Is the Atlantic hurricane dead? Not so fast, say some

Tuesday, September 8, 2015, 6:01 PM - Are the quiet Atlantic hurricane seasons of the past few years part of a trend towards fewer, weaker storms in the years to come?

A new study may be pointing to this possibility, but leading experts in the field are pointing only to flaws in the study.

While researchers long ago coaxed short-term patterns out of the "peaceful" Pacific Ocean - such as El Niño - it's only been within the past twenty years that a significant longer-term pattern has been found lurking in the data from the more turbulent Atlantic Ocean.

This pattern, known as the Atlantic Multidecadal Oscillation (AMO), describes a trend of increasing and decreasing sea surface temperatures in the North Atlantic, which varies on a scale anywhere from 50-70 years long.

The AMO (blue) and its "sister" pattern, the Pacific Multidecadal Oscillation (green) result in a overall Northern Hemisphere Multidecadal Oscillation (black). Credit: Michael E. Mann/RealClimate.org

With quieter Atlantic hurricane seasons over the past few years, and this year being rather mediocre so far, researchers from Colorado State University published a letter in Nature Geoscience this week. Describing how the strength and longevity of hurricanes - specifically the Accumulated Cyclone Energy (ACE) - can be related to these sea surface temperature changes, they sought to show how this could be related to the AMO. Comparing the most recent period of relatively low ACE values (from 1971-1994) to the most recent active period of relatively high ACE values (from 1995-2012), they attempted to show how the recent downturn points towards the latest active period possibly coming to an end.

Michael E. Mann, the Distinguished Professor of Meteorology at Penn State University, who has been delving into AMO research for years, was quick to point out the flaw in the study.

According to Mann, the method the Colorado State researchers used - known as "linear detrending" - simply involves "subtracting off a linear trend [from the data] and calling what's left over the AMO."

Nearly a decade ago, Mann, along with MIT meteorology professor Kerry Emanuel, studied the possible connections between Atlantic hurricanes and the AMO, and found that linear detrending led to the wrong conclusions.

Using statistical methods to separate the estimated human-enhanced climate change signal from the natural variability in North Atlantic temperatures, they found that the human-caused factors, along with cooling due to increases in aerosols in the atmosphere, were the more likely influences over hurricane activity in the Atlantic. Using their tested methodology for extracting the AMO from the data, they found that it only had a modest contribution to the pattern over the past two decades. Of the natural components, it's been the much stronger changes in the PMO (as shown in the graph above) that have had the bigger influences on the overall pattern.

Also, over the past year or so, three different studies either led by or participated in by Mann have delved deeper into separating out the natural AMO signal. Running multiple climate model simulations, their findings further support the view that the linear detrending method simply is not robust enough to properly describe what is going on.

"I must say that is disconcerting, that some scientists continue to use this deeply flawed procedure," Mann wrote in an email to The Weather Network. "Any conclusions that result from an analysis that uses that procedure - including this latest study - are useless. The study can tell us nothing about the relative role of external (including human and natural) and internal 'oscillatory' climate variability."

Where are we in the 2015 Atlantic season?

As of the first week of September, seven storms have developed in the Atlantic Ocean, including two hurricanes - one of which reached category 3 strength. So far, this is in line with the National Hurricane Center forecast update in August, of 6-10 named storms, 1-4 hurricanes and 0-1 major hurricanes.

Tropical Storm Ana touched things off in early May - forming in the pre-season period and becoming the earliest tropical cyclone ever to make landfall in the United States.

Tropical Storm Bill followed in mid-June, then Tropical Storm Claudette in mid-July, and Hurricane Danny (which reached category 3 strength), Tropical Storm Erika, Hurricane Fred (Cat 1) and Tropical Storm Grace have all been packed into the latter half of August and the beginning of September.

Credit: NASA and WikiProject Tropical cyclones/Tracks

With 2015 shaping up as a strong El Niño year, that would suggest it should be a very quiet Atlantic hurricane season. El Niño years tend to suppress Atlantic hurricanes, due to increased the wind shear (the difference in wind speeds between the surface and higher up) over the region. The stronger winds higher up tilt the storms' profiles, limiting their ability to grow in height and thus strength.

However, while that is true, an El Niño doesn't necessarily mean no hurricanes in the Atlantic.

The 1997/98 'super' El Niño still "allowed" eight Atlantic storms to develop in 1997, including a category 3 hurricane, and 14 storms developed in 1998, 10 of which developed into hurricanes (3 of them major hurricanes, including category 5 Mitch).

During the 1986/87 El Niño, which had an unusual late development very similar to the pattern this year's seems to be following, there were also 7 storms, including 3 hurricanes (one of which was Category 3).

The Atlantic hurricane season lasts until October 31, giving nearly two months for more tropical storms and hurricanes to potentially develop.

Sources: Nature Geosciences | Prof. Michael E. Mann | Mann/Emanuel/Eos (pdf) | University of Illinois | NOAA | Wikipedia