15 Nov – Tropical cold eddy


SST Anomaly for Australia – RAMSSA

There has been no significant change in the surface anomaly pattern this past week, with weak cool anomalies persisting in the west and south, while the east waters are neutral to warm.

SST Anomaly for northwest WA – RAMSSA

The inshore waters of northern WA show signs of recent heating, with positive SST anomalies developing along the Broome coast. However, west of the NW Cape it remains cooler than normal.

Tropical cold core eddy

Sea level anomaly for far northern WA waters – OceanMAPS

In our far north west there is a negative sea level anomaly which is acting as a cold core eddy, with the word ‘cold’ here being a relative word. Sea surface temperatures are 28-29 deg C.

Try to spot the cold core eddy in the SST image below. Where is it?

Spot the cold core eddy, where is it in this SST picture? – OceanMAPS

The eddy is only seen at depth.

Temperature profiles at three location with the blue dot indicating the cold core eddy – OceanMAPS – click to expand

An examination of three temperature profiles in adjacent areas shows that the cold core eddy (blue dot and line) is significantly colder than surrounding water between 34 m and 233 m depth.

The peak difference is at 85 m depth. At this depth we can look at the temperature percentiles in the image below.

5th percentile exceedence at 85 m depth – OceanMAPS – click to expand

We can see that at 60-90 m depth the temperatures are up to 6 degrees cooler than the 5th percentile! And 4 degrees cooler than the 1st percentile (not shown). This is a significant deviation.

Salinity profile with background view of values at 85 m – OceanMAPS – click to expand

Salinity data shows that the cold core eddy is also a region of higher salinity. The cold, salty water is a lot more dense than the surrounding areas. Is this water sinking?

Why has this cold core eddy formed? It probably contains some of upwelled cooler water that was along the south coast of Java through the winter.  We had a look at the Indonesian Throughflow back in September when it was at its seasonal maximum. Recap on the blog post here.

The Java upwelling season has now finished because the southeast monsoon has eased and the monsoon trough is moving southward. Later this week some W to NW winds develop along this coast and we may actually see some downwelling.

However, the old upwelled water is still around. It looks like this water has been entrained into the eddy and is slowly sinking as it moves south, with warmer water filling in over the top.

Series of temperature slices at 48 m depth for 1st Oct, 21st Oct, 7th Nov and 15th Nov – OceanMAPS

The much cooler water in this cold core eddy would mean a much lower tropical cyclone heat potential (TCHP). This is a depth-integrated measure of heat down to the 26 deg C isotherm.

There are some potential TCs to the immediate west of this area at the moment. For more details, see the BOM TC Outlook. If these lows move eastward over the cold core eddy (unlikely but possible) then TC development will be affected by a reduction in heat availability.

Gradient level wind analysis with arrow pointing to two tropical lows to the far northwest of Australia, and a hand-drawn blue dot representing the location of the cold core eddy – BOM

A rough TCHP calculation from the 15th of November OceanMAPS analysis is shown below. See how the TCHP values are very low in the vicinity of the cold eddy, but increase to the east.

Tropical Cyclone Heat Potential for 15th Nov 2018 at 12z – OceanMAPS

A sunken pool of cold water in ‘TC alley’ will be of interest to watch over the coming months.


The BOM wave spectra plots were updated a fortnight ago. Those with access to the BOM intranet can find the plots here.

The improvements made are:

  • Removed blank sector (usually white sector to north) from the background
  • Added figure legend and fine tuned the contour intervals
  • Added direction lines and labels e.g N, NE, NW etc
  • Modified the period labels
  • Changed the normalization design and conduct normalization in percentage to the maximum value for each site over all forecast lead time.  The original plots show normalisation for all sites and all forecast lead time. This change would not allow simple comparison of wave spectra for different sites though; however you can still see the evolution of wave energy built over the forecast lead time for each location
  • Removed the minimum wind limit to only display the wind vector in every figure.
  • Use a fixed length and big arrow for all wind vectors as the wind speed is listed under the graph.

case study

By way of an example, let’s look at the Brisbane wave buoy location at 27.5 S 153.6 E. Today, a swell of 9-10 seconds from the east is affecting the location. The energy is very concentrated in direction. Note that the spectra are in oceanographic convention.

The blue arrow in the centre shows the local wind direction. The wind is coming from the northeast, but the wave energy is coming from the east-southeast.

Wave spectra for Brisbane buoy location at 12Z on 15th Nov – Auswave-R

This swell is being generated north of New Zealand, where easterly winds have persisted on the north side of a strong high pressure system.

MSLP analysis chart for 13th Nov, arrow showing wind direction – BOM

On Saturday a southerly change is expected along the southeastern coast of Australia. In the wave spectra we can see a new peak of wave energy developing at 8-9 seconds travelling northward. This is the swell of the southerly change reaching the location before the wind has changed. At this timestamp the wind is a very light 2 knots from the north.

Wave spectra forecast for Brisbane buoy location at 00Z on 17th Nov – Auswave-R

Fast forward 21 hours and the wind will have changed direction to the south to south-east and picked up to 22 knots. The wave peak from the change has now intensified and spread in direction. The period has also reduced to 6 seconds which indicates it is mostly wind waves. The easterly swell continues, but may no longer be the dominant wave energy, or if it is, only just.

Wave spectra forecast for Brisbane buoy location at 21Z on 17th Nov – Auswave-R

How does this look on a 2D map?

Answer: a mess.

Total wave height for 21z on 17th Nov – Auswave-R

What about the buoy measurements?

Answer: They simply can’t tell the full story.

Brisbane wave buoy measurements 14-16th Nov – Queensland Government

Wave spectra plots are powerful tools to understanding a full sea state.

Do you have any feedback about the new plots?



One thought on “15 Nov – Tropical cold eddy”

  1. Plots nice.
    To tackle next may be :
    – verification
    – use of spectral in-situ observations
    – visualisation of satellite obs

Leave a Reply

Your email address will not be published. Required fields are marked *