Colorado State University Forecast Q&A

 

Following are frequently asked questions about Colorado State University's hurricane forecasting:

 

Q: Why do you issue seasonal hurricane forecasts?

 

A: There is an inherent curiosity amongst the general public about how active or inactive the coming season is likely to be. Using historical data, there is considerable hindcast (using the past to predict the future) skill available for predicting the upcoming season. However, one must realize that these are statistical forecasts which will fail in some years. We find that we learn a lot from our forecast errors. Our end-of-the-season verifications give much information on explaining what the factors were that dictated the number and frequency of storms. Some of these factors may not have been considered in our forecasts for that particular year, and we often add new predictors in a quantitative or qualitative manner based on our end-of-the-season verifications.

There is also an educational component to these forecasts. For example, it was discovered about 25 years ago that El Nino reduced hurricane activity in the Atlantic. Through the issuing of these seasonal forecasts, this relationship has become well-known amongst the general public. Also, these seasonal hurricane forecasts have taught us many new relationships between climate features and Atlantic basin hurricanes such as sea surface temperatures, sea level pressures and levels of vertical wind shear in the tropical Atlantic.

Q: Should coastal residents prepare differently if an active or inactive season is predicted?

 

A: Coastal residents need to prepare for every hurricane season, regardless of seasonal predictions. There is inherent uncertainty in seasonal predictions. Also, seasonal forecasts do not say anything about when or where storms are going to make landfall. This information is typically only available 3-5 days before a storm actually makes U.S. landfall.

Coastal residents need to realize that the probability of landfall for any one point along the coastline is quite small in any year. However, one must also realize that it only takes one storm making landfall in your neighborhood to make it an active season for you. Major hurricanes have made U.S. landfall in inactive seasons (e.g., Hurricane Alicia - 1983 and Hurricane Andrew - 1992).

Q: How are hurricane insurance rates set? (answer provided by Insurance Services Office, Inc.)

 

A: While procedures vary, insurers generally use sophisticated probabilistic models to assess the wide range of hurricane activity that can occur over the long-term. These models use a meteorological database of tropical cyclones of over 100 years; sophisticated wind field algorithms; actual historical hurricane characteristics and costs; engineering expertise; etc. to develop reliable estimates of expected hurricane losses -- by geographic area, by construction type, by deductible purchased, etc. These models provide a much more stable and reliable measure of insurance rates than would result from the use of past hurricane claims alone -- because hurricane activity is very volatile from year to year. If such volatile hurricane loss data were to be primarily relied upon, rather than stochastic models, insurance rates could swing wildly upward and downward from one year to the next -- solely due to the occurrence (or absence) of a major hurricane. So, in addition to bringing more accuracy to the ratemaking process, probabilistic models also enhance rate stability -- a main actuarial consideration in ratemaking.

Of course, new information emerges when hurricanes do occur. For example, new knowledge about the performance of various construction types may become available which can further refine future loss estimates. So, while hurricane models, reinsurance, and other mechanisms greatly help mitigate the "shock" effects of new hurricane events on insurance rates, it is natural and appropriate to expect some changes as well.

Q: How do I find the probabilities of landfall for my region, sub-region or county?

 

A: Probabilities of landfall are available from the following webpage: http://www.e-transit.org/hurricane. From this webpage, click on Interactive Landfall Probability Display, then click on the county for probabilities. Or you can obtain probabilities from a spreadsheet located here: http://www.e-transit.org/hurricane/Full Data Table.xls

 

Q: I've noticed that we've had a lot more activity in the Atlantic since 1995. Is global warming the cause of this increased activity?

 

A: The Atlantic basin tends to go through periods of about 25-35 years with heightened major (Category 3-4-5) hurricane activity and then similar periods of about 25-35 years with less activity. We believe that these multi-decadal variations are mostly due to changes in large-scale ocean circulations referred to as the Atlantic Ocean thermohaline circulation. For a more detailed discussion, please refer to the following webpage and read Section 9: http://tropical.atmos.colostate.edu/Forecasts/2008/april2008/apr2008.pdf.

 

Q: How accurate are your forecasts?

 

For a complete spreadsheet with all of our forecasts from 1948-2012, view the following MS Excel file: http://tropical.atmos.colostate.edu/Includes/Documents/Publications/forecast_verifications.xls

 

Our forecasts for June and August that have been issued since 1984 show significant skill when evaluated against climatology or the previous five-year mean. Our earlier seasonal forecast, issued in early April, shows somewhat less skill.

 

Q: Why do you issue a forecast during the middle of the season (August 1)?

 

A: Although the Atlantic basin hurricane season starts on June 1, more than 90 percent of all tropical cyclone activity and 95 percent of major hurricane activity occurs after August 1 in an average season. In general, our seasonal forecasts issued on August 1 show the greatest skill.

 

Q: How can you issue a forecast so far in advance of the season (e.g., April 1) when scientists generally can't issue skillful forecasts for more than one to two weeks in advance?

 

A: The issue at hand is the difference between statistical and dynamical forecasts. Dynamical forecasting, which takes initial value observations and integrates them forward in time, usually loses all skill after about 2 weeks. Statistical forecasting uses large-scale ocean/atmosphere empirical relationships to issue forecasts. Much more in-depth discussions of statistical versus dynamical forecasting are available in Section 1 here: http://tropical.atmos.colostate.edu/Forecasts/2006/june2006/jun2006.pdf

 

Q: How does your forecast compare to the forecast issued by the National Oceanic and Atmospheric Administration (NOAA)?

 

A: The Colorado State University forecasts are issued independently of NOAA. For the latest forecast from NOAA, visit the following webpage:

 

http://www.cpc.ncep.noaa.gov/products/outlooks/hurricane.shtml

 

NOAA makes two forecasts per year (late May and early August) and issues a prediction for a range of named storm, hurricane, major hurricane and Accumulated Cyclone Energy (ACE) index values. NOAA does not issue specific two-week forecasts or landfall probability forecasts.

 

Q: Do other groups issue Atlantic basin seasonal hurricane forecasts?

 

A: Yes, several other groups issue Atlantic basin forecasts. Among these are Tropical Storm Risk (TSR) and the Cuban Institute of Meteorology. TSR's forecasts are located here: http://www.tropicalstormrisk.com

 

The Cuban Institute of Meteorology's forecasts are located here: http://www.met.inf.cu/

 

Q: How much activity occurs in an average year before August 1?

 

A: On average, 1.4 named storms, 0.6 hurricanes and 0.1 major hurricanes occur before August 1. The average full season between 1950 and 2000 witnessed a total of 9.6 named storms, 5.9 hurricanes and 2.3 major hurricanes.

 

Q: When is the most active part of the average Atlantic basin hurricane season?

 

A: About 90 percent of all Atlantic hurricane activity takes place during the months of August, September and October. The most active 30-day period during the season is approximately August 25-September 25.

 

Q: Does an active June-July mean that the rest of the season is likely to be more active?

 

A: In general, June-July activity does not correlate very well with the rest of the season's activity. However, if storms form in the deep tropics (i.e., south of the Tropic of Cancer - 23.5N - and east of the Leeward and Windward Islands) during June and July, this, in general, means that the season is going to be very active.

 

Q: Do you issue forecasts for other basins (e.g., the West Pacific, East Pacific)? Does anyone issue forecasts for these basins?

 

A: We do not issue forecasts for other tropical cyclone basins. NOAA issues forecasts for the East Pacific, and these forecasts are available here: http://www.cpc.ncep.noaa.gov/products/Epac_hurr/Epac_hurricane.html

 

Tropical Storm Risk issues forecasts for several tropical cyclone basins. Their forecasts are available here: http://www.tropicalstormrisk.com/

 

The City University of Hong Kong issues forecasts for the West Pacific. Their forecasts are available here: http://weather.cityu.edu.hk/. Prof. Johnny Chan issues these forecasts. He is a former graduate student of Bill Gray.

 

Q: Is there an FAQ site available with answers to general questions about hurricanes?

 

A: Chris Landsea, Science and Operations Officer at the National Hurricane Center, maintains a detailed FAQ webpage available here: http://www.aoml.noaa.gov/hrd/tcfaq/tcfaqHED.html

 

Q: I'm traveling to Florida in August. What are the probabilities of landfall during that time period?

 

A: Currently, our landfall probability webpage does not provide these shorter-period probabilities. We intend to include this information in the future. Useful maps of landfall probabilities for shorter periods are available from Chris Landsea's Hurricane FAQ webpage: http://www.aoml.noaa.gov/hrd/tcfaq/G13.html

 

Q: Does Colorado State University issue forecasts for individual storms?

 

A: No. Colorado State issues seasonal and two-week forecasts, that is, they predict how many storms are likely to occur during an individual season or two-week period. Individual storm track and intensity predictions are only issued by NOAA's National Hurricane Center. Their forecasts are available here: http://www.nhc.noaa.gov

 

For questions about individual storms, please contact Dennis Feltgen, NOAA Public Affairs Officer for the National Hurricane Center at dennis.feltgen@noaa.gov.

 

Q: Why are we performing research forecasting hurricanes from Colorado?

 

A: Herbert Riehl, who started the Atmospheric Science department at Colorado State University in 1962, was a world-renowned hurricane researcher in the 1950s at the University of Chicago. One of his doctoral students and researchers at the time was Bill Gray. Gray followed Riehl from the University of Chicago to Colorado State University and has been a professor at Colorado State University since 1964.

 

Q: How long has Dr. Gray been issuing hurricane forecasts?

 

A: Dr. Gray has been issuing Atlantic basin seasonal hurricane forecasts since 1984 (30 years) and has been studying hurricanes for over 50 years.

 

Q: Is Dr. Gray retiring?

 

A: Dr. Gray has officially retired as a professor from Colorado State University. He is now an Emeritus Professor and is free to spend more time in his research endeavors. He plans to be involved in the issuing of the seasonal hurricane forecasts for the foreseeable future. He is also deeply engrossed in trying to better understand the links between climate change, global warming and hurricanes.

 

Q: How active was the 2005 season?

 

A: The 2005 hurricane season broke many records with its level of activity. A total of 27 named tropical storms and one unnamed subtropical storm formed during the year, along with 15 hurricanes and seven major (Category 3-4-5) hurricanes. The 27 named tropical storms broke the old record of 21 named tropical storms set in 1933 (although several named storms may have been missed in 1933 since there was no satellite monitoring or aircraft reconnaissance at that time). The 15 hurricanes broke the old record of 12 hurricanes set in 1969. The seven major hurricanes did not break the old record of eight major hurricanes set in 1950. These types of unusually active seasons occur about once every 40-50 years.

 

Q: What conditions came together to make the 2005 season so active?

 

A: A combination of several features came together to make 2005 active. There was very low vertical wind shear, especially in the Caribbean during the August-October period. In addition, warm sea-surface temperatures and abundant moisture dominated the tropical Atlantic providing very favorable conditions for hurricane development and intensification. Refer to our 2005 forecast verification for a more in-depth discussion of features in place that made the 2005 hurricane season so active. The forecast verification is available here: http://tropical.atmos.colostate.edu/Forecasts/2005/nov2005/