Hurricane Katrina, in August 2005, was one of the most devastating hurricanes in the US history.

Data used

We will take advantage of the fact that several satellites were providing data at the time (August 2005), and use the GDRs from Envisat, Jason-1, Topex/Poseidon and GFO.


We will use the Broadview Radar Altimetry Toolbox to have a look at significant wave heights and wind speed modulus as provided in the standard datasets.

Geographic extraction

We will be looking at individual ground tracks (one for each satellite), between 17°N and 30°N

Data selection and BRAT Datasets definition

To select the right tracks, you can have a look at the pass locator using Google Earth available through Aviso.


fig 1. Katrina at different dates with T/P (red) and Jason-1 (blue) passes, Envisat (and ERS-2), and GFO passes. The images were taken within 20 minutes of the altimeter passes.(Credits NOAA/Altimetrics LLC)

The tracks and cycles closest to hurrican Katrina are:
–   Envisat, cycle 040, pass #351
–   Jason-1, cycle 134, pass #026
–   Topex/Poseidon, cycle 477, pass #026
–   GFO, cycle 157, pass #409

Name the dedicated BRAT workspace you are using for this job. Within this workspace, name your datasets. We will need 4 datasets: one for each satellite, with one file in each of them.

BRAT Operations definition

In the ‘Operations’ tab, name your operation (you will need 4 operations, too, one for each satellite), then select your dataset and data record. In ‘Data mode’ keep Y=F(X) for a plot.
Enter your data expression:
we will define two data fields, one for Significant Wave Height in Ku-band, one for Wind speed modulus.

For the X field the variable is the latitude (beware that default is longitude, so you have to change it, and change the data type).

The only thing we will put in ‘Selection criteria’ is the ocean flag. “Classical” data editing (the one provided as pre-defined formula within BRAT) is not in order here, since we are dealing with a phenomena out of the ordinary (and thus over usual thresholds). Optionally, you can constrain the number of values used to compute a 1-Hz value (depending on the satellite; see the ocean data editing for ideas of significant values) and the wet tropospheric correction (between -1 and 1 m).


Once the different operations are executed, you can plot them in the “Views” tab. Define minimum and maximum X for the plot (between 17 and 30°N); you can also use maximum Y at 15 m for SWH or 30 m/s for wind speed (minimum being 0). You can define two views : one for SWH, one for wind speed. Resulting plots should look like the ones below:


fig 2. First three plots are SWH, second three are wind speed. T/P and Jason-1 are together on the first (T/P in green, Jason-1 in blue-grey), Envisat on the second (red) and GFO is on the third (blue).


  • Topex and Jason-1 recorded significant wave heights up to 5 metres, and wind speeds around 20 m/s.
  • Envisat crossed Katrina very near the eye of the storm and measured wave heights well beyond 10 metres and wind speeds around 30 m/s (wind speeds higher than 20 m/s cannot be measured reliably; the algorithm used at that time in Envisat processing is making the plateau in the wind speed line).
  • GFO caught up with Katrina shortly before making land fall. Wave heights and wind speeds sustained at 10 metres and near 30 m/s, respectively, with higher values measured near the coast.
  • Wind speeds and wave height distributions appear asymmetric, with higher values measured windward of the eye.

See Fig. 3 in the Application about Cyclones, hurricanes and typhoons
Scharroo, R., W. H. F. Smith, and J. L. Lillibridge, The impact of dynamic topography on the intensification of hurricanes, 15 years of progress in radar altimetry Symposium, Venice, Italy, 2006