Hurricane Katrina - Image Large 01 (compressed)

Unusual and devastating hurricanes in 2005

29 August 2005 00:00 UTC-5 December 00:00 UTC

Hurricane Katrina - Image Large 01 (compressed)
Hurricane Katrina - Image Large 01 (compressed)

Hurricane season in 2005 was deadly because of Katrina and unusual because of Vince.

Last Updated

14 March 2023

Published on

02 March 2023

By Gordon Bridge, Livia Briese, Marc Jenner, Jochen Kerkmann, Francois Parisot, HansPeter Roesli, Jarno Schipper (ZAMG), Cecilie Wettre


29 Nov-9 Dec, Bermuda, Azores

In many respects Hurricane Epsilon was an exceptional event: it was the 26th hurricane occurrence in the Atlantic during 2005. However, even Epsilon was not the last such event in the Atlantic, when Zeta, the 27th tropical storm; developed on 30 December 2005; but did not reach hurricane force (see Meteosat-8 RGB composite image, 3 January 2006, 07:00 UTC).

Epsilon's birth place; between Bermuda and the Azores Islands; as well as its almost exclusively eastward track have probably no counterpart; according to recent records. In addition, and bearing in mind its quite northerly position in the Atlantic, the hurricane-force period of Epsilon was quite long, lasting from 2 December 2005 18:00 UTC to 7 December 2005 06:00 UTC, with the hurricane having an open eye throughout this period.

The hurricane first appeared in an area of convective clusters, embedded in an upper-level low, that slowly organised themselves into a cyclonic structure with an eye already on 1 December 2005.

Met-8, 05 December 2005, 06:00 UTC
Figure 1: Meteosat-8 RGB Composite WV6.2–WV7.3, IR9.7–IR10.8, WV6.2, 5 December 2005 06:00 UTC. Animation (Six-hourly intervals) (29 November 00:00 UTC–9 December 00:00 UTC)
Met-8, 05 December 2005, 06:00 UTC
Figure 2: Meteosat-8 Channel 09 (IR10.8), 5 December 2005 06:00 UTC. Animation (storm centered)
(2 December 18:00 UTC–7 December 12:00 UTC, six-hourly, showing details of eye region)
Met-8, 05 December 2005, 06:00 UTC
Figure 3: Meteosat-8 Channel 05 (WV6.2), 5 December 2005 06:00 UTC. Animation (storm centered)
(2 December 18:00 UTC–7 December 12:00 UTC, six-hourly, showing upper tropospheric humidity)
Met-8, 05 December 2005, 06:00 UTC
Figure 4: Meteosat-8 Difference Image WV6.2–WV7.3, 5 December 2005 06:00 UTC
Animation (storm centered) (2 December 18:00 UTC–7 December 12:00 UTC, six-hourly, showing mid tropospheric humidity)


8-10 Oct, Maderia, Spain

On Sunday 9 October 2005 Tropical Storm Vince developed a few hundred kilometers northwest of Madeira island, and was upgraded to a hurricane later that day. Although rather small in horizontal dimension, it was classified as a category 1 hurricane with a maximum speed of 125km/h. Vince caught forecasters by surprise, since the waters in the area are some degrees colder than the 26.5ºC typically needed for a tropical storm.

Hurricane Vince gradually weakened as it moved over cooler waters, and was classified as an extratropical cyclone the next day. Vince made landfall on the Iberian peninsula near Huelva, Spain at 09:00 UTC on 11 October, and was downgraded to a tropical depression. Vince was the first tropical cyclone on record to make landfall in Spain.

Warm water (or, more specifically, the moisture in the air above it) is the energy source for tropical cyclones. A sea surface temperature above 26.5ºC to at least a depth of 50 meters is one of the conditions normally required to make tropical cyclone formation possible. Vince is unusual in that it formed over significantly cooler water.

Tropical cyclones do occasionally form even where sea temperatures do not meet the above-mentioned condition, but only when associated with specific weather disturbances. One such disturbance is an upper tropospheric trough, or a cold-core upper level low. A tropical cyclone may then develop when the upper-level disturbance produces deep convection, which becomes organised.

On the Airmass RGB image and animation (below, lower right), the narrow, upper tropospheric trough can be seen as a red contour around the 'mini' hurricane, indicating descending stratospheric air with high potential vorticity. Hurricane Vince at its peak lies exactly in the axis of this trough and it looks like it is driven by strong shear vorticity. In addition, it is pictured and surrounded by blue colours indicating cold, polar airmasses. Finally, a secondary smaller centre of vorticity can be seen running around the southern side of the trough.

Met-8, 08 October 2005, 12:00 UTC
Figure 5: Meteosat-8 RGB Composite VIS0.8, IR3.9r, IR10.8, 8 October 2005, 12:00 UTC
Met-8, 08 October 2005, 12:00 UTC
Figure 6: Meteosat-8 RGB Composite: NIR1.6, VIS0.8, VIS0.6, 8 October 2005, 12:00 UTC
Met-8, 09 October 2005, 12:00 UTC
Figure 7: Meteosat-8 RGB Composite: VIS0.8, IR3.9r, IR10.8, 9 October 2005, 12:00 UTC
Met-8, 09 October 2005, 12:00 UTC
Figure 8: Meteosat-8 RGB Composite: NIR1.6, VIS0.8, VIS0.6, 9 October 2005, 12:00 UTC
Met-8, 09 October 2005, 12:00 UTC
Figure 9: Meteosat-8 RGB Composite, HRV, IR3.9, 9 October 2005 12:00 UTC
Met-8, 09 October 2005, 12:00 UTC
Figure 10: Meteosat-8 RGB Composite WV6.2–WV7.3, IR9.7–IR10.8, WV6.2, 9 October 2005 12:00 UTC
Animation (source: ZAMG) (9 October 10:00 UTC–10 October 03:00 UTC, timestep 1 hour)

Note: The airmass RGB (lower right image) shows not only the height of the clouds, but also the airmass characteristics in cloud-free areas. The WV6.2 channel shows the Upper Tropospheric Humidity (UTH), the WV6.2–WV7.3 difference shows the vertical distribution of humidity and the IR9.7–IR10.8 difference is related to the total ozone content and the height of the tropopause.

Combined together in a RGB, the result is an image where high clouds appear in white colour, mid-level clouds in light ochre colour and cloud-frees areas in dark green colour (warm air mass with high tropopause) or blue colour (cold air mass with low tropopause). A particular feature of this RGB is that the dry descending stratospheric air is marked by a red colour.


23-29 Aug, USA

Hurricane Katrina was the 11th named storm and fifth hurricane of the 2005 Atlantic hurricane season. It was the costliest natural disaster, as well as one of the five deadliest hurricanes, in the history of the United States. Overall, at least 1,833 people died as a result of the hurricane and subsequent floods, and total property damage was estimated at $108 billion. (Source: Wikipedia )

Background and history

Hurricane Katrina was first classified as a Tropical Depression on 23 August 2005, when it was located between the Bahamas and eastern Cuba (see Track of hurricane Katrina, source: NOAA/CIMSS). Within two days it developed into a category 1 hurricane (on the Saffir-Simpson Hurricane Scale, i.e. mean wind speed between 115 and 147km/h), moving towards Florida where it made landfall on 25 August (Figure 11).

Figure 11: GOES IR, 25-29 August 2005, three-hourly

After crossing southern parts of Florida it moved into the Gulf of Mexico, first in a south-westerly, and then in a north-westerly, direction. Picking up energy from the warm waters of the Gulf, it quickly developed into a deadly category 5 hurricane (i.e. mean wind speed of more than 260km/h), which is the maximum intensity classification.

Unlike hurricane Isabel in September 2003 (also a category 5 hurricane), which degraded to a category 2 hurricane before making landfall in Northern Carolina and Virginia, hurricane Katrina maintained its full force making landfall over Grand Isle (Mississippi River Delta) at about 10:00 UTC on 29 August with winds reaching 250 km/h (category 4 hurricane).

The MODIS image from 17:00 UTC on 28 August shows the fully developed hurricane 17 hours before landfall (Figure 12). The image shows the very symmetric cloud structure of this huge hurricane and the relatively large eye, with the eyewall and the Central Dense Overcast (CDO) region.

Hurricane Katrina hits coast of USA
Figure 12: Terra MODIS True Color RGB, 28 August 2005 17:00 UTC. Source: NASA

Altimeter data

It is well known that hurricanes require warm water, and, often, a threshold value of 26–28°C is mentioned in the literature as the minimum sea surface temperature needed for hurricane formation. However, the SST is very variable (e.g. it changes rapidly due to mixing processes) and it does not appropriately represent the heat content that is stored in the upper ocean.

A much more reliable dataset for the forecast of the intensity of hurricanes is altimeter data (as provided by Jason-1 and Topex/Poseidon, and, in future, by Jason-2, see Figures 13-16) .

Altimeter instruments measure the height of the ocean surface, or the height anomaly as compared to a reference height (Figure 13). The ocean height is mainly related to its internal thermal structure, i.e. it represents a vertically integrated measurement of the density, and, thus, the temperature of the ocean: the warmer the ocean, the higher the ocean surface.

Therefore, while sea surface temperatures (SST) derived from instruments such as AVHRR represent the temperature of the top (skin) layer of the ocean, from altimeter measurements one can derive information about the ocean temperature at deeper layers. An example is given below (Figure 14). It shows the depth (in metres) of the 26°C isotherm, which is a critical parameter for the forecast of hurricane formation/intensity, since it is much closer related to the heat stored in the upper layers of the ocean. The rule of thumb based on this parameter is: hurricane formation is possible when the sea temperature is above 26.5°C to at least a depth of 50m.

In the case of hurricane Katrina, a warm anticyclonic ring with an average depth of the 26°C isotherm of 90m was observed in the central part of the Gulf of Mexico (Figure 14). When hurricane Katrina reached this warm ring it rapidly intensified from a category 3 to a category 5 hurricane (as indicated by the coloured circles that mark the track of the hurricane). The ring of warm water is not visible in the SST field (Figure 15), because it is masked by the very thin, warm and stable upper layer formed in the Gulf of Mexico (with sea surface temperatures up to 32°C). This example demonstrates that altimeter-derived fields such as the depth of the 26°C isotherm or the Tropical Cyclone Heat Potential (not shown here) are essential for the forecast of the intensity of hurricanes. For further details, read GOM Surface Dynamics Reports from 29 August 2005, from G. Goni.

Satellite-derived products

Satellite(altimetry)-derived field of
Figure 13: Satellite(altimetry)-derived field of sea height anomaly (SHA), 28 August 2005.
Satellite(altimetry)-derived field of
Figure 14: Satellite(altimetry)-derived field of estimated depth of the 26°C isotherm, 28 August 2005.
Source: NOAA
Satellite-derived field of
Figure 15: Satellite-derived field of sea surface temperature (SST), 28 August 2005. Source: NOAA
Satellite(altimetry)-derived field of
Figure 16: Satellite(altimetry)-derived field of surface currents, 28 August 2005. Source: NOAA

Note: The Sea Height Anomaly (SHA) field used is from the Naval Research Laboratory (NRL) based on Jason-1, GFO and Envisat. The SST field comes from TMI data and is produced by Remote Sensing Systems, which are sponsored by the NASA Earth Science REASoN (Research, Education and Applications Solution Network)) DISCOVER Project. The data is available at

Media reports

Tropical Cyclone Report Hurricane Epsilon (James L. Franklin, 2005)
Report on NASA Earth Observatory Newsroom web page
Report on NASA Earth Observatory Natural Hazards web page

Other image source

Colour-enhanced GOES IR image (29 Aug. 2005, 12:45 UTC, source: NOAA/CIMSS)