As of October 2, 2016, Hurricane Matthew is poised to impact the vulnerable and disaster-prone nation of Haiti, targeting the island nation with the most dangerous north-east quadrant of its 140 mile-per-hour Category 4 winds. October hurricanes, when they have occurred in Haiti over the past decade, have caused tremendous disaster impacts. In October of 2012, Haiti’s agricultural sector was decimated by Hurricane Sandy. In October 2007, Hurricane Noel was the second wettest tropical cyclone on record in Haiti causing massive flooding and mudslides.
2012, 2007, and 2016. Sound familiar? These years also represent the very worst years of peak Arctic sea ice melt extent in terms of area in square kilometers at peak melt point. Is it possible that there could be a potential teleconnection between the worst-of-the worst of Arctic sea ice melt years and Haiti’s October Hurricanes?
To explore this hypothesis, ReliefAnalysis analyzed a period of record of 12 years of peak Arctic melt seasons with data from the National Snow and Ice Data Center to confirm Arctic sea ice minimum melt extent since 2005. The results were then sorted into clusters of Severity Categories to include “Worst” representing the 2012/2007/2016 minimums, “Strong,” “Medium,” and “Less Severe” to determine a spectrum of values within this spectrum of accelerating sea ice loss.
From there, 21 named storms that have affected since Haiti were analyzed in chronological order, with primary data from the Center for Research on the Epidemiology of Disasters and Weather Underground. Key data elements were Haiti impact date, Cyclone name, maximum Category on the Saffir-Simpson Scale, and “Number Affected” – a proxy for impact given that damage and fatality figures for Haiti have some historical data challenges. Then, these 21 named storms were cross referenced against the key elements of Arctic sea ice patterns…did the storm occur after peak Arctic Melt Season for that year–which also occurs about half way through the most active peak of the Atlantic Hurricane Season. In addition, did that particular storm occur in a “Worst,” “Strong,” “Medium,” or “Least Severe” sea ice loss category. Any storms that occurred after peak melt season are highlighted in yellow.
The period of record chosen was intentional. Haiti has a very long, and devastating history of hurricane impacts. The 12 year period of record is short enough that it can represent some granularity and separation with our recent experience with Arctic amplification and allow for some nuanced sea ice extent categories. (Sea ice extent was chosen as the variable;, had thickness data been available, 2016 may well have been the “Worst” year.) Expanding the record too far would simply have clustered the last decade of the sea ice loss spiral into a less useful composite category. At the same time, the record was expanded to include the hyperactive 2005 Atlantic Hurricane Season, when the storm naming convention used Greek letters well into December, and before sea ice extent began to plunge below 5 million square kilometers consistently.
The numbers show us that looking at “Number Affected” being greater than 100,000 people as a proxy for major hurricane impact, only Noel and Sandy had tremendous impacts, and occurred post-peak melt season (in fact in October). Noel (2007) and Sandy (2012), fell in the “Worst” category of recent Arctic sea ice loss years.
2010’s Tropical Storm Matthew and Hurricane Tomas occurred during a very dangerous Atlantic Hurricane Season with hundreds of thousands displaced in the wake of the Port-au-Prince Earthquake. Their significant but relatively less severe impacts occurred post-peak melt season and were in the “Strong” category of sea ice loss. Hurricane Ike, in 2008, had a very significant impact on Haiti, but occurred pre-melt season peak and in the heart of the Atlantic Hurricane Season’s active window.
Looking deeper into Haiti’s deeper hurricane history, post melt-season peak devastating Haiti hurricanes do occur, and significantly so. Here are six examples between 1935-2004–a 69 year period of record averaging a return frequency of 11.5 years:
- Hurricane Jeanne impacted Haiti on September 17, 2004 and claimed 3,000 lives.
- Hurricane Flora made a Category 4 impact on Haiti October 3, 1963 and caused 8,000 fatalities.
- An unnamed storm in 1935 claimed over 2,000 lives during a mid-October impact.
- Hurricane Hazel impacted Haiti as a Category 3 storm in mid-October 1954, claiming 1,000 lives.
- Hurricane Gordon’s November 1994 impact claimed 1,000 lives.
- In 1998, Hurricane Georges caused a major agricultural disaster and also claimed 400 lives, while making an impact on September 21.
Haiti has a long history of devastating hurricane impacts. From 1935-2004, six such disasters occurred after the peak melt Arctic sea ice loss, before our current era of increasing Arctic amplification. These events occurred on average once every 11.5 years.
Since 2005 (a much shorter period of record of 12 years), we have been 3 for 3 in terms correlating years of very significant sea ice loss, to a strong October Haiti hurricane impacts–with Matthew’s current approach included. The return frequency for this smaller period of record is once every 4 years.
Of course, this is not a meteorological analysis by any means–merely an overlay of patterns and humanitarian impacts. But certainly a humanitarian trend worth continuing to evaluate in future years as Arctic amplification continues.
Programming Note: I have had the honor of finishing up the audio of a discussion with Brad Baldilla, a prominent activist in the Asian American and Filipino American community who is very involved in Business Process Outsourcing–one of the increasingly prominent trends in globalization and international labor markets. The implications are significant, and I look forward to sharing to kick off the fall series of ReliefAnalysis podcasts shortly.