Category 2/3 Hurricane Isabel making landfall on Thursday, September 17, 2003 along the Outer Banks of North Carolina. While the intensity of this storm rivaled Hurricane Ivan's Category 5 status at one point, it would never cause nearly as much widespread damage as any ONE of the FOUR hurricanes which impacted Florida last summer. The cleanup from that unprecedented season continues to this day and will so for years, even as some families in the Mid-Atlantic are still regaining parts of their lives shattered by Isabel. The long term effects of so many storms in such a short time may never be fully known or understood. However, the silver lining is that mountains of scientific data were generated as a result of last summer's tropical reign of terror. This data has proven invaluable in helping forecasters improve computer modeling and even begin to make more accurate long range predictions on climate trends which given rise to hurricane seasons of monstrous, or minor, porportions.
Everyone who follows storms, especially hurricanes, has the same questions on their minds as you: "Will there be a repeat of last season?" "Will Florida see such a barrage of storms again anytime soon?" The answer my friend is blowing in the wind, or more rightly put, over the water.
My seasonal forecast for tropical cyclone activity is partly based on water temperature anomaly patterns over the past several years. I realize that "anomaly patterns" is a misnomer of sorts, but I believe as many others in meteorology also do, that changes in the water temperature profile over time reveal a lot of information on how the atmosphere and oceans interact and affect each other in ways we still don't fully understand. As Dr. William Gray of Colorado State University has masterfully demonstrated in his 21+ years of hurricane forecasting, there is an amazing complexity of variables which factor into what type of hurricane season will unfold. I will be tackling in the short term just the water temperature profile issues, and later move on to some of the atmospheric factors such as the North Atlantic Oscillation, the Southern Oscillation Index, influence of El Nino and other data.
The next set of graphics referred to as the SST anomaly for the past three years, recorded in the first week of May by satellite analysis and bouy data worldwide. SST stands for "Sea Surface Temperature" anomaly, meaning the amount of deviation in degrees Celsius from the climatological norm which surface water temperatures were at the time of the analysis. These unclassified maps are available for use by the general public on the Navy's Fleet Numerical Meteorology and Oceanography Center, known as the FNMOC.
While I know you want me to get right to the forecast, I have to take you through the analysis FIRST this time. If you want to skip the analysis, then scroll directly to Part 3 at bottom of this post. More information and detail will be added to that section with time.
To adequately understand what this map is telling you, we must review the evidence presented in the climatological archives from previous hurricane seasons.
NEXT, THE SST ANOMALY MAP FOR EARLY MAY, 2004
You'll notice from this map that as of May 2004, the entire Florida peninsula is surrounded by water that was at least 1 degree C above normal, as was most of the western and central Atlantic. Contrast that with the western Gulf of Mexico...where temperatures were 1-2 degrees below normal. You'll also notice the large swath of much below normal water off the coast of Peru extending into the eastern Equatorial Pacific. This mini "La Nina" (cooling of Pacific waters) has the effect of lessening the intensity of the westerlies blowing across North America. As a result, tropical cyclones under this wind regime can just drift along until they strike land. Weak westerlies and strong easterlies are the reason why Ivan headed north west and then recurved back into the Gulf instead of going out to sea, and Jeanne/Frances both struck along the same portion of coast in a 3 week period. There was no strong force of wind to redirect them out to sea. If your memories of the destruction wrought by Ivan on the central Gulf coast have faded, prepare to grimace and clench your teeth as you view these before and after photos of coastal damage. Below is a big picture composite of all the major storms to impact North America last hurricane season...why it has been called the parade is easy to understand. Credit for this photo goes to Hays Cummin's Tropical Weather Page.
Hurricane Ivan about to wreak havoc on the Alabama-Florida coastline
THIRD, THE SST ANOMALY MAP FOR EARLY MAY, 2003
What should stand out to the careful observer of this map is the extremely large coverage of much above normal water in the western and central Atlantic. This is water that was 2-3-4 degrees C warmer than normal heading into the start of the 2003 hurricane season. Looking at the tracking chart of storms from that summer, it should be clear to understand that strong westerlies and weak easterlies combined with warm ocean surface temperatures made for a busy season. However, much to the benefit of coastal areas, almost all the storms stayed offshore. The westerlies coming across the Gulf of Mexico were strong and persistent enough to serve as a shield against many storms charging up the coast. Except of course for Isabel.
PART 1 SUMMARY: Pre-season Global SST anomalies indicate that changes in sea surface water temperature trends as a result of a previous hurricane season activity may play a major role in determining the frequency and intensity of storms in the following season. Evidence suggests that the U.S. Southeast Coast, including Florida, is much less likely to experience the number of landfalling systems in 2005 as was observed in 2004. Based on observed SST since 2002, the most probable areas of above normal landfall of tropical cyclones in 2005 is focused on the eastern seaboard from North Carolina to Maine, and the western Gulf Coast from New Orleans to Galveston. Tropical activity in the southern and western Caribbean is also expected to be above normal. Development of the first system is projected to be as early as the first 2 weeks of June, and the first area of landfall may be in the central Gulf Coast.
Part 2: Climate indices and hurricane development
As briefly discussed above, there are a wide range of other factors which have a role in the frequency, intensity and duration of tropical cyclone development in the Atlantic Basin. This list are among the major factors to be included in the final hurricane forecast. The part 2 analysis will focus on just the climate indices.
WORLDWIDE SEA-SURFACE TEMPERATURE (SST) FACTORS:
- Disparities in the Gulf of Mexico, below normal in the Western Atlantic, above normal Caribbean and Southeast Atlantic, expansive area of below normal in central and Norther Pacific.
TROPICAL INLAND FACTORS:
- Extent and duration of Monsoon in India and Southeast Asia, seasonal rainfall in central Africa, persistence of offshore winds from central Africa to the Southeast Atlantic.
- The North Atlantic Oscillation, the Arctic Oscillation, the Pacific-North American Index, the Madden-Julian Oscillation, the Southern Oscillation Index
OCEAN-ATMOSPHERIC INTERACTIVE FACTORS:
- The ENSO (El Nino-Southern Oscillation Index) Cycle of multi-year differences in Sea Surface Temperature of Eastern Equatorial Pacific waters in coastal Peru.
Please note that the data analysis for Part 2 is in development and will be posted soon.
Part 3: A timeline of events this hurricane season
I believe the overall hurricane season will unfold like a play with three acts. ACT I will be the inaugural storms to affect the central Gulf and southern Carribbean as those areas feature the warmest SST's at present. Atlantic waters are still too irregular to support anything beyond a fledgling tropical storm until at August.
ACT II will take shape after the mid-point of summer. I believe we will see the westerlies begin to slacken heading into late July, as cooling in equatorial pacific dampens previous warming. This has been alluded to in the Climate Prediction Center's El Nino report as of May 5. Equatorial Pacific cooling will coincide with the seasonal, gradual warming of the western Atlantic and the Gulf. I believe the persistent cooling of water in vicinity of Florida is going to force a westward migration/expansion of the Bermuda High. Cooler-than-normal waters tend to promote high pressure becoming established above. With the westerlies reduced because of a neutral El Nino/Southern Oscillation, and nature abhoring a vacuum, the western Atlantic ridge will take over, reinforcing the easterlies, channeling any Southern Caribbean storms right into the central Gulf much like Ivan's path in mid September 2004.
ACT III will take shape past the climatological mid-point of hurricane season, generally known to tropical meteorologists as September 10. On any given year, records show this point on the calendar, there are one or more tropical cyclones in the Atlantic basin. As waters in the eastern Gulf and western Atlantic warm, the Bermuda High will back off. This is also the time when the NAO can be in a strongly positive phase. With the westerlies picking up again in September, and the Bermuda High eases a bit, a storm is already in development at the time of this transitionary period could get caught between the changing air masses, and get funneled up the eastern seaboard. This scenario happened in 1985 with Gloria and in 1938 New England's "Long Island Express" (as pointed out by loyal reader terpboy). Thus the third area at risk in the latter half of the season will be the Mid-Atlantic and Northeast...from September 10 to mid October.
I am working on a review of my 2004 forecast and a month-by-month summary of what I believe is in store for the Gulf and East Coasts this coming summer. As you have figured out, hurricane forecasting is much more complex and heavy reading that winter storm forecasting. And I have promises to keep, and miles to go before I sleep.