The term Hurricane originated from the name Hurakan³, an ancient Mayan storm god that was considered evil by the Arawakan Indians of the Caribbean islands and South America. Hurricanes are TCs that occur in the North Atlantic Ocean, North East Pacific Ocean and South Pacific Ocean. These storms are warm-core, rapid-onset and low-pressure intense tropical weather systems with sustained winds of 74 mph (64 knots) or higher⁴. When the same storms occur in the North West Pacific Ocean, they are called Typhoons. When they occur in the South West Pacific and Indian Oceans, the word “Cyclone” is used [16] [17] [18] [19] [20] (Figure 1).

Hurricanes are also called severe tropical storms. The phrase tropical indicates the region and environmental condition in the world where they form—warm

oceans and seas between the Tropics of Cancer and Capricorn, between longitudes 100 - 300 North and South of the equator. It should be noted, however, that not all tropical waves, disturbances, or storms develop into hurricanes. Particular environmental conditions are required for the formation of hurricanes, including: 1) Warm Tropical Ocean waters of at least 27˚C/80˚F with a depth of about 46 - 50 m—the heat in these warm waters is necessary to fuel the storm; 2) An atmosphere that cools fast enough with height such that it is potentially unstable for moist convection—the thunderstorm activity allows the heat stored in the ocean waters to be released and used for tropical cyclone development; 3) The mid-troposphere (5 km or 3 miles altitude), must contain enough moisture to sustain the thunderstorms—dry mid-levels are not conducive to the continuing development of thunderstorms; 4) There must be an initial disturbance of air near the surface of the ocean (within the tropics at a minimum distance of at least 500 km from the equator)—which causes the air to begin to spin and has low-level inflow of moist air; 5) The Coriolis force must be present⁵—without the Coriolis force, the low pressure of the disturbance cannot be maintained and 6) There should be a vertical wind shear⁶ of less than 10 m/s⁷ (between 20 - 23 mph) between the surface of the ocean and the upper troposphere [17] [19] - [24].

TC activity varies in the different oceans. Around 12% of all TCs develop in the Atlantic Ocean, 12% develop in the southern Pacific Ocean, 15% develop in the eastern Pacific Ocean, 30% develop in the western Pacific Ocean and 31% in the Indian Ocean both N and S of the equator (including 7% NE of Australia) (Figure 1). Their movement is affected by the prevailing wind system—from East to West under the influence of trade winds [17] [25] [26] [27].

Each ocean has its own particular seasonal patterns of tropical storm activity that constitutes the seven tropical cyclone seasons as shown in Table 2. TC activity tends to peak worldwide in September, while the storms are least active in May.

There are different stages in the formation of TCs. It begins as a tropical disturbance with thunderstorm activity that might stretch about 100 miles and remain active for about 24 hours. In the Atlantic Ocean, these disturbances start as tropical waves from two dominant source regions North and South of latitude 150 over the coast of Africa and move westward often with wind speeds of less than 25 mph [28] [29]. If the right environmental conditions exist (see Section 2.2), a tropical disturbance can develop further and begin to spin around a low-pressure center. When this happens, it is known as a tropical depression if it has wind speeds between 25 - 38 mph. Under suitable conditions, the pressure in the center of the tropical depression can drop, causing air to rush inwards. If the wind speed is strong enough to go beyond 39 mph but less than 73 mph, the system is called a tropical storm. If favorable conditions persist, the pressure in the center of the tropical depression drops further and the winds pick up more speed. When the wind speed in this growing, swirling, vortex of air and clouds exceeds 74 mph, the system is dubbed a TC or hurricane depending on where the storm developed.

A full-grown hurricane has a distinct easily identifiable structure with a spiral shape as shown on Figure 2. At the center of the spiral is a visible “eye” (empty space) where air sinks (downdraft), in an attempt to fill the empty space. The eye is a clear, calm region (light winds of no more than 15 mph with no clouds) that

Related references: Authors, adapted from various sources including the websites of the NOAA’s Hurricane Research Centre (HRC) and National Weather Service (NWS), the UK Met Office.

can extend to diameters of a hundred miles or more. Bordering the eye is the eye wall that consists of bands of clouds, very strong winds and heavy rainfall. Beyond the eye wall are feeder or rain bands, with intense thunderstorms, gusty winds, and often heavy rain showers that spiral inward toward the eye wall. Feeder bands can be very extensive, going outwards for many miles and they often cause tornadoes. The spiral shape or rotating vortex can have an average diameter of about 500 km and can extend to a height of about 15 km. A fully formed TC may only cease after making landfall when the heat source is terminated or when the moisture content is cut off on land [16] [19] [20] [25] [30] [31] [32] [33].

Hurricanes are ranked into five different intensities based on their wind speed; strength and damage potential according to the Saffir-Simpson hurricane wind scale (Table 3). The ranks increase from one to five as the wind speed increases. Category 1 is the weakest, with wind speeds of 74 - 95 mph while category five is the strongest and most dangerous with wind speeds exceeding 157 mph.

Concise metrics have been developed to summarize a hurricane “seasonal activity”⁸ indicating how busy or intense a hurricane season or regional storm activity has been. This is determined by calculating the season’s accumulated cyclone

Related references: Authors, adapted from various sources including the websites of the NOAA’s Hurricane Research Centre (HRC) and National Weather Service (NWS), the UK Met Office.

energy (ACE) index⁹ or Track Integrated Kinetic Energy (TIKE) metric¹⁰ [34] [35]. The ACE index of a season considers the number, frequency, strength or intensity and duration of all the tropical storms and hurricanes in the season. It is computed taking into account the lifetime of the storms and the maximum sustained wind speed [36]. The TIKE, however, is influenced by the size and strength of each storm, rather than solely focusing on maximum intensity without consideration for storm structure, as is the case with ACE [35].

The Climate Prediction Centre of the US government’s National Atmospheric and Oceanic Administration (NOAA) classifies, or ranks, seasons based on the ACE index and the number of storms that have formed within a particular season as above normal, near normal, or below normal. An above-normal season is one that has an ACE index above 111 × 104 kt² and at least two of the following conditions: 1) thirteen or more tropical storms, 2) seven or more hurricanes and 3) three or more major hurricanes. A near-normal season has an ACE index below 66 × 104 kt² and the following three conditions: 1) nine or less tropical storms, 2) four or less hurricanes and 2) one or less major hurricanes. A season is called below normal when it has an ACE range of 66 - 111 × 104 kt² and 1) four or less tropical storms and 2) two or less hurricanes. In particular, long-lived, intense hurricanes have a high ACE index, while short-lived, weak tropical storms have a low value [37]. Table 4 shows the ten extreme seasons from 1960-2017 based on their ACE index.

Before the season starts, meteorologists forecast how active the upcoming season may be. Forecast agencies including governments, universities and private companies issue publicly available AHS predictions. The lead agencies in this field

Related references: Authors, Adapted from [38].

are the NOAA agency, UK government’s National Weather Service (Met Office), the Tropical Meteorology Project (TMP) at Colorado State University (CSU), North Carolina State University, University College London’s (UCL) Tropical Storm Risk (TSR) Consortium and the Commercial forecaster AccuWeather.com (see Table 8). The pre-season outlooks and/or forecasts¹¹ include guesstimates of the number of named storms, hurricanes, and major hurricanes. These are usually issued twice each year—April or May prior to the start of the season in June, and then an update in August [39].

Forecasting hurricanes have several limitations. It is based, for instance, on statistical schemes, which, owing to their intrinsically probabilistic nature, can fail in some years and forecasts do not predict with precision where the storms will strike within the Atlantic basin [40]. Moreover, forecasters caution that, within a single year, there is no solid relationship between the number of storms in a season and the number of landfalls because local weather conditions determine whether hurricanes will make landfall. For example, Hurricane Andrew (1992) occurred in one of the quietest AHS; yet at the time, it was the most expensive hurricane ever to hit the US. These shortcomings highlight the fact that despite huge efforts in prediction/forecasts, there has been little improvement in hurricane intensity forecasts since 1990 [41].

Regardless of the shortcomings in forecasting, an undisputable fact is that seasonal forecasts are good for hurricane risk reduction and management. Indeed, TC mortality and injury have been reduced by improved forecasts and preparedness, especially in developed countries [4]. People may sometimes base key decisions, such as to buy insurance on forecasts of an inactive season. However, the danger is that seasonal forecast can by grossly misinterpreted making it counterproductive. This is because seasons with few hurricanes do not necessarily imply the hurricanes will not be very powerful. For example, as mentioned earlier, the relatively quiet 1992 season produced devastating hurricanes like Hurricane Andrew [41].

The NOAA, UK Met Office and forecast agencies track, monitor and document all the storms that develop every year. In 2017, the Atlantic region was hit by a frenetic stretch of long-lived tropical monster storms, many of them major and dramatic. The 2017 AHS produced 17 storms. Ten were hurricanes (six category 3 or stronger) and the other seven were tropical storms. The season started on April 19 with tropical storm Arlene and extended to November 9, lasting 7 months (Figure 3). The maximum wind speed recorded was 185 mph (295 km/h) by Hurricane Irma.

A prominent caveat of the 2017 AHS is the occurrence of several storms and back-to-back hurricanes (Table 5) in a season classified as above average (see Section 2.7).

The first hurricane of the season was “Franklin” that developed into a hurricane on August 8 and made landfall in Lechuguillas (Mexico). It quickly dissipated into a tropical storm two days later. On August 13, storm “Gert” intensified to a Category 1 hurricane but just 12 hours after achieving peak intensity, the storm weakened below hurricane status. While Gert was dissipating, Hurricane Harvey was brewing in the Caribbean and attained hurricane status on August 24. After leaving a trail of destruction in the Caribbean and South East USA, Harvey spun down into a tropical storm on the August 26. Barely 7 days after Harvey’s onslaught, Hurricane Irma came into line and developed into Category 5 intensity. Irma ravaged several Caribbean islands and barreled directly into Florida before losing its hurricane character. Irma was downgraded into a tropical storm on September 11.

While Irma was still battering the US, storm Jose attained hurricane intensity on September 6 in the Atlantic Ocean and peaked to Category 4 hurricane on September 9. After leaving a lasting impact in the Caribbean islands, Jose stalled offshore into the North East USA on September 22 as a post-tropical cyclone [46]. On September 6, in the Gulf of Mexico, storm Katia also became a hurricane

Related references: Authors, compiled from various sources including the websites of the NOAA’s NHC/NWS, UK Met Office; Caribbean Weather/Weather.org.

but peaked only at Category 2. After making landfall in Tecolutla (Mexico) as a Category 2 hurricane, the system quickly dissipated on September 9. While Irma was still active, storm Maria developed into a hurricane on September 16 and quickly attained Category 5 status on September 18. After devastating the Eastern Caribbean Islands and some areas of South East USA, Maria transitioned into a tropical cyclone on September 30. “Nate” developed into TS from the SW Caribbean Sea on October 5. After making landfall in the US on October 8 as Category 1 hurricane, it quickly dissipated into a tropical depression the next day. Storm Ophelia attained hurricane status on October 11. It dissipated into a tropical storm on October 16 shortly before making landfall in the British Isles—although still alive even at this stage with gusts in excess of 90 mph that devastated parts of Northern Ireland (causing three fatalities).

Table 6 is a summary of the unique characteristics of the 2017 AHS, including new meteorological records set when compared to other seasons.

The 2017 AHS is compelling and stands out as one of the most notorious

Related references: Authors, compiled from various sources including NOAA’s NWS, UK Met Office [19] [47] [48] [49].

when compared to other hurricane seasons. Many new hurricane records were set. Since the development of the satellite for observing hurricane formation in the early 1960s, 2017 is the only year where there have been ten consecutive hurricanes in the Atlantic basin. The 2017 AHS is the most active since the record-breaking season of 2005 and is now ranked amongst the top ten most active seasons since records began. Compared to historical database for land falling hurricanes (Figure 4), the 17 storms that formed in 2017 is above the 1950-2016 average¹³ (11.6) and the 1996-2016 average (14.4). The number of hurricanes (10) is above both the 1950-2016 average (6.2 hurricanes) and 1996-2016 average (7.2). The major hurricanes of the season (6) surpass the average of 1950-2016 (2.5) and 1996-2016 (3.3).

Philippe Klotzbach¹⁴ tweeted that there were “only eight other years, dating to the mid-19^th century that had generated seven or more hurricanes by September 17”. Indeed, it was the first time two Category 4 hurricanes (or stronger) had made landfall in the USA from the Atlantic Ocean since weather records began in 1851. It is rare for a Category 4 (or stronger) Atlantic hurricane to reach the US. Only 29 hurricanes (including Harvey, Irma, Jose and Maria that formed in 2017) have been documented since 1851; of which four were Category 5 hurricanes. Since Hurricane Charley devastated South West Florida in 2004, Harvey’s strike on Texas was the first to occur in the US and landfalls of that strength have not happened since Hurricane Andrew (Category 5) devastated South Florida in 1992. In the Caribbean, the situation is similar. Harvey was the first Category

5 hurricane to hit Dominica and Puerto Rico since at least 1851 and 1928 respectively.

During 2017, hurricanes developed in the Atlantic for more than 52 days. That is the sixth highest number of days since 1851, the most since 2004, and is more than double the seasonal average. Eminent meteorologists Bob Hansen¹⁵ described the season: “This season has been an overachiever by almost every index...we are running at about twice the pace of a typical season”. Eric Blake¹⁶, a scientist at the NOAA similarly remarked: “Never seen anything like this in the modern record”.

In addition, it is notable that during the 2017 AHS six major hurricanes (>Category 3) passed through the Atlantic hurricane basin. It is rare for two hurricanes to pass through the same region in the same year. The last time this happened was in 1829 when the northern Leeward Islands were hit by two major hurricanes in the same season. Maria (Category 4) was the strongest storm to have hit Puerto Rico in 85 years. The 30-year average number of hurricanes for an entire AHS is six. Yet by mid-October 2017, 10 hurricanes had been recorded in the Atlantic.

The first major hurricane of the season, Harvey (Category 4) was the 18th strongest hurricane to make landfall in the US since 1851 and ninth strongest to hit Texas. Harvey has gone down on record as the wettest tropical cyclone in the US history, dropping 34 trillion gallons of water across East Texas and western Louisiana and almost 52 inches of rain in six days [49]. The storm produced 51.8 inches of rain at Cedar Bayou, Texas—the most ever recorded in US mainland from a tropical cyclone, breaking the previous record of 48 inches set in Texas by Tropical Storm Claudette in 1979. Philippe Klotzbach tweeted that, “Harvey broke the longest-running US land falling major hurricane drought at 4,323 days”.

Irma was the 11th most intense hurricane on record in the Atlantic basin, the strongest Atlantic storm in Caribbean Sea on record and the only Category 5 storm to have made direct impact in the northern Leeward Islands [46] [50]. Hurricane Irma’s peak wind speed of 185 mph is the second highest¹⁷ on record, only surpassed by Hurricane Allen (1980) which reached wind speeds of 190 mph. Moreover, Irma is on record as the only tropical storm worldwide to have sustained very intensive wind speeds at 185 mph (298 km/h) for about 37 hours, beating the previous record for a storm of that intensity held by Typhoon Haiyan, which was at maximum intensity for 24 hours. Furthermore, Irma remained a Category 5 Hurricane for the longest time since satellite storm tracking began (at 3.25 days), tied only with the 1932 Cuba Hurricane and the first Category 5 Hurricane to have hit Cuba since 1924 [50].

Moving north-westward at 28 mph (44 km/h), Hurricane Nate was the fastest-moving storm ever recorded in the Gulf of Mexico, according to the National Weather Service. In addition, Nate was the first tropical cyclone to move ashore in the state of Mississippi since Hurricane Katrina (Washington Post, edition of October 8, 2017). Hurricane Ophelia was the easternmost Atlantic major hurricane on record and the worst storm on record to hit N Ireland with gusts of 90 mph that caused three fatalities [45] [51].

Also remarkable during the 2017 AHS is the stretch of hurricanes. This was the longest since the early twentieth century (Figure 5). The timeline for each hurricane to form—one after the other, was very short, averaging only about 6 days between Hurricanes Franklin, Gert and Harvey. Indeed, for three weeks in September 2017, five hurricanes occurred concurrently in the NE Atlantic. These hurricanes were also, notably, long-lived. Included in this series was Harvey (15 days), followed by Irma (14 days) then by longest lived Jose (18 days) and the next longest-lived Maria (15 days) and then Nate (6 days).

The 2017 AHS had an ACE index of 226 that is above the 1950-2016 (101 ACE) average and ranked the fourth largest since 1950 (Figure 6). Only three other seasons (1995, 2004 and 2005) have had a higher ACE since 1960 (Table 4).

The 2017 AHS is the most expensive on record, culminating in more than $202.6 billion in damage [52] [53]. The havoc caused was immense. Table 7 shows the storms that caused considerable damage during the 2017 AHS, and their estimated financial and human costs.

Whilst the full economic impact of the 2017 hurricanes continues to be assessed, preliminary analysis indicates large gross losses from affected short-term investment flows and lost value of capital stock (damages to roads, infrastructure, housing, commercial property, agricultural). Arguably, 2017 is one of the most expensive seasons on record including in the USA especially when compared to Hurricane Katrina and other major hurricanes in 2005 that caused about $143.5 billion in damage.

In the USA, Harvey caused damages of $126 billion, mainly in Texas. Indeed, Harvey and Irma have joined the list of “billion-dollar hurricanes”, the latter being the fifth most expensive hurricane to hit the US [11]. The secondary effects to the US economy could be dire. For instance, it was estimated that Harvey could have reduced Texas’s GDP growth to 2.7% from 4.3% in 2017, while Irma also reduced Florida’s real GDP growth from 3.6% down to 2.5% in 2017. According to the US Energy Information Administration, Hurricane Harvey caused the US refining capacity to drop from 97% on August 25% to 78% on September 8 and this caused retail gasoline prices to increase by 32 cents (13.2%) between August 21 and September 11 [49] [54].

Notwithstanding the relative size of the US economy, the much smaller economies of the Caribbean nations were seriously affected. Prior to 2017, the largest hurricane cost in the Caribbean was caused by Hurricane George (1998), which ravaged six islands causing almost US$11 billion in damages [10]. In 2017, however, the cost was far higher. Irma and Maria caused losses worth $51 billion and $91 billion respectively in Puerto Rico [11]. Irma’s damages amounted to $1.5 billion for Saint Martin, $500 million for the Turks and Caicos Islands, $480 million for St. Bathelemy, $150 - 300 million for Barbuda, $190 million for Anguilla, and $200 million (uninsured losses) for Cuba [50]. Indeed, such economic losses have the potential to set back development in the Caribbean for several decades. The Washington Post (edition of September 15, 2017) reported the Antigua and Barbuda’s ambassador to the USA describing the damage: “For the first time in 300 years there’s not a single living person on the island of Barbuda—a civilization that has existed on that island for over 300 years has now been

Related references: Authors, compiled from various sources including [11], [46] [47] NOAA’s NHC/NWS, UK Met Office.

extinguished”. He said reconstruction could cost up to USD $300 million.

The impact on the population and to the environment was considerable. Irma caused untold sufferings in the Caribbean. Puerto Rico experienced island-wide power outages due to extensive damages on the energy grid. More than one million people were left without power and 6,000 others sort refuge in shelters. In Tortola, St. Thomas, St. John, Dominican Republic and Barbuda, Irma’s severe wind and storm surge caused substantial structural damages to hundreds of buildings and the communication systems. Prime Minister Gaston Browne of Antigua and Barbuda claimed that the Islands were barely habitable afterwards, with 95% if the buildings destroyed. The international NGO, Oxfam, estimated that more than 24,000 people were displaced in the Dominican Republic and more than 5,000 people were evacuated. In Haiti, the storm caused flooding and mudslides that damaged hundreds of homes. In the USA, Irma’s landfall was associated with riverine flooding that reached up to 1.8 meters in several locations in Florida, causing significant damage.

Maria caused substantial structural and roof damage to buildings in St. Croix and Palmas Del Mar, Guayama, Rio Grande, Fajardo, Caguas and Arecibo. Low lying areas at St. Thomas suffered flooding damage. Trees in the region were stripped of their leaves and Guadeloupe’s banana plantations were severely damaged. In Puerto Rico, Maria exacerbated the power outages caused by Irma, causing significant damage to several large wind turbines and damaged weather radars on the Island. The storm also caused 70,000 people to be evacuated in northern Puerto Rico due to the danger of dam failure [47] [48] [49] [55].

The insurance industry also suffered huge losses from the major hurricanes. Estimates indicate insured losses from $25 to $40 billion for Harvey, $35 to $55 billion for Irma, $15 to $30 billion and $15 to $30 billion for Maria. These estimates represent insured loss associated with storm surge, coastal and inland flood and wind damage in the US and the Caribbean. Implications to the banking industry include higher delinquency rates and deterioration in the quality of collaterals in affected areas [47] [56].

The ramifications of the losses are even dire when the knock-on effects to the economies of the Caribbean countries are considered! These include a short-term curtailment of economic activity resulting from a reduction in trade and investment, income, consumption and value of services. These could seriously affect the national economies of the countries impacted in the long-term.

The severity of the storms reignited discussions and arguments that seek to explain why the 2017 AHS season was so active. At the center of these discussions is whether global warming and/or Climate Change (CC) are playing a significant role. Many scientists argue, and we agree, that this seems to be the case, evidenced by the increasing number and intensity of the storms.

Scientists have established that ocean temperatures in the Atlantic have been increasing steadily over the last few decades due to CC, and this is fueling the frequency, intensity and tracks of TCs [3] [15]. For instance, they argue that global warming is contributing to an increase in water vapor in the atmosphere. Indeed since 1988, the amount of total water vapor column over the global oceans has increased by 1.3% per decade [57]. This is causing an increase in the rainfall potential with the prediction that this will lead to an increased ferocity of TCs in the second half of the 21st Century [3]. In fact, climate predictions based on global warming suggest that average-intensity TCs will likely get even more intense in the coming decades as the earth continues to warm¹⁸ [15]. Actually, a 0.5 centigrade increase in sea surface temperature (SST) can lead to a 40% increase in hurricane activity [58].

The 2017 AHS was definitely characterized by highly favorable environmental oceanic and atmospheric conditions that favored cyclogenesis (see Section 2.5) and promoted the above-average hurricane activity (see Section 2.7). During the first half of 2017, the central equatorial Pacific Ocean warmed to borderline El Niño¹⁹ conditions. At the start of the 2017 AHS, tropical Pacific Ocean SSTs and upper ocean heat content cooled rapidly, and El Niño-Southern Oscillation (ENSO)²⁰, which significantly influence inter-annual hurricane activity, quickly transitioned toward La Niña conditions by mid-summer²¹. The lack of El Niño, therefore, leads to a reduction in mid-troposphere vertical wind shear in the Western Pacific Ocean in 2017 [15] [46]. This was highlighted by an atmospheric scientist at Massachusetts Institute of Technology, Kerry Emanuel, who remarked? “El Niño is stuck in neutral this year...there’s currently little difference in wind speeds near the surface and those roughly 10 miles up, which ensures that miles-tall hurricanes can form and remain stable”. Atlantic hurricanes are suppressed when an El Niño is underway in the Pacific because its shearing winds²² can dissipate nascent storms [21] [29].

In August 2017, the NOAA explained that unseasonably warm ocean temperatures and the lack of El Niño would cause the busiest season in seven years. And this was the case in the main Atlantic basin where SSTs increased considerably²³ (Figure 7(a)), reaching their warmest since 2010. The high temperatures caused instability in the atmosphere, which led to the low-level convergence of moisture, winds, and latent heat creating very favorable conditions for systems to develop rapidly and sustain strong intensities for long periods of time. The warmer oceans arguably made Irma and Harvey far more destructive than they would have been in previous decades. Meteorologist Sean Sublette²⁴ stated that: “CC makes these bad storms worse... CC can make it totally disastrous or catastrophic”. If global temperatures continue to rise, it is estimated that the average annual hurricane damages in the Caribbean will increase between

22% and 77% by the year 2100 [10].

The entire Atlantic basin (Gulf of Mexico, Caribbean Sea and Western Atlantic) experienced above-average moisture content (Figure 7(b)) especially in September (the peak months of the season) that produced the major hurricanes. As mentioned in Section 2.2, a higher concentration of water vapor in the atmosphere fuels storm development and intensification [59] [60] especially if there is no dust blowing across the Atlantic from Africa, which tends to have a drying effect on developing storms [28] [61]. In fact, the Climate Prediction Centre asset that in 2017, a strong West African monsoon aligned to make the Caribbean Sea and part of the tropical Atlantic particularly well suited to hurricanes. Furthermore, the sea level pressure across the Atlantic basin in 2017 was generally below average (Figure 7(c)); thereby enhancing cyclone activity. Furthermore, a proper steering pattern was in place—the Atlantic subtropical high extended much further west than in previous decades, creating a predominantly steering flow towards land (Figure 7(d)). This position, therefore, directed the major storms on a westward trajectory towards the Caribbean Sea and South East USA [45] [46]. These hurricane-friendly environmental and atmospheric conditions in 2017 arguably contributed not only to more, but stronger, hurricanes in 2017.

The pre-2017 season forecast did not reflect what actually transpired. Indeed, there was no indication of the hyperactivity of the season that involved the major hurricanes and associated ACE index observed. Rather, most pre-season forecasts issued between April and June indicated a slightly below-average to above-average activity.

The TMP’s pre-season forecasts in April 2017 predicted a slightly below-average activity for the 2017 Atlantic basin hurricanes [48]. In fact, the CSU also made an early forecast that the season would be about 30% below both the 1950-2016 long-term norms and the recent 2006-2015 10-year norm [58]. Prior to the official start of the hurricane on June 1, the main forecast agencies (CSU and TSR) [38], predicted an average to slightly above-average activity (Table 8). In May 2017, the NOAA said the season had a 45% chance of an above average season²⁵. In August, the NOAA upgraded their forecast that: “the season has the potential to be extremely active and could be the most active since 2010”. Accordingly, the TMP and TSR later updated their forecasts to indicate a busier season (above average). The agencies attributed the hyperactivity of the 2017 AHS to environmental anomalies in the central Atlantic. However, the disparities in the earlier forecasts show that it is impossible to precisely predict a season’s hurricane activity in early April (see Section 2.7) [39].

Related references: [46].