-recorded storm surge.
Contemporary accounts of the storm reported a waist-deep wall of water inundating a 40 ft tall ridge where several law enforcement officers were camped during the storm, dolphins being found stranded atop 15 m high cliffs after the storm had passed, and fragments of Aboriginal canoes being deposited 70 to 80 feet above the normal high tide.
TC Mahina was a undoubtedly a monster storm. With sustained winds of over 175 mph, it sank 54 ships (mostly pearling vessels) and killed more than 300 people, sweeping devastation across the Bathurst Bay region of northeastern Australia.
Yet despite its impressive statistics and a number of contemporary (albeit generally third-person) accounts of storm-related inundation, meteorologists have long regarded the 13 m storm surge record skeptically. It just didn’t seem possible. The commonly reported central pressure of 27 inches of mercury (914 mb), while extremely low, just doesn’t support a storm surge as high as a four-story building.
Previous studies that used computer models to estimate storm surge given the most likely track and intensity of the cyclone over this topographically complex region had come up well short of 13 m, and field work in the region had not found evidence of debris deposits to the reported 13 m height.
Something was amiss – either the central pressure was lower than 27 inHg or the storm surge wasn’t actually 13 m high. Possibly both.
Despite the contradictory evidence, little research had been done to set the record straight, until recently. In the May 2014 issue of the Bulletin of the American Meteorological Society (BAMS), several Australian scientists revealed the results of their forensic analysis of TC Mahina’s storm surge.
In that analysis, they utilized methods that forensic meteorologists often use to evaluate much more recent weather events: they investigated historical records, examined the physical evidence, and modeled the event. What they learned is that, as is often the case, the devil is in the details.
Previous modeling studies had relied upon a thirdhand account of Mahina’s central pressure published in an anonymously authored report several months after the cyclone made landfall. That central pressure – 27 inches of mercury (or 917 mb) – was simply too high to produce a 42 foot storm surge.
By combing through the historical record, though, the authors of the May 2014 study found several references to the storm’s central pressure as an astonishing 26 inHg (880 mb). All of those accounts ultimately were traceable to the same man – a ship captain whose schooner was the only vessel to experience – and survive – the eye of the cyclone. That captain, William Field Porter, also happened to write a letter to his parents recounting his harrowing experience. In it, he stated plainly that “the barometer was down to 26 [inches of mercury].”
So it would seem that the central pressure that had been used in previous modeling studies – studies that had failed to reproduce anything nearing a 13 m storm surge – had been too high. Perhaps the lower pressure of 26 inHg would create the reported record storm surge?
Before jumping straight into the modeling, though, the authors also re-examined the physical evidence – debris that was washed up and deposited by the storm. They found wave-deposited sandy sediments 6.6 meters above mean sea level at Ninian Bay, the location where law enforcement officers reported the 13 m storm surge, but they found no evidence of inundation above that.
This doesn’t necessarily rule out the 13 m water level, however. During tropical cyclones, the highest debris tends to consist of biological material that floats, like leaves, sea grasses, and small marine animals. This material also tends to biodegrade after a few years or decades, leaving no trace for forensic meteorologists peering 115 years into the past. Observations after more recent tropical cyclones suggest that sandy sediments can be deposited at only half the height of maximum inundation, so it is quite possible that the water did reach a height of 13 m above mean sea level at Ninian Bay, where those sandy deposits were found at 6.6 m.
Once the authors had concluded that there was a decent probability that the storm’s central pressure really was 26 inHg and that the waves at Ninian Bay really did reach 13 m above mean sea level, they set about to model the storm surge based on a range of storm forward speeds and storm tracks suggested by ships’ wind and pressure recordings as well as damage assessments after the storm passed.
What they discovered was that even in a worst-case scenario (Mahina approached Bathurst Bay from the northeast with a central pressure of 26 inHg), the storm surge would “only” be about 9 m.
And here is where the devil is in the details. You see, there’s a difference between storm surge and maximum inundation. Storm surge is the abnormal rise of water generated by a tropical cyclone, over and above the natural (astronomical) tides. Click here for an animation of storm surge in an area of steep topography, similar to Bathurst Bay. Storm surge is influenced by the size of the storm, winds speed within the storm, the forward speed of the storm as it approaches land, the angle of approach to the coast, the topography of the sea floor and coast, and the storm’s central pressure.
Maximum inundation, on the other hand, is just what it sounds like – the maximum height that water reaches above mean sea level. Maximum inundation is influenced by the height of the storm surge, the timing of astronomical tides, and various types of wave action like wave setup and wave run-up. It’s the high water mark.
And the high water mark is almost always higher than the storm surge. In fact, in severe tropical cyclones in northeastern Australia, wave and tidal effects have added approximately 25% to the height of maximum inundation.
What this means for Tropical Cyclone Mahina is that the 1899 accounts of a monster cyclone that brought the sea to the top of a 40 ft high cliff may in fact have been accurate. If the central pressure was actually 26 inHg and the storm approached from the northeast, it could have generated a storm surge of up to 9 meters (30 feet). Mahina struck during astronomical high tide, and that combined with wave setup and run-up could have added an additional 4 m (12 feet) of water on top of the storm surge.
So, the record for highest storm surge may have to be revised downward (Mahina’s storm surge was probably 9 meters or less), but its maximum inundation may still take the gold. It is entirely possible that on March 5th, 1899, men waded through seawater atop a 40 ft high cliff and dolphins swam through the tops of 50 ft tall trees.
Blue Skies Meteorological Services offers forensic meteorological analyses of a wide range of weather events — from hail storms to lightning strikes, from flooding to tornadoes, from fog to sun glare. We typically don’t look 115 years into the past, but we’re always up for unique and interesting challenges! Give us a call or send us an email to discuss your weather-impacted legal case, insurance claim, or investigation.