GFDL - Geophysical Fluid Dynamics Laboratory

Skip to main content

Large-scale Climate Projections and Hurricanes

View of model-projected changes in vertical wind shear (warm colors mean higher shear) to global warming.
(image by Remik Ziemlinski)
Click here for Images and Animations

See also the GFDL Hurricane Portal

Global Warming Projections and Hurricanes Activity

How hurricane (or more generally "tropical cyclone") activity will respond to human-induced global warming is a topic of much popular interest and scientific debate. Recent studies (cf. Tom Knutson's "Hurricanes and Global Warming Page") suggest that global warming may act to increase tropical cyclone activity due to a rise in ocean surface temperatures. However, there are a number of environmental factors besides ocean surface temperature which also influence the development and intensification of hurricanes, such as upper atmospheric temperature, relative humidity, and wind shear (see below).

The study of Vecchi and Soden (2007) explored changes in these environmental factors in a set of 18 'state-of-the-art' coupled climate models simulations for the 21st Century (PCMDI). These climate model simulations were performed by research laboratories all over the world in support of the IPCC Fourth Assessment Report (IPCC AR4)

These models combine our best understanding of the physical processes controlling the climate system (atmosphere, ocean, land surface and sea ice), with estimates of possible concentrations of greenhouse gases over the coming century, to provide projections of future climate changes. The models are not perfect, with uncertainties arising from our inability to fully represent certain physical processes and imperfect knowledge of future emissions of greenhouse gases and aerosols. However, climate models have proved skillful in reproducing many aspects of past climate change and are essential to making projections of future climate change.

A key finding of this study is the projected increase in wind shear over the tropical Atlantic and eastern Pacific ocean basins during the next century. Wind shear results from a change in direction or speed of winds between the lower and upper levels of the atmosphere and is widely recognized to inhibit the development and intensification of tropical cyclones. The models also project a decrease in Central and West Pacific wind shear in the 21st Century (see below or Vecchi and Soden (2007))

The increased Atlantic and East Pacific shear is a common feature of climate model projections for the 21st Century and is tied to an overall weakening of the tropical "Walker circulation" - a vast loop of winds that influences climate across much of the globe, and varies in concert with naturally-occuring El Niño and La Niña oscillations.

Because current climate models do not resolve tropical cyclones explicitly, the study focused on the models' projections of changes in large-scale environmental parameters associated with tropical cyclone activity and intensity, and NOT on simulations of tropical cyclones themselves. The net effect of increased wind shear, warmer oceans, and other environmental changes on the number and intensity of tropical cyclones will require further investigation with more detailed models. However, the current study does point out the presence of other key environmental changes over the Atlantic and eastern Pacific that are comparable in magnitude to the impact of warming oceans, but with an opposing effect on tropical cyclone activity.

 


Summary Table

The following table summarizes the general effect of the model-projected changes in parameters relevant to hurricanes over the various basins. See below or the Vecchi and Soden (2007, GRL) manuscript for a more detailed discussion.

Ocean Basins

Effect of Ocean Warming

Effect of Wind Shear Change

Effect of Relative Humidity Change

Net Effect

Atlantic and East Pacific

???

West Pacific and Indian Ocean

 

Model Projected Changes in Hurricane-relevant Parameters:

The following plots offer summaries of the "multi-model" average change in various "hurricane-relevant" parameters. The changes are computed by differencing the models' 2081-2100 projection, with their 2001-2020 projection. More details are available in Vecchi and Soden (2007). Changes in the quantities are plotted for separately for the two six month seasons of June-November and December-May.

The quantities we have considered are:

  1. Vertical wind shear:
    (Vs) Magnitude of the difference between lower (850hPa) and upper (200hPa) tropospheric winds. Large values of wind shear are historically connected with reduced tropical storm activity and intensification.
  2. Mid-tropospheric relative humidity:
    (RH700) This is the relative humidity at 700hPa. Large values of mid-tropospheric relative humidity are conducive to tropical storm activity.
  3. Emanuel's Maximum Potential Intensity:
    (MPIv) This is a measure of the thermodynamic limit on the intensity of a storm, which is based on the theory developed by Kerry Emanuel - who has a discussion of MPI Theory and Fortran Code for computing MPI. Large values of MPIv are associated with enhanced tropical storm activity and intensity.
  4. Emanuel and Nolan (2004) Genesis Potential Index:
    (GPI) This is a metric that combines the values of shear, mid-tropospheric relative humidity, MPIv and large-scale vorticity to estimate the potential for a storm to develop. Larger values of GPI are associated with enhanced tropical storm development.

Projected Changes in June-November

Click on figures for higher resolution.

Wind Shear

The model average projects a decrease in June-November wind shear over much of the Northern Hemisphere tropics, with the notable exception of the tropical Atlantic and East Pacific. This local shear increase is related to a weakening of the Pacific Walker circulation in these model projections.

Mid-tropospheric Relative Humidity

The model average projects regions of both increasing and decreasing relative humidity. Over much of the tropical Atlantic Ocean the models show a decrease in mid-tropospheric relative humidity, which would tend to be unfavorable to tropical storms. However, over the rest of the Northern Hemisphere, the models project modest and large increases in mid-tropospheric relative humidity.

Changes in modeled mid-tropospheric relative humidity appear tied to changes in circulation (shown in contours, with dashed lines indicating anomalous ascent). Regions of anomalous ascent are associated with increasing relative humidity, regions of anomalous subsidence are associated with drying.

Maximum Potential Intensity

The model average projects an increase in Emanuel (1995) maximum potential intensity (MPIv) over most of the northern hemisphere tropics. The change is on the order of a few percent per degree warming, consistent with the findings of Knutson and Tuleya (2004). However, over the tropical Atlantic Ocean there is a large swath where MPIv actually decreases in a warming climate.

The local decrease in MPIv in the tropical Atlantic is associated with a minimum in sea surface temperature warming there (contours show the relative surface warming, with dashed lines indicating regions that warm less than the tropical mean). Even though model surface temperatures are warming in the global tropics, the tropical Atlantic warms less than the rest of the tropics. Because the upper tropospheric temperature changes are tied to the changes across the entire tropics, this local minimum in warming results in stabilization of the atmosphere.

Genesis Potential Index

The model average projects an overall increase in June-November Genesis Potential Index (GPI) across the northern hemisphere tropics. However, the Tropical Atlantic and East Pacific generally have smaller changes than the other basins, with regions of both increase and decrease of GPI. The Western and Central Pacific exhibit the largest change in June-November GPI, with substantial increases; the northern Indian Ocean also shows a moderate increase in GPI.

Projected Changes in December-May

Click on figures for higher resolution.

Wind Shear

The model average projects an increase of shear across most of the subtropics, and a decrease near the Equator. In the Indian Ocean there is a decrease in shear in the region 5S-10S.

Mid-tropospheric Relative Humidity

The model average projects regions of both increasing and decreasing relative humidity.

Changes in modeled mid-tropospheric relative humidity appear tied to changes in circulation (shown in contours, with dashed lines indicating anomalous ascent). Regions of anomalous ascent are associated with increasing relative humidity, regions of anomalous subsidence are associated with drying.

Maximum Potential Intensity

The model average projects an increase in Emanuel (1995) maximum potential intensity (MPIv) over most of the tropics. The change is on the order of a few percent per degree warming, consistent with the findings of Knutson and Tuleya (2004). The regions of local minima in warming in the subtropics are associated with a decrease in MPIv, however those are regions where tropical storms do not generally form. Over regions where tropical storms tend to form and intensify ini the austral winter, MPIv tends to be projected to increase.

Genesis Potential Index

The model average projects an overall increase in June-November Genesis Potential Index (GPI) across much of the tropics. The southern Indian Ocean exhibits the largest change in December-May GPI, with a substantial increase; the western Pacific Ocean also shows a moderate increase in GPI - although there is a region in the central Pacific that shows a decrease in GPI.

 

Coming Soon

  • Plots for each of the models
  • NetCDF data of the multi-model averages
This work was a collaboration between:
Prof. Brian Soden at the Rosenstiel School of Marine and Atmospheric Science (RSMAS) at the University of Miami.
and
Dr. Gabriel Vecchi at NOAA's Geophysical Fluid Dynamics Laboratory