WunderBlog Archive » Category 6™
Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.A 65% chance of an above-average 2008 hurricane season : NOAA
There is a 65% chance that year's Atlantic hurricane season will have above average activity, and only a 10% chance it will be below average, according to today's seasonal hurricane forecast issued by the National Oceanic and Atmospheric Administration (NOAA). In an effort to emphasize the uncertainties in their forecast, NOAA for the first time included a probability with their "numbers" forecast. The numbers NOAA predicts are 12-16 named storms, 6-9 hurricanes, 2-5 intense hurricanes, and an Accumulated Cyclone Energy (ACE) index 100%-210% of the median. They put the odds of experiencing those numbers at 60-70%. This probability is based on the numbers seen in past hurricane seasons that had May climate patterns similar to those of 2008. An average hurricane season has 10 named storms, 6 hurricanes, and 2 intense hurricanes. The forecasters also call for a 70% chance of a below normal season in the eastern Pacific off the coast of Mexico, and a below normal season in the Central Pacific near Hawaii.
The forecasters considered two main climate signals to make their forecast:
1) Continuation of conditions that have been conducive to above-normal Atlantic hurricane activity since 1995, including above-average sea-surface temperatures in the eastern tropical Atlantic Ocean (Figure 1). Between 1995-2007, the Atlantic has averaged 14.5 named storms, 8 hurricanes, and 4 major hurricanes, with an average ACE index of 167% of the median. NOAA classifies nine of the thirteen seasons since 1995 as above normal, with seven being hyperactive (ACE > 175% of median). Only four seasons since 1995 have not been above normal. These include the three El Niño years (1997, 2002, and 2006) and the 2007 season.
2) La Niña or neutral conditions during the peak months (August-October) of hurricane season. The current weak La Niña episode has substantially weakened the upper-level west-to-east winds over the tropical Atlantic. These winds create much of the wind shear that inhibits hurricane development. La Niña is expected to persist (or weaken slightly to neutral) by hurricane season, and these weak westerly winds are predicted to persist over the Atlantic, bringing low levels of wind shear.
Figure 1. SST image from NOAA's May 22, 2008 seasonal hurricane forecast.
Where will this year's storms hit?
NOAA does not present any forecasts of where the steering currents might take this year's hurricanes. In general, steering currents are not predictable more than 5-7 days in advance, although sometimes one can stretch this to two weeks. As far as U.S. landfalls go, I believe that the climatology of landfalls since 1995 offers the best forecast. Between 1995-2007, the U.S. experienced an average of 4.7 named storms making landfall, of which 1.9 were hurricanes and 0.8 were intense hurricanes.
How believable is the NOAA forecast?
Mathematical analysis of forecasts made in late May and early June by two groups--Tropical Storm Risk Inc., and the Colorado State University group led by Phil Klotzbach and Dr. Bill Gray--have shown that these forecasts had reasonable skill over the past decade or so. Unfortunately, NOAA has never performed a mathematical analysis of their May forecasts, so we don't know whether their forecasts have any skill. Their forecast did poorly last year, and the head of NOAA's forecast team, Dr. Gerry Bell, provides an interesting analysis of why last year's forecast failed.
Commentary
NOAA's forecast is a reasonable one--but then, since 1995, it's always been smart to forecast an above-average hurricane season, barring an obvious El Niño event. I like that NOAA is looking to insert more uncertainty into their forecast by providing a percent chance of their expected numbers to verify. However, they really need to provide some verification numbers of how much skill these forecasts have, if they want to make them more useful. In addition, they should consider dropping the "numbers" forecast of named storms, since the number of named storms often does not give a true picture of how active a hurricane season is. For example, recall the 15 named storms from 2007, which was about 50% above average! Yet the ACE index, a more true measure of the destructive power of the storms, was 31% below average. ACE index is a better measure because it uses the square of each storm's peak winds, summed up over the lifetime of the storm. Wind damage from a hurricane is proportional to the square (and possible the cube) of the peak winds, and total damage increases by a factor of 5-10 for each Saffir-Simpson Category between Cat 1 and Cat 4. According to Pielke and Landsea (1998), a doubling of wind speed--from a 74 mph Category 1 hurricane to a 148 mph Category 4 hurricane--has historically resulted in 250 times more damage for U.S. landfalling hurricanes. This figure includes storm surge, flooding, and wind damage.
The Colorado State University group led by Phil Klotzbach and Dr. Bill Gray issues their seasonal hurricane forecast on June 3 this year, and I'll be sure to provide additional commentary then.
Jeff Masters
The forecasters considered two main climate signals to make their forecast:
1) Continuation of conditions that have been conducive to above-normal Atlantic hurricane activity since 1995, including above-average sea-surface temperatures in the eastern tropical Atlantic Ocean (Figure 1). Between 1995-2007, the Atlantic has averaged 14.5 named storms, 8 hurricanes, and 4 major hurricanes, with an average ACE index of 167% of the median. NOAA classifies nine of the thirteen seasons since 1995 as above normal, with seven being hyperactive (ACE > 175% of median). Only four seasons since 1995 have not been above normal. These include the three El Niño years (1997, 2002, and 2006) and the 2007 season.
2) La Niña or neutral conditions during the peak months (August-October) of hurricane season. The current weak La Niña episode has substantially weakened the upper-level west-to-east winds over the tropical Atlantic. These winds create much of the wind shear that inhibits hurricane development. La Niña is expected to persist (or weaken slightly to neutral) by hurricane season, and these weak westerly winds are predicted to persist over the Atlantic, bringing low levels of wind shear.
Figure 1. SST image from NOAA's May 22, 2008 seasonal hurricane forecast.
Where will this year's storms hit?
NOAA does not present any forecasts of where the steering currents might take this year's hurricanes. In general, steering currents are not predictable more than 5-7 days in advance, although sometimes one can stretch this to two weeks. As far as U.S. landfalls go, I believe that the climatology of landfalls since 1995 offers the best forecast. Between 1995-2007, the U.S. experienced an average of 4.7 named storms making landfall, of which 1.9 were hurricanes and 0.8 were intense hurricanes.
How believable is the NOAA forecast?
Mathematical analysis of forecasts made in late May and early June by two groups--Tropical Storm Risk Inc., and the Colorado State University group led by Phil Klotzbach and Dr. Bill Gray--have shown that these forecasts had reasonable skill over the past decade or so. Unfortunately, NOAA has never performed a mathematical analysis of their May forecasts, so we don't know whether their forecasts have any skill. Their forecast did poorly last year, and the head of NOAA's forecast team, Dr. Gerry Bell, provides an interesting analysis of why last year's forecast failed.
Commentary
NOAA's forecast is a reasonable one--but then, since 1995, it's always been smart to forecast an above-average hurricane season, barring an obvious El Niño event. I like that NOAA is looking to insert more uncertainty into their forecast by providing a percent chance of their expected numbers to verify. However, they really need to provide some verification numbers of how much skill these forecasts have, if they want to make them more useful. In addition, they should consider dropping the "numbers" forecast of named storms, since the number of named storms often does not give a true picture of how active a hurricane season is. For example, recall the 15 named storms from 2007, which was about 50% above average! Yet the ACE index, a more true measure of the destructive power of the storms, was 31% below average. ACE index is a better measure because it uses the square of each storm's peak winds, summed up over the lifetime of the storm. Wind damage from a hurricane is proportional to the square (and possible the cube) of the peak winds, and total damage increases by a factor of 5-10 for each Saffir-Simpson Category between Cat 1 and Cat 4. According to Pielke and Landsea (1998), a doubling of wind speed--from a 74 mph Category 1 hurricane to a 148 mph Category 4 hurricane--has historically resulted in 250 times more damage for U.S. landfalling hurricanes. This figure includes storm surge, flooding, and wind damage.
The Colorado State University group led by Phil Klotzbach and Dr. Bill Gray issues their seasonal hurricane forecast on June 3 this year, and I'll be sure to provide additional commentary then.
Jeff Masters
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.