Severe Weather: Hurricanes Jim Kossin Cooperative Institute for Meteorological Satellite Studies University of Wisconsin—Madison Madison, WI [email protected] http://www.ssec.wisc.edu/~kossin. National Press Foundation, “Understanding Violent Weather” Program, 12 March 2007.
Operational hurricane forecasting challenges globally each year.
Hurricane track forecast errors cut in half in past 15 years globally each year.
Little increase in intensity forecast skill
Prediction of large-scale (environment) versus small-scale (hurricane)
Figures from Franklin et al. (2005)
Present skill (track and intensity)
How does the environment control hurricanes? globally each year.
Genesis: warm SST, low shear, …, an incipient vortex.
Intensity change: SST, shear, …, landfall, Saharan dust.
Track: larger-scale circulation patterns.
Need to predict the environment that the hurricanes are traveling through. (Requires good track prediction)
Suppose we could predict the environment and track perfectly. Could we then have perfect intensity predictions?
The challenge: globally each year.
Hurricane-scale processes control / modify intensity
Model simulation of eye / eyewall exchange (mixing): globally each year.
Hurricane Alberto (2000)
(figure adapted from Kossin and Eastin 2001, Kossin et al. 2002)
Hurricane Isabel near local sunrise on 12 Sep 2003. globally each year.
Hurricane globally each year.Isabel
(figure adapted from Kossin and Schubert 2001, 2004)
Why do we care? globally each year.
Small-scale mixing affects intensity change
(figure adapted from Kossin et al. 2006)
Primary eyewall globally each year.
Eyewall replacement cycles
Cat 3 to Cat 5 in 12 hours globally each year.
Eyewall replacement cycles usually cause rapid intensity swings. Particularly problematic as storms approach land (Hurricane Andrew 1992).
(figure adapted from Willoughby et al. 1982)
Longer-term forecasting challenges globally each year.
How are hurricanes affected by climate and climate change?
(How is the climate affected by hurricanes?)
What changes in the level of hurricane activity can we expect during the next 5, 10, 50, 100 years?
Are there cycles and / or trends?
Are the cycles / trends natural or man-made?
How do we meaningfully measure the changes? globally each year.
Accumulated Cyclone Energy (ACE)
Power Dissipation Index (PDI)
Number of Cat 4-5 storms
Systematic (i.e. not random) track changes
Direct relationship between SST and hurricane intensity globally each year.
The theory of Potential Intensity (PI, MPI) suggests that, all other things being equal, an increase of underlying SST will lead to an increase in the maximum intensity that a hurricane can achieve.
Relationship between SST and hurricane “activity” globally each year.
(figure adapted from Emanuel 2005)
Reflects changes in frequency, intensity, and duration
Changes in frequency of the most intense hurricanes globally each year.
(figure from Webster et al. 2005)
Relationship between SST and hurricane location / track globally each year.
(figure from Kossin and Vimont 2007)
The relationship with SST is part of a more general relationship with the Atlantic Meridional Mode (AMM). The AMM describes large-scale circulation patterns that go beyond the limitations of Potential Intensity theory.
Data issues… globally each year.
A Brief History of the Global Hurricane Record globally each year.
Age of the Weather Satellites
Post WW-II Aircraft
Maximum intensity rarely measured.
Storm counts may be low.
West Pacific until 1987.
Maximum intensity rarely measured.
Storm counts are better.
The counts are good.
The existing record is inconsistent by its nature and its construction.
Data reanalysis globally each year.
Variability and increases in the Atlantic verify well.
Global trends may be inflated.
(figure adapted from Kossin et al. 2007)
(from Vimont and Kossin 2007)
Changes in the long-term records markedly to warming SST than storms in other ocean basins?
Human-induced variability and increases, natural cycles, or both?
The Atlantic Multi-decadal Oscillation (AMO) markedly to warming SST than storms in other ocean basins?
Hypothesis #1: The AMO is a natural cycle related to periodic changes in the thermohaline circulation (a.k.a. the Atlantic conveyor belt). This natural cycle is superimposed on a smaller man-made trend. Under this hypothesis, SST is expected to eventually decrease to a long-term cooler regime.
Hypothesis #2: The signal known as the AMO is actually just a superposition of two human-induced signals – anthropogenic greenhouse gas warming and sulphate aerosols. Under this hypothesis, SST is expected to continue its present rate of human-induced increase with no natural cycle to help offset it.
Summary: markedly to warming SST than storms in other ocean basins?
Hurricane forecasting faces many challenges in both an operational setting and toward long-term risk assessment.
Operational intensity forecasts are challenged by the broad spectrum of scales that matter (environment to hurricane scale).
Long-term forecasting is challenged by our present lack of understanding of the relationships between hurricanes and climate change.
How will frequency, intensity, and tracks change?
Could the effects of increasing SST be offset by more frequent eyewall replacement cycles and/or mixing events?
Why is the Atlantic changing so profoundly?
Is the present high Atlantic activity just a phase of a cycle or will it continue indefinitely?
Relationship between the AMM and the 3 factors comprising hurricane activity (frequency, duration, intensity)
raw / low-pass / high-pass
Raw time series
Raw time series hurricane activity (frequency, duration, intensity)
An explanation for the variability of duration
There is a systematic shift of the mean tropical cyclogenesis region to the southeast (northwest) during positive (negative) phases of the AMM. Since storms generally track westward to northwestward, a southeast shift allows storms to last longer before reaching hostile environments (land, cold SST, high shear).