At the Midpoint of the 2008 Atlantic Hurricane Season Editor s note: It has been an anxious couple of weeks for those with financial interests in either on- or offshore assets in the Gulf of Mexico and a busy time at AIR for the tropical cyclone team that develops loss estimates for approaching storms in real time. In this article, Dr. Peter Dailey reviews the 2008 hurricane season at its midpoint and looks ahead to what might be in store for the remaining weeks and months of the season. 09.2008By: Dr. Peter S. Dailey, Director of Atmospheric Science Introduction So far, the 2008 Atlantic Hurricane season, which officially kicked off on June 1st, has produced two very notable storms: Hurricanes Gustav and Ike, both achieved Category 4 status during their life cycles and both wreaked havoc in the Caribbean. Ike, in particular, may turn out to be comparable in some respects to storms of the very active 2004 and 2005 seasons. This article examines the 2008 hurricane season at its midway point, by which time the number of tropical storms and hurricanes had clearly exceeded long-term averages. What does one expect in the typical season? What has influenced this year s activity, and how does it compare half way through to the last several seasons? How have the seasonal forecasts been updated, and what do they indicate for the remainder of the season? Atlantic Tropical Cyclone Climatology at the Season s Midpoint Figure 1 shows how one would expect the typical Atlantic hurricane season to evolve. Tropical storms tropical cyclones that achieve a wind intensity of 39mph typically begin to form in the months of May and June, with the first expected by the end of June. Activity accelerates in the most active months of August and September, and then tails off by the season s official end in November. By season s end, one expects about eleven tropical storms to have formed. By the end of a typical season, about 11 named tropical storms will have formed. As Figure 1 also illustrates, hurricane formation follows a similar pattern over time, with activity peaking in August and September. About 58% of tropical storms achieve hurricane status, with winds of at least 74mph. About 45% of hurricanes reach major hurricane status Category 3 or higher on the Saffir-Simpson scale, with winds exceeding 110mph. About six hurricanes and three major hurricanes will have developed by the end of a typical year.
Figure 1 Seasonal Evolution of Atlantic Tropical Cyclone Activity The traditional midpoint of the season, as defined by historical activity levels in the Atlantic, occurs around September 8th. In 2008, as of the September 8th midpoint, there had been 10 named tropical cyclones, among them 5 hurricanes, and 3 major hurricanes. Hurricane Bertha achieved major hurricane status early on in the season, but the storm remained over the open Atlantic throughout its life cycle and did not impact land interests. Gustav achieved Category 4 status just before crossing the west coast of Cuba, but maintained that strength for just 12 hours, ultimately arriving on the coast of Louisiana as a strong Category 2 hurricane. Hurricane Ike achieved Category 4 status as well, causing significant damage in the already hurricane worn Caribbean, but was unable to re-intensify to major hurricane status after being subjected to the mountainous terrain of Cuba. Hurricane Ike is now known to have made landfall near Galveston, TX in the early morning hours of September 13th, but since this occurred after the season s midpoint, its landfall will not be considered in this discussion. Again, in summary, as of the midpoint of the 2008 season, there have been 10 tropical storms, 5 of which reached hurricane status. Among the hurricanes, there has been one each of CAT1, CAT2, CAT3, and two CAT 4s. Two Dolly and Gustav made U.S. landfall at hurricane strength. (We now know that Ike will be included in the end of season totals.) By comparison, in an average season at the midpoint, one would expect 5 to 6 tropical storms, about 3 hurricanes, and between 1 and 2 major hurricanes. One average, one would expect the U.S. to have experienced one hurricane landfall at the season s midpoint. Thus, the counts so far in 2008 are above average. Comparing the Midpoint of 2008 to Previous Seasons How does 2008 compare to the last several seasons at midpoint? Figure 3 shows the evolution of named tropical cyclone counts relative to the long-term average (Figure 1) for the 2004, 2005, 2006, and 2007 seasons. First, it is clear that each season is unique. The 2004 season started off slowly, but by the middle of August the tropical storm count reached average levels, and then activity took off such that by the midpoint there were 8 named storms. The 2005 season started off early and fast, and maintained rapid storm development reaching 13 named systems by September 8th. The 2006 and 2007 seasons were both fairly active, though by mid-season, both had seen six named storms, which is near average. Of the other two hurricanes, Dolly made landfall as a marginal Category 2 just north of the border with Mexico, causing low-level damage centered near Brownsville, TX. Hanna achieved Category 1 status for less than a day and made landfall near the border between South and North Carolina as a strong tropical storm. Figure 2 Summary of Tropical Storms and Hurricanes So Far in 2008 2
How does 2008 compare to the last several seasons at midpoint? Figure 3 shows the evolution of named tropical cyclone counts relative to the long-term average (Figure 1) for the 2004, 2005, 2006, and 2007 seasons. First, it is clear that each season is unique. The 2004 season started off slowly, but by the middle of August the tropical storm count reached average levels, and then activity took off such that by the midpoint there were 8 named storms. The 2005 season started off early and fast, and maintained rapid storm development reaching 13 named systems by September 8th. The 2006 and 2007 seasons were both fairly active, though by mid-season, both had seen six named storms, which is near average. and 6 hurricanes respectively. On the other hand, in both 2006 and 2007, less than half of the six tropical storms had reached hurricane strength. None of the three hurricanes that developed by mid-season in the previous two years reached the U.S. coastline at hurricane strength. Overall, it is clear that the 2004 and 2005 seasons were very active from the perspective of both tropical storm and hurricane counts. Further, in both of these seasons, there had been two U.S. hurricane landfalls, most of which occurred at major hurricane strength. It was clear, from any perspective, that even by the midpoint of these seasons, activity was well in excess of average. In 2006 and 2007, activity levels were fairly normal as of mid-season. As it turned out, the indications at the midpoint of all four of these Atlantic seasons were good predictors of the activity for the remainder of the season; that is, 2004 and 2005 were very active, and 2006 and 2007 were relatively average. Note also that by the midpoint of the 2004 and 2005 season, the U.S. had seen a number of hurricane landfalls, several of them at major hurricane intensity. Figure 3 Evolution of Mid-Season Tropical Storm Counts from 2004 to 2007 Figure 4 shows the evolution of hurricane counts relative to the long-term average for the same four seasons. Recall that, on average, just over half of named storms reach hurricane status, and that by mid-season one would expect about 3 hurricanes to have developed from about 6 tropical storms. In 2004 and 2005, by mid-season, it was clear that hurricane activity was well in excess of the mean, with 5 Figure 4 Evolution of Mid-Season Hurricane Counts from 2004 to 2007. The percentage in the upper left indicates the proportion of tropical storms that went on to reach hurricane strength at the season s midpoint. Storms that made U.S. hurricane landfall are highlighted in red. 3
Indications for the Remainder of the 2008 Atlantic Hurricane Season So, how does the remainder of the 2008 season look? To put the midpoint of this season in perspective, Figure 5 shows the progression of tropical storm and hurricane count as of September 8th, 2008. Note that the season started early with Tropical Storm Arthur, which was named on May 31 just before the official start of the season. Arthur was a short-lived storm only marginally meeting the requirements for a named system; it formed from the remnants of a tropical cyclone from Mexico s Pacific coast and a pre-existing tropical wave. It is questionable whether a storm like Arthur would have been named at all ten years ago. Considering this, the 2008 season is still above average in terms of overall activity, and more typical in terms of how storms are intensifying with roughly half of the named systems reaching hurricane strength and again about half of them reaching major hurricane status. Of course, meteorologists and seasonal forecasters pay less attention to the midpoint statistics and much closer attention to the forecasted ingredients for hurricanes in the coming months. The main source of energy for hurricanes is the heat and moisture of the ocean surface, so one of the key factors that forecasters scrutinize is the projected sea surface conditions for the remaining months of the hurricane season. Figure 6 shows the Atlantic sea surface temperature (SST) anomalies forecasted by the IRI group at Columbia University for the months of October through December 2008. This is one of several operational ocean forecast models in operation and provides a good indication of how the ocean conditions are likely to develop through the remainder of the season. The graphic shows the projected anomalies (deviations from average), with positive or warm anomalies indicating regions of the ocean that are likely to be more conducive than normal to hurricane development. Figure 5 Evolution of Mid-Season Tropical Storm and Hurricane Counts for 2008. One-half of tropical storms reached hurricane strength (5 of 10) and about one third reached major hurricane strength (3 of 10) as of September 8th. Figure 6 SST Forecast for the Months of October, November and December 2008 Of course, the experience through mid-season is not always indicative of what is to come for the remainder of the season. But, at least for the last several seasons, the counts of tropical storms and hurricanes through the season s midpoint have been a reasonable predictor for the season as a whole. The forecast indicates that much of the tropical Atlantic will be within a quarter degree Celsius of average (though a large expanse of ocean well outside the tropical latitudes in which hurricanes form may be significantly warmer than average). Because many tropical cyclones form over the middle and far eastern part of the Atlantic in the second half of the season, there is the potential for such storms to be fueled by warmer than average SSTs, as indicated by the forecast. 4
Another indicator of activity is the level of Atlantic wind shear, which describes the difference between surface and upper-level winds. Wind shear, which prevents tropical cyclones from reaching their full potential, is correlated with the SST anomaly pattern over the eastern Pacific known as ENSO. In the La Niña phase of ENSO, wind shear in the Atlantic is reduced, thereby leading to elevated levels of activity. For the September to November timeframe, the IRI team has recently increased the estimated probability of an El Niño event, if a weak one. 1 It is interesting to note that the two most recent disturbances being tracked in the Atlantic by the National Hurricane Center are not likely to develop because of particularly strong shear conditions. Of course, this may be purely coincidental and, in fact, the most likely scenario is for ENSO neutral conditions, which suggests that ENSO is not likely to play a critical role in activity levels for the remainder of the 2008 season. The question as to how many of the hurricanes that form later this year will make landfall along the North American coastline relates mainly to the steering currents in place at the time the storms are tracking. It is well known that storm movement is largely a reflection of the large-scale atmospheric winds that change on a weekly or even daily basis. Thus, steering is a function more of weather than of climate, and short-term changes in the atmospheric circulation can make a significant difference in the propensity of storms to reach the coastline. Figure 7 Evolution of the 2008 Atlantic Seasonal Forecasts Summary The atmospheric conditions that lead to storm development, such as sea surface temperatures and environmental wind shear, are close to average, with some of the tropical Atlantic expected to be warmer than average. This combined with neutral ENSO conditions will likely lead to a continuation of what we ve seen so far, namely, higher-than-average levels of activity in the basin. As to how many more hurricanes make landfall along the U.S. coastline, highly variable weather conditions are the primary determinant and are difficult to predict well in advance. Updates to 2008 Atlantic Hurricane Seasonal Forecasts What are the seasonal forecasters saying about the remainder of the 2008 season? Several forecast organizations have updated their early season projections. The evolution of these forecasts is shown in Figure 7. Virtually all have revised their predictions upward. Part of the explanation for these adjustments is the warmer-thanaverage SST conditions being forecast by groups like IRI Columbia combined with the expectation that the ENSO cycle will not counteract the effect of a warm ocean with higher-than-average shear conditions. 5
1 See http://iri.columbia.edu/climate/enso/currentinfo/figure3.html About AIR Worldwide Corporation AIR Worldwide Corporation (AIR) is the scientific leader and most respected provider of risk modeling software and consulting services. AIR founded the catastrophe modeling industry in 1987 and today models the risk from natural catastrophes and terrorism in more than 50 countries. More than 400 insurance, reinsurance, financial, corporate and government clients rely on AIR software and services for catastrophe risk management, insurance-linked securities, site-specific seismic engineering analysis, and property replacement cost valuation. AIR is a member of the ISO family of companies and is headquartered in Boston with additional offices in North America, Europe and Asia. For more information, please visit www. air-worldwide.com. 2008 AIR Worldwide Corporation. All rights reserved. 4