Uncovering New England's Prehistoric Hurricane Activity (Part 1 of 3)

We have a passion for documenting and preserving as much weather history as we can here at New England Storm Center. As part of this, we have taken a dive into the world of paleo-tempest-ology to get a sense of potential storm activity before history was documented in this region. Paleo (meaning old or ancient) tempest (meaning a violent storm) ology (meaning the study of...) is a form of research to try to get a picture of prehistoric storms.

Introduction to paleotempestology (the study of ancient violent storms)

In a nutshell, paleotempestology is the study of past tropical cyclone activity using geological proxies and historical records. The primary focus of the research is on sediment displacement along coastlines. This type of research has grown in popularity in recent decades. 

The Atlantic Hurricane Database is only as old as 1851, with sporadic historical records dating back to the 1600s. With paleotempestology, records of possible hurricanes dating back over 1,500 years can be found. These studies have primarily focused on the east coast of the United States over the past couple decades.

Before delving into the findings of studies conducted in Massachusetts and Rhode Island, it is important to note the limitations and issues that arise from this type of research. To begin with, tropical systems are not the only types of storms that can leave behind evidence in sediment. Non-tropical winter storms can also leave behind evidence, although they tend to cause less storm surge than tropical systems. This fact will be discussed more later. These studies focus solely on storm surge, which is not necessarily an indicator of overall storm intensity. Evidence can also vary significantly from one point to another, even within close proximity to a potential landfall.

Overwash deposit analysis is a major focus of research in New England. Overwash deposits are studied in marshes and coastal lakes. When a storm hits these areas, storm surge can overtop barriers (usually sand dunes) and deposit sediment in the marshes behind the barriers. When a powerful storm overtops the barrier, it can cause fanned out sediment deposits. The surge transports sediment from the beaches over the barrier and into the back marshes.

Through the years, these fanned out areas get separated into multiple layers. The back marshes’ sediment layers can be cored out and studied. Each sediment deposit fan layer can represent a powerful storm. The deeper the layer is found, the earlier in time the storm struck. A study at Salt Pond in Massachusetts, for example, drilled a core 800cm (26ft) deep. This revealed potential storm activity records dating back to approximately AD 150.

Below: A sediment core taken from Salt Pond in Massachusetts:

There are also specific limitations to studying overwash deposits. It can be difficult to distinguish storm deposits from other deposits (such as deposits from sediment filtering, wind-driven transport and non-storm related high tides). Storms can erode the barriers that protect marshes and ponds. This can reduce the height of the barrier, making it easier to overtop during future storms. Storms may appear to be stronger than they actually were as a result of the reduced height of the barrier. Fluctuations in sea levels over the course of centuries will also cause varying data and potential errors. 

Overwash deposits can also provide incomplete hurricane data. The more powerful the storm, the more evidence of that storm was left behind. This means that overwash deposits may only show very intense storms and may not show weaker tropical systems such as tropical storms or weak hurricanes. Because of these constraints, all information in this series is potential and nothing can be truly confirmed or rejected.

AD 540: An Extreme Hurricane?

As previously stated, Salt Pond was cored down to 800cm in order to study tropical activity. Falmouth, Massachusetts' Salt Pond is a tidal area. The core was transported to a laboratory to be examined for coarse-grained overwash deposits specifically. The more coarse-grain (versus fine grain) sediment found in the event bed, the stronger the storm was most likely to have been, especially older storms that most likely had to have sediment transported over a larger area versus in the modern era.

Below: Research team collecting a sediment core from Salt Pond in Falmouth, Massachusetts (Photo from Woods Hole Oceanographic Institution:

The bottom of this core dates back to AD 150 (radiocarbon dating is used in these studies), which is by far the furthest a paleotempestology study goes back in New England. Twenty-three event beds were discovered between the years 150 and 1150, indicating potential hurricanes. An additional 10 event beds were found between 1400 and 1675. By far the largest event bed with the most coarse grain was discovered at 693cm, dating from around AD 540. This bed contains a sand fraction of around 36-37% (the larger the percent, the larger the event likely was).

In comparison, there is evidence of the Great Colonial Hurricane, which occurred in 1635, at 253cm. This is a well-documented, historic storm that was most likely one of the most powerful to hit New England. Southern New England experienced widespread damage consistent with a strong category three hurricane. The sustained winds were estimated to be 130 miles per hour, just one mile per hour shy of category four strength. The sand fraction in this storm is 6-7%. If the AD 540 storm was accurate, it may have been truly unprecedented (or even unthinkable) by modern standards. The study wrote:

"The largest coarse anomaly peak occurs at 693 cm (event bed #26) and dates to ca. 540 C.E.. A rip-up clast of fine-grained organic sediment incorporated in the ca. 540 C.E. quartz sand deposit further attests to the layers origin from a high-energy event. While the amount of coarse fraction transported is only one metric for ascertaining the local intensity of an event, these large coarse fraction peaks suggest that the competence of local event driven waves and currents to transport sand-sized particles was greater during recent prehistory than experienced over the last ca. 400 years."

Below: Location of where two cores were taken from Salt Pond:

There were other event beds consistent with historic hurricanes that were well documented and included in the Atlantic Hurricane Database to support the accuracy of this study. The first event bed was discovered only 10cm into the core. The age models place this bed between 1982 and 2005. Hurricane Bob, a category 2 storm, passed within 60 kilometers of Salt Pond in 1991, right in the middle of the 10cm age range. Likely significant hurricanes in 1635 and 1675 also have event beds that correspond to them.

Below: Event bed records found in Salt Pond (Sand fraction is represented by the red line). Hurricane Bob, near the top of the chart is #1. The Great Colonial Hurricane is #3 and the 540 event is #26 toward the bottom:

Since official records began in 1851, this was the only category 2+ storm to pass within 100 kilometers of Salt Pond. Salt Pond was photographed immediately after Bob's passing. Clear washover fans overtopping the barrier in front of Salt Pond caused by Bob's storm surge was seen.

One of the limitations of this type of study, as mentioned in the 'paleotempestology' section above, is that severe winter storms and nor'easters can also cause storm surge, which could theoretically overtop a barrier. Winter storms have not produced storm surge strong enough to breach the barrier in the past, according to Salt Pond's historical data. Hurricanes that passed offshore of Massachusetts have also been unable to generate storm surge powerful enough to breach the barrier and leave event beds in the pond, according to this study. This means Salt Pond likely contains records of only more intense storms whose centers passed within a relatively close range of the pond. The study wrote:

"The last 350 years of sediment accumulation at Salt Pond indicates that only relatively intense hurricanes making a close landfall (∼100 km) to the west of the site have left event beds."

This study concluded that many prehistoric storms may have generated a greater intensity than that of modern hurricanes for the New England region. The study wrote:

"Out of the 32 prehistoric event beds, 12 beds contain more coarse sediment than that deposited by the Great Colonial Hurricane in 1635 C.E. (event bed #3), despite likely being transported a greater distance due to barrier transgression related to sea-level rise...This implies that many of the prehistoric hurricanes may have locally been more intense than those impacting the region historically."

Below: Coarse-sediment anomaly. For an anomaly to be considered an event bed, it must exceed the event bed threshold of 1.34% (which is the black dotted line toward the bottom of this chart. You can see how many events exceed the 1635 hurricane and Hurricane Bob:

The study also showed that the western Atlantic was particularly active between 250 and 1150 AD with another burst of high activity between 1400 and 1675 AD. That first burst of activity (250-1150) saw approximately 23 severe hurricanes strike New England (about once every 39 years on average). These hurricanes may have been equivalent to what we know refer to as category 3 or 4 hurricanes. Since 1635, five hurricanes of category three intensity have made landfall in New England. This averages out to one every 78 years.

You can read the study used for this article in its entirety here: Climate forcing of unprecedented intense-hurricane activity in the last 2000 years

You can also read another recap of this study from the Woods Hole Oceanographic Institution here: Monster hurricanes reached U.S. during prehistoric periods of ocean warming