This weekend, an East Coast Low (ECL, also known as an East Coast Cyclone) has been ravaging the east coast of Australia, causing very heavy rain, flooding, strong winds, rough seas, and coastal erosion. These sorts of storms are not uncommon – there were at least four East Coast Lows that caused major impacts last year, including a very severe ECL that hit Sydney and the Hunter in mid-April, while June 2007 had five East Coast Lows in a single month.
However, when people start asking questions like, “How often do these storms happen?” or “How does La Niña influence East Coast Lows?”, the answers aren’t quite as simple as you might think. Quite simply, there is still a lot of uncertainty on what actually makes an East Coast Low.
However, when people start asking questions like, “How often do these storms happen?” or “How does La Niña influence East Coast Lows?”, the answers aren’t quite as simple as you might think. Quite simply, there is still a lot of uncertainty on what actually makes an East Coast Low.
“I know it when I see it”
There are some East Coast Lows that everyone can agree on, like the Pasha Bulker storm of 2007 or the Sygna storm of 1974. These storms tend to have the following characteristics:
But when trying to put together a database of events for science and analysis, it starts to get trickier, and more questions arise.
Where do you draw the line?
Is it still an East Coast Low if there is a surface cyclone, but there isn’t any severe weather?
The most widely cited database of cyclones, by Speer et al. (2009), defined cyclone based on how they looked on a weather map. Using that approach, there are about 22 ECLs a year, but only 7-8 have widespread rain and 2-3 have “explosive” development. In contrast, work by Jeff Callaghan and Scott Power started by looking at storms with big impacts on the coast, and then tried to see what systems caused them. Both these approaches have value, but they give you very different databases.
Is the surface low the most important part?
We know that how strong the low pressure system looks on the weather map isn’t really a good indicator of how big its impacts are – if you look at all the maps for all ECLs in June 2007, you wouldn’t necessarily guess that the most impactful one was 7-8 June. So some studies just look at the upper level circulation instead of the surface low, as that might be a better guide to impacts.
How can a computer best identify cyclones?
Looking through weather maps is hard and time-consuming, so we need a better way to get databases of cyclones. There are a lot of different automated methods these days for identifying cyclones from gridded pressure data. But they all identify slightly different systems, and they also depend on which pressure data you use. This can mean some quite large differences between automated databases in identifying smaller or less severe systems, and mean you get very different correlations between ECL frequency and things like El Niño and La Niña. Annoying.
How close to the coast does an ECL need to be?
The Speer et al. (2009) database looked at cyclones anywhere from the coast to 160° E, but the most impactful cyclones occur much closer to the coast. On the other hand, a very large cyclone could have impacts further from the low centre, and a very restrictive definition might mean important cyclones get missed. Where do you draw the line? Current definitions also tend to look at cyclones between Gippsland and Fraser Island – but what about cyclones in the Bass Strait, such as during the infamous Sydney-Hobart race of 1998, or near Tasmania?
What about the “type” of ECL?
East Coast Low is a funny category, and often lumps together very different types of systems – big extratropical cyclones that come from the westerlies to the south of Australia, ex-Tropical Cyclones that come from the north, cyclones that develop in a trough right along the east coast. All of these systems can have quite different characteristics and impacts, as well as different relationships with things like La Niña. But there are a lot of different classifications used by different groups, and it can be difficult to apply them in an automated way, so lazy people like me often lump all the storms together. This makes analysis easier, but may lose some of the fine detail we need to understand cyclones properly.
Ultimately, it’s all a bit subjective, and every study uses a slightly different definition. So, if you ask me “How often do storms like this happen?” or “Does La Niña mean we’ll get more storms like this this year?”, at the moment my answer is “Well, it really depends on what you mean by East Coast Low…”
There are some East Coast Lows that everyone can agree on, like the Pasha Bulker storm of 2007 or the Sygna storm of 1974. These storms tend to have the following characteristics:
- A closed low pressure system on the surface
- Forms or intensifies directly near the east coast (this can be very rapid, a so-called “explosive” cyclone or “bomb”)
- Has some component of movement parallel to the coast
- Very severe weather, particularly on the south side of the low
- A “cold pool” or low in the upper atmosphere
But when trying to put together a database of events for science and analysis, it starts to get trickier, and more questions arise.
Where do you draw the line?
Is it still an East Coast Low if there is a surface cyclone, but there isn’t any severe weather?
The most widely cited database of cyclones, by Speer et al. (2009), defined cyclone based on how they looked on a weather map. Using that approach, there are about 22 ECLs a year, but only 7-8 have widespread rain and 2-3 have “explosive” development. In contrast, work by Jeff Callaghan and Scott Power started by looking at storms with big impacts on the coast, and then tried to see what systems caused them. Both these approaches have value, but they give you very different databases.
Is the surface low the most important part?
We know that how strong the low pressure system looks on the weather map isn’t really a good indicator of how big its impacts are – if you look at all the maps for all ECLs in June 2007, you wouldn’t necessarily guess that the most impactful one was 7-8 June. So some studies just look at the upper level circulation instead of the surface low, as that might be a better guide to impacts.
How can a computer best identify cyclones?
Looking through weather maps is hard and time-consuming, so we need a better way to get databases of cyclones. There are a lot of different automated methods these days for identifying cyclones from gridded pressure data. But they all identify slightly different systems, and they also depend on which pressure data you use. This can mean some quite large differences between automated databases in identifying smaller or less severe systems, and mean you get very different correlations between ECL frequency and things like El Niño and La Niña. Annoying.
How close to the coast does an ECL need to be?
The Speer et al. (2009) database looked at cyclones anywhere from the coast to 160° E, but the most impactful cyclones occur much closer to the coast. On the other hand, a very large cyclone could have impacts further from the low centre, and a very restrictive definition might mean important cyclones get missed. Where do you draw the line? Current definitions also tend to look at cyclones between Gippsland and Fraser Island – but what about cyclones in the Bass Strait, such as during the infamous Sydney-Hobart race of 1998, or near Tasmania?
What about the “type” of ECL?
East Coast Low is a funny category, and often lumps together very different types of systems – big extratropical cyclones that come from the westerlies to the south of Australia, ex-Tropical Cyclones that come from the north, cyclones that develop in a trough right along the east coast. All of these systems can have quite different characteristics and impacts, as well as different relationships with things like La Niña. But there are a lot of different classifications used by different groups, and it can be difficult to apply them in an automated way, so lazy people like me often lump all the storms together. This makes analysis easier, but may lose some of the fine detail we need to understand cyclones properly.
Ultimately, it’s all a bit subjective, and every study uses a slightly different definition. So, if you ask me “How often do storms like this happen?” or “Does La Niña mean we’ll get more storms like this this year?”, at the moment my answer is “Well, it really depends on what you mean by East Coast Low…”