Special Investigation: Something's Up Down Under (pt.3: Local Effects)

In my previous posts, I have investigated tropical cyclones that last a remarkably long time over Australia by using a sample of eighteen such storms over the past ten years. First I arranged the storms by the month of the year they occurred in, and found that there is a tendency for these storms to occur near the ends of the tropical cyclone season. However, for several reasons, this correlation was not considered very strong. Next, I compared the timing of the sample storms to the state of the El Nino/Southern Oscillation (ENSO). I found a rather strong correlation between the storms and La Nina conditions, and a weak correlation to El Nino conditions. This data was quite clear and straight forward, so I considered this to be more scientifically sound than the time of year analysis. Then I continued my investigation by looking into the Madden-Julian Oscillation (MJO), which has a profound impact on Australia. I found that the sample storms occurred most often when the active phase of the MJO was weak and when it was located over central Australia or slightly east of Australia. I found the latter correlation to be the stronger of the two, since it closely corroborated with a small article I found on the MJO. In this final post, I will look at orographic (terrain induced) and meteorological factors that may play a part in maintaining these tropical cyclones over land for such a long time.



Local Effects

For this final section of the investigation I utilized storm tracks, topographical data, and 30 year averages of December through March from the 20th Century Reanalysis dataset.

I began with the topography of Australia, as that had been my hypothesis years ago when I first discovered this trend. I plotted most of the storm tracks in the sample (not all of the tracks were available) over a topographical map of Australia. The map's color scheme was exaggerated to help terrain features stand out. When looking at the tracks, it appears that only the raised terrain in the far west of Australia could possibly have any real effects. Since the dominate wind direction in southern Australia is from the west, air will have to rise some to traverse the mild terrain. This rising air could help fuel the convection that maintains storms that pass roughly parallel to the terrain. The airflow of the storms that pass right over the terrain would be forced to rise, and thus support the storm. This is vaguely similar to what happens on the great plains of North America where the gentle slope towards the Rocky Mountains helps fuel the severe thunderstorms common to the region.

One factor that ended up being irrelevant was the moisture content of the soil. The idea was that very moist soil, such as a swamp, could help maintain a tropical cyclone over land. As you can see in map below, there does not appear to be any correlation between soil moisture and the tracks of the storms, in fact, several long lived storms tracked right over some of the driest soil in Australia. Therefore, this factor is dismissed.

The next parameter I looked at was vorticity (specifically relative vorticity). vorticity is calculated from the components of wind and is essentially the amount of "spin" in the air at any given location. Thus, air with high cyclonic vorticity will help fuel a storm. Below is a three dimensional map of the regions with reasonably high amounts of cyclonic vorticity in an average of 30 Januarys. Notice that nearly all of the vorticity of this strength in Australia occurs directly over the bulk of the storm tracks. This is certainly an useful factor to consider.

Where there is upward motion, there are often convective clouds, therefore I looked at omega (ω). Omega is the pressure vertical velocity, since pressure decreases with height, the more negative omega is, the faster it is rising. The animation below shows areas of decent omega values; upward motion is in red and downward motion is blue. Typically, the greatest values will be greatest at mid-levels and lowest at the surface and at the tropopause. Below that is a two-dimensional map of omega values at mid-levels (850 hPa). Note the strong upward motion just south of the terrain feature discussed earlier. This could very likely be due to the air rising up the terrain.

In conjunction with omega is an investigation on horizontal convergence and divergence. When air converges at the surface, the only direction available for it move is up, creating negative omega. The opposite is true with divergence in which air spreading out causes air to descend to take its place. For a tropical cyclone to sustain itself, it must have strong convergence at the surface to force air to rise, and divergence at its top to help pull the air up and exhaust the "spent" air away from the storm. The animation below shows 3-D surfaces of convergence (red) and divergence (blue). Notice how much of the storm tracks lie in regions of surface convergence, and some areas even have higher level divergence. Vorticity, omega, and convergence all have high values in the same general region, lending support to the hypothesis that there is something unique about this region.

The final factor to consider are the horizontal winds and how they are related to the environmental conditions. First of all, it is important to understand the sea surface temperatures (SST) around Australia. During the tropical cyclone season (December through March in this case) the average SSTs are highest in the Gulf of Carpentaria and around Darwin. Also important that there is a sharp temperature gradient along the western coast of Australia, compared to a much more gradual gradient on the east coast. Using streamlines of the average low level flow (950 hPa), one can see that most of the air flowing into the region containing the sample storm tracks comes from over very warm ocean water, especially in the Gulf, and is likely quite moist.

What causes this wind pattern? A look at the sea level pressure shows a region of low pressure. As one would expect with any low, air flows around it cyclonically and slightly inward. The 950 hPa streamlines follow this rule very well.

As far as the moisture that this airflow might be bringing in is concerned, I made a couple of cross-sections to get a profile on relative humidity. In general, one of the largest contributing factors to the weakening of tropical cyclones is dry air entrainment at mid-levels, which is usually considered to be around the 700 hPa level. That being said, I was rather amazed that a limb of moist air, at mid-levels, extended far inland, even over very dry surface air. This was perhaps the best piece of data I had found to explain the long lived storms. Even though they were over the bone dry desert, these storms were pulling in moist air from afar.

 

Conclusion

With the investigation done, I have a clearer picture of what causes these long lived Australian tropical cyclones, although I feel I do not have enough insight to make a truly solid statement. Keeping this in mind, I have come up with the likely conditions for these storms. First of all, they seem to occur almost exclusively during the height of the tropical cyclone season and are more likely to occur under La Nina conditions. As far as the Madden-Julian Oscillation (MJO) is concerned, these storms tend to occur when the MJO's active phase is over Australia or a little east of it, but a strong active phase seems to work against the storms, possibly inhibiting something. Australia itself might contribute to these tropical cyclones due to subtle topographical features that might enhance upward motion. Upward motion in general occurs over a large portion of the tracks of these storms. Coinciding with this is a large area of low level convergence. The existence of higher level divergence is exactly what would help maintain a tropical cyclone. Finally, a large area of low pressure over northwestern Australia seems to induce an airflow pattern that brings warm, moist air far inland. Much of this moisture is found at mid-levels, where tropical cyclones are often the most vulnerable to dry air entrainment.
If I were to do this study again, I would want to use a much larger sample size of at least thirty years. I would also develop a more objective way of classifying tropical cyclones as fitting this long lived storm scenario. Throughout the study I would use more numbers and less estimations based on diagrams and figures. Ideally, I would like the climate data to be higher resolution, that would especially improve the accuracy of streamlines and cross-sections. Finally, I would want to pursue a potentially crucial fact: these storms are the only tropical cyclones in the world that form near the west coast of a continent. This means that all of them that get caught in the westerlies track over land, versus over open ocean, as seen in the diagram below depicting the tracks of all recorded tropical cyclones. These factors open up many possibilities that have not been covered in this study.

As one final piece to this investigation, I have put together an animation loop of satellite images of 06U in January, 2014, the tropical low that first inspired this post. The black box frames the system. The track data does not cover the storm's entire life so at the beginning and end the box will not move even though the system is. This storm spent over eleven days over land.