Hurricane Forecast Update: William Gray and Colorado State Team Expect Lower-Than-Average Storm Activity in August and September

Note to Editors: Forecast totals are in the attached chart. The complete hurricane forecast, including the new August, September and October monthly forecasts, a detailed description of forecast factors, press releases and access to downloadable audio and video clips, is available on the Web at

With four named storms and two hurricanes forming in the Atlantic before the end of July, the 2003 hurricane season is off to a strong start. However, William Gray and the Colorado State University hurricane forecast team issued a report today calling for a change in this pattern and predict lower-than-average storm activity for August and September. Gray and his team slightly reduced their seasonal forecast and are calling for near average storm activity for the remainder of the 2003 hurricane season, but maintain their prediction for an above-average season overall.

"Based on atmospheric changes from June to July that resulted in a changed global circulation pattern, we have slightly decreased our seasonal storm activity forecast," said Gray. "We expect storm activity in August and September to be below average and October to be a bit more active than normal. Overall, we think the remainder of the 2003 Atlantic basin storm season will be about average, leaving the entire season still somewhat above average."

As detailed in today’s update, which can be accessed on the Web at, Gray and his colleagues maintain their call for 14 named storms to form in the Atlantic basin this year. Of the 14 storms, eight are predicted to become hurricanes and three are anticipated to evolve into intense hurricanes (Saffir/Simpson category 3-4-5) with sustained winds of 111 mph or greater. These numbers mirror the team’s May 30 forecast: although the storm activity for August through October is now expected to be only about average, the numbers remain the same because of such an active early season.

The updated forecast reduction totals are seen primarily in the amount of days that storms are expected to last. The predicted total of named storm days, or the number of days throughout the season on which named storms occur, was reduced from 70 days to 60 days. The forecast number of hurricane days was also reduced from 35 days to 25 days, and intense hurricane days have dropped from eight days to five days. The team also reduced the season’s overall storm activity from 145 percent of the long-term average to 120 percent of average.

The long-term average is 9.6 named storms, 5.9 hurricanes and 2.3 intense hurricanes per year, with 49.1 named storm days, 24.5 hurricane days and 5.0 intense hurricane days. Last year there were 12 named storms but only four hurricanes and two intense hurricanes.

For the second year, Gray and his team are also issuing storm activity forecasts for specific months. Along with today’s seasonal forecast, Gray and his colleagues issued individual monthly forecasts for August, September and October. The monthly forecasts use different parameters than the seasonal forecasts to predict storm activity within shorter time periods. The individual monthly forecasts aid with the seasonal predictions and can be found included in today’s updated forecast report.

For the month of August, Gray and his team forecast three named storms, one hurricane and one intense hurricane for the Atlantic basin. For September, the team predicts four named storms, two hurricanes and one intense hurricane. For October, three named storms, two hurricanes and no intense hurricanes are forecast.

"The same factors that can make individual months active or inactive are often not the same factors that can make the entire season active or inactive," said Philip Klotzbach, an atmospheric science researcher and member of Gray’s forecast team. "We are continually improving our forecasts to provide people with specific monthly hurricane forecasts and specific landfall probability forecasts."

As part of today’s updated report, the Colorado State team slightly reduced their assessment of hurricane landfall probability. However, the probabilities for landfalling hurricanes for the remainder of the season are still above average.

"This is better news for coastal residents. However, this does not mean that there will not be significant United States and Caribbean hurricane-spawned destruction," said Gray. "The probabilities for landfalling major hurricanes for 2003 are still above the long-period yearly average."

According to today’s forecast, there is a 65 percent chance of a major hurricane hitting somewhere along the U.S. coastline in 2003 (the long-term average is 52 percent). For the U.S. East Coast, including the Florida Peninsula, the probability of an intense hurricane making landfall is 44 percent (the long-term average is 31 percent). For the Gulf Coast from the Florida Panhandle westward to Brownsville, the probability is 36 percent (the long-term average is 30 percent). The team also expects an above-average major hurricane landfall risk in the Caribbean.

The last eight years have witnessed 106 named storms, 62 hurricanes and 29 major hurricanes in the Atlantic basin. During that period, only three of the 29 major hurricanes (Opal, Bret and Fran) crossed the U.S. coastline. Based on historical averages, one in three major Atlantic basin hurricanes comes ashore in the United States. Before Lili made landfall as a category 2 hurricane in October 2002, a record 21 consecutive Atlantic basin hurricanes formed without making landfall along the U.S. coast.

"The United States has been very lucky over the past three decades in witnessing very few major hurricanes making landfall in Florida and along the East Coast. At the same time, we have seen large coastal population growth," said Gray. "Regardless of whether a major hurricane makes landfall this year, it is inevitable that we will see hurricane-spawned destruction in coming years on a scale many times greater than what we have seen in the past."

The storm seasons spanning 1995-2002 comprised the most active eight consecutive hurricane years on record, and the Colorado State forecasting team believes that the Atlantic basin has entered a new multi-decadal cycle of increased hurricane activity and landfall probability as was witnessed in the 1940s and 1950s.

The Colorado State forecast team does not attribute changes in recent years and projected Atlantic hurricane activity to human-induced global warming or any other human-caused phenomenon. They believe the changes are a natural consequence of climate variability that has been a continuing feature of atmosphere-ocean changes since the last Ice Age.

Gray and his team are continually working to improve their forecast methodologies based on a variety of climate-related global and regional predictors.

"We are also performing new research to forecast the month-to-month variability of Atlantic basin hurricane activity and landfall probability," said Gray. "Our forecast accuracy has been improving in recent years, and we feel ongoing research will add further improvement and understanding."

The research team will issue seasonal summary updates and outlooks for the 2003 Atlantic basin hurricane activity on Sep. 3 and Oct. 2. The updates will be available on the team’s Web site at

In addition to Gray, team members include Christopher Landsea, Philip Klotzbach, Eric Blake, William Thorson, Jason Connor and others.


Tropical Cyclone Parameters and 1950-2000 Climatology (in parentheses) Extended Range Forecast for 2003 April 4, 2003 Update May 30, 2003 Update August 6, 2003 Update
Named Storms (9.6)* 12 12 14 14
Named Storm Days (49.1) 65 65 70 60
Hurricanes (5.9) 8 8 8 8
Hurricane Days (24.5) 35 35 35 25
Intense Hurricanes (2.3) 3 3 3 3
Intense Hurricane Days (5.0) 8 8 8 5
Hurricane Destruction Potential (72.7) 100 100 100 80
Net Tropical Cyclone Activity (100%) 140 140 145 120

* Number in ( ) represents average year totals based on 1950-2000 data.

Note to Editors: The National Oceanographic and Atmospheric Administration (NOAA) also has begun to issue Atlantic basin seasonal hurricane forecasts. The NOAA forecasts are independent of the Colorado State University forecasts although they use prior Colorado State research which NOAA augments by their own insights. The public should not expect the NOAA and Colorado State forecasts to necessarily be in agreement.