Part 2: Madmen and mad data downloads

Part 2: Jesse Greenslade continues his summer student blog. You can read Part 1 here.

In the past two weeks I’ve joined the library and the PCYC and watched episodes 1 through 8(out of 12) of Mad Men season 3, 4, and 5. I find the missed episodes of Mad Men add much flavour and mystery to the show.

Thankfully I ran into the one person I know in Hobart and got invited somewhere with a radio to listen to the hottest 100, which made Australia day much more fun than it would have been had I sat silently in my disconnected unit watching the clouds pass by.

In the meantime, my supervisor suggested using a fourier bandpass filter on the ozone columns in order to characterise STTs (Stratosphere to Troposphere ozone transport events) and given suitable boundaries it has proven very robust. The image below shows the effect of the bandpass filter, which is meant to remove high and low frequency ozone perturbations.


So now I’m using this filter to automatically screen the ozonesondes datasets and pick out just the STTs. One goal is to analyse/determine the regional weather (data from ECMWF ERA-Interim) which causes these events. However downloading ERA data from the website is like pulling out teeth. It is only possible to download a couple of months worth of data at once, and there are 30 or so checkboxes to tick that each cause the website to ‘verify’ something for a few seconds. Once every 100 clicks or so the page will crash and you need to restart the most recent clickfest.

I spent three hours downloading 2004-2011 weather data and have attached here an image of the weather and wind over Melbourne during an event on the 3rd of Feb, 2005.


There is a big low pressure system right on top of Melbourne and the winds are circling clockwise.

UNSW logo ANU logo Monash logo UMelb logo UTAS logo