Musings: CSU Hurricane Forecast Trimmed Once Again
As we write this article the first tropical storms of the season have formed In the Atlantic basin and Gulf of Mexico. The latest forecast from the Colorado State University (CSU) for this year's tropical storm season authored by lead scientists Dr. Phillip Klotzback and Dr. William Gray of the Department of Atmospheric Science shows another reduction from their prior forecasts. The team now looks for a total of 10 tropical storms with only four becoming hurricanes and two of them significant hurricanes (Category 3, 4 or 5).
The official forecast says they "anticipate a below-average Atlantic basin tropical cyclone season in 2009," which is attributed to the development of El Niño in the Pacific Ocean. The new forecast also anticipates below-average probabilities for major hurricane landfalls on the U.S. coastline and in the Caribbean.
As the El Niño phenomenon developed this year, the CSU forecasters pointed out the potential for its dampening effect on tropical storm formation when issuing earlier forecasts. The CSU team has been gradually reducing its projected number of storms by category and the corresponding number of storm days. With reduced expectations for the number of storms this season, especially since two of the storm season's six months have passed with no storms, the probabilities of major hurricane landfalls have also declined, but by a smaller reduction than the number of storms since projecting a storm's path before it forms is difficult. As we all know, it only takes one storm making landfall, and it doesn't have to be among the strongest storms, to cause loss of life and extensive damage while creating extreme chaos.
Despite the challenge of predicting landfalls, the CSU forecasters have been issuing probability forecasts for a few years for each of the 11 coastal regions of the United States and 205 coastal and near-coastal counties from Brownsville, Texas, to Eastport, Maine. The team believes developing better landfalling forecasts is an important benefit of their work since it allows earlier and presumably better storm preparation.
In the current forecast, for the first time, the CSU team prepared landfall probabilities for the 18 coastal states using data from 1856- 2008. They have also prepared landfall probabilities for the Mid- Atlantic and Northeast U.S. regional groupings of states. The CSU forecasters showed how the historical landfall probability for both hurricanes and major hurricanes for selected states and regions is reduced by the formation of El Niño.
One of the interesting phenomenons impacting the formation of hurricanes appears to be the relative warmth of the Atlantic basin coupled with the strength of El Niño. From May to July, the tropical Atlantic has warmed by 0.25-0.5°C. Despite this warming, the Atlantic remains slightly cooler than the 1995-2008 average. What also has been learned from the historical data is that the ratio of named storms forming south of 23.5°N and east of 75°W with
storms forming north of 25°N changes considerably between neutral and La Niña years to El Niño years. The CSU forecasters have determined that from 1950-2008, approximately 45% of all storms form north of 25°N in a La Niña year or a neutral year compared to 60% of all storms forming north of 25°N in El Niño years.
The CSU report contained a chart showing the number of storms and their tracks below that cut-off in the six warmest El Niño years and the six coldest La Niña years. The ratio between the aggregate numbers of storm days during La Niña years to those in El Niño years is 4.1:1, a significant difference. Based on this data and the CSU team's assessment of the future weather patterns that will be in existence during the mid-August to mid-October period, it believes conditions facilitating the formation of Atlantic basin hurricanes will be moderated.
The other forecasting change the CSU team has introduced is a twoweek forecast rather than its traditional monthly forecast. We assume this change is an attempt to be timelier in its forecasting than it has been in the past. Shortly after issuing the revised 2009 hurricane seasonal forecast, the CSU team issued a forecast of the August 6-20 period. The team believes this will be a below average period compared to the historical data for 1950-2000.
The team's forecast is backed additionally by a chart from the National Oceanic and Atmospheric Administration (NOAA) showing the current forecast period compared to the climatology record, which shows storm activity traditionally picking up in early August. Activity peaks in early September but continues strong until the latter part of October.
While we can take some comfort in the revised CSU forecasting team's outlook for a below-average storm season, the fact we are just entering the peak seasonal storm period should be a reminder to be on your guard and ready for action. This is the latest formation of the first storm. In 1992, the first storm developed late in the season. It was Hurricane Andrew that devastated south Florida.