Hurricanes surely are the most visible and violent natural phenomena encountered on the Gulf Coast. A hurricane is an intense, whirling storm similar in character to other rotating storms in the atmosphere on larger (midlatitude storms) and smaller (tornados) scales. Nevertheless, while storms may produce winds and rain over a broad region and tornados generate destructive winds over a small area, hurricanes are unique in their ability to produce both devastating winds and intense rainfall for hundreds of kilometers around the storm center. Although about 85 hurricanes occur each year in all the world’s tropical oceans, only a fraction in the North Atlantic pose threats to the United States. A still smaller fraction actually makes landfall, and only a subset of these affect the Gulf Coast.
Hurricanes are distinguished by their driving force—heating from an underlying warm ocean—and by their intensity. A category 1 hurricane possesses winds in excess of 74 miles per hour (33 meters per second), whereas a category 5 has winds greater than 157 miles per hour (70 meters per second). Winds in excess of 179 miles per hour (80 meters per second) have been recorded. Once formed, a hurricane will persist unless it is perturbed by internal instabilities, cold water, friction from the underlying land surface, or winds aloft that change its vertical structure. The high winds are of particular significance since they are the primary force behind much of the damage that occurs. Direct wind damage increases roughly with the cube of the wind speed. Economic losses vary with the fourth power of the wind speed. Storm surges driven by a hurricane’s low pressure and destructive winds are major killers that can lead to complete destruction of low-lying coastal communities and industries. As the hurricane moves inland, heavy rain and associated flooding become the major contributors to damage and loss of life. As we have seen in New Orleans, flooding damage and inundation also can last for days and sometimes months.
Hurricanes are among the most destructive elements of nature and are the cause of the largest number of fatalities related to natural disasters in the Western Hemisphere. For this reason, great resources are expended on preparation, protection, and response to hurricanes. Specialized aircraft reconnaissance, advanced satellite systems, and enhanced coastal observing systems are utilized to provide essential data for hurricane analysis and prediction. Since the advent of satellite observations, all hurricanes are detected long before landfall.
Predictions of hurricane paths have improved markedly due to dedicated research efforts and advances in numerical forecast models, but predictions of hurricane intensity and detailed structure have made very little progress. Despite the urgency of the problems faced, there has been an alarming decay in the resources provided for hurricane research and development in the past decade. Urgent action is needed to reverse this trend and increase support for multidisciplinary approaches to ameliorating the impact of these dangerous systems.
Uncertainty in predictions requires that hurricane warnings cover long coastal stretches that could be threatened. As a result, many communities incur unnecessary costs in preparation and/or evacuation. If the forecasts of landfall and intensity are made in probabilistic form, regional and local authorities can conduct risk analyses that allow reasoned decision-making processes in which the costs of evacuation are compared with the potential cost of damage and loss of life. Thus improved probabilistic forecasts have a direct benefit in reducing the secondary economic and social impacts of hurricanes.
Role of the Ocean
The transfer of heat from the ocean to the atmosphere is the primary driver of hurricanes. Viewed as a heat engine, a hurricane extracts heat energy from the warm ocean surface and exhausts it into the upper atmosphere where excess energy is radiated to space and transported toward higher latitudes. Hurricanes require sea surface temperatures (SST) above 26 degrees Celsius (79 degrees Fahrenheit) to form. A hurricane’s maximum strength is governed largely by the amount of energy it can extract from the underlying ocean. In general, the warmer the SST, the more intense the hurricane. The overall correlation between SST and intensity is relatively small (0.35). When the SST is substantially higher than the threshold temperature, then intense hurricanes can form. The reason the correlation is low is because lower-category storms also form in this high SST regime, but other factors such as vertical wind shear reduce their potential intensity.
The tropical ocean is strongly stratified with relatively warm surface waters, usually a few tens of meters thick, overlaying colder subsurface waters. As a hurricane travels over the ocean it generates a cold-water wake produced both by wind mixing the warm surface water with colder water below and by extraction of heat through evaporation that fuels the hurricane. This cold water limits the hurricane’s strength by reducing its energy supply and similarly limits the intensity of succeeding hurricanes that might follow.
In some regions of tropical oceans the warm surface layer is thicker, reaching more than 100 meters (329 feet). Such a layer contains a great deal of heat energy and vertical mixing, but the mixing may not be sufficient to bring cold subsurface waters to the surface. Tropical cyclones often intensify rapidly while passing over such regions, where the ocean surface remains warm. For example, Hurricane Katrina intensified first over a warm eddy and then strengthened very rapidly from category 1 to category 5 as it passed over deep warm waters in the Gulf.
The primary ocean data required for hurricane prediction are the SST and the heat content of the warm layer, referred to specifically as the tropical cyclone heat potential. Sea surface temperature is observed routinely by satellites through infrared and microwave emission sensors. Sea surface temperature is more difficult to measure in cloudy conditions such as occur in the disturbed tropical environment. Only research satellites such as the Tropical Rainfall Measuring Mission (TRMM) currently possess this crucial capability to measure through clouds. Measurement of ocean heat content requires additional observations of temperatures within the ocean—either direct temperature measurements by expendable probes or indirect inference from satellite-measured sea surface height.
Role of the Surrounding Atmosphere
The movement, initial development, and intensification of hurricanes are in response to their near environment. The major factors are the capacity to sustain convective clouds, vertical wind shear, and the degree of midlevel moisture. Vertical wind shear has a negative effect on intensification and especially on initial development. A dry middle level can induce cool, dry downdrafts from convective systems that impair a hurricane’s capacity to take up energy from the ocean surface.
In the decade since 1995, sustained changes in the broad flow over the equatorial North Atlantic have produced a distinct change in the numbers and characteristics of hurricane developments. Thirty percent of the waves entering the North Atlantic basin developed into tropical storms in 2005, compared with the normal 10 percent. Recent research has shown that the increased number resulted from variations in the tropical trade winds that flow across the North Atlantic.
In the 2005 hurricane season, there were 27 named storms in the North Atlantic. Nineteen of those formed from waves in the westward flowing tropical trade winds, principally in the eastern portion of the basin.
Current observing approaches utilize advanced satellite sensors and aircraft reconnaissance to target regions where observations are required, both within the hurricane and in its surrounding environment. These resources are under threat from budget reductions, delays in implementing new systems, and delays in replacement of existing systems currently depended on for monitoring and prediction.
There currently is insufficient skill in empirical predictions of the number and intensity of storms in the forthcoming hurricane season. Predictions by statistical methods that are widely distributed also show little skill, being more often wrong than right. Advanced global models are beginning to show some ability to predict seasonal characteristics. Examples include the coupled ocean-atmosphere climate models used in extended prediction by Meteo-France, the European Centre for Medium-Range Weather Forecasts (ECMWF), and the UK Meteorological Office. On 1 May 2005, the Meteo-France model predicted 22 named tropical storms and hurricanes for the 2005 hurricane season in the North Atlantic. On 1 June the ECMWF and the UK Met Office integrations were showing similar results. What was extraordinary about these forecasts was that their predictions, some months in advance of the hurricanes, were two standard deviations above the already elevated 1995–2004 mean. These models also forecast a reduced number of storms for the northwestern Pacific during the same period. In hindcast mode these three models have outperformed statistical forecasts over the previous 10-year period of elevated storm activity. Yet despite these successes and the clear promise of the techniques, no operational model within NOAA is making extended range forecasts with climate models.
Major improvements in hurricane track forecasts have been made in recent years. The 72-hour forecast is now approaching the skill of a 24-hour forecast two decades ago. Four- to five-day forecasts also are increasingly skillful. Nevertheless, there has been little improvement in the forecasting of hurricane intensity that currently relies primarily on statistical techniques. High-resolution models such as the Advanced Research Weather (ARW) model of the National Center for Atmospheric Research (NCAR) have demonstrated that such models can yield impressive capabilities for prediction of both intensity and track. NOAA is developing an operational version of the ARW for implementation in 2007, but this will not have sufficient resolution to replicate the improvements noted in the NCAR model.
The most immediate need is to improve the range, accuracy, and precision of real-time predictions of hurricane track, intensity, and structure. This will require continuing improvements in observations and forecast models.
The most urgent observational needs are for maintenance of the current atmospheric observing systems and further development of new capabilities.
For aircraft observations, the implementation of a Stepped Frequency Microwave Radar on U.S. Air Force Hurricane Hunter aircraft will lead to substantial improvements in observing the ocean surface and surface winds.
Unmanned aircraft systems have promise for obtaining direct observations of the near-surface wind conditions that are critical for many aspects of the overall forecast approach including surge modeling and predicting wind damage impacts.
Satellite remote sensing remains one of the most powerful tools, especially the space radars, scatterometers, and multispectral systems. Improvements in the ocean observing system are also vital.
Major improvements are needed for the assimilation of all available and relevant atmospheric and ocean data in real time into numerical predictions. At present, much data that are potentially usable for predictive models, for example, hurricane reconnaissance radar data, are not utilized operationally or inserted into the models as initial data.
Prediction models offer promising prospects for improvement through research.
Better coupling between ocean and atmosphere in numerical models is needed. There are still grave uncertainties in the magnitudes of evaporation and sensible heat exchange in high-wind environments. Most parameterizations have only been validated for wind speeds less than 45 miles per hour (20 meters per second). The Coupled Boundary Layers/Air-Sea Transfer (CBLAST) program, in which buoys were dropped into the path of a hurricane, has provided some clues as to the magnitude of these transfer coefficients, but funds have not been available for a thorough analysis of the data.
The effects of ocean waves on the transfer of heat and momentum between the atmosphere and the ocean are not well modeled. Without additional research on the interchange processes between the atmosphere and the ocean, progress in modeling and prediction will be limited.
Nested high-resolution models should be incorporated increasingly into operational practice, despite current uncertainties in some physical processes. The research community is developing advanced atmospheric and ocean models, covering limited areas with high resolution, that are nested within the larger-scale models used for operational weather forecasting.
Atmospheric and ocean predictions should be coupled to storm surge models to improve prediction of flooding and inundation. This also would aid in planning reconstruction.
NOAA should explore the use of advanced climate simulation models for extended prediction of hurricane characteristics.
In the longer term, collaboration between research and operational centers should be fostered in both development and application of hurricane prediction models. It is clear that current collaborations are not optimal and are in need of sustained effort to make the best use of the substantial investment that is currently made in these areas. The Gulf Coast will continue to be affected by hurricanes for the foreseeable future. Indications are that the effects may worsen. A combination of improved research and operational applications of this research, especially through numerical models, will prove to be a prudent and productive investment in improved hurricane predictions.