Friday 9 July 2010

Update from Dalhousie....

Post written by Eddy Barratt and Rob Trigwell

Well we’ve had a little practise and I (Eddy) had my first shift running the lidar solo on Monday night, though a combination of instant messaging, e-mails and phone calls meant that I had all the support that I needed. You might expect that the lidar can pretty much run itself, but there are still quite a lot of tasks that kept me busy. The most important of these is to do with aircraft safety.  We have a radar that shuts down the laser whenever an aircraft is in our vicinity. When this happens I have to go outside, look for the aircraft in question, then log it’s height and it’s direction before turning the laser back on. There are also helicopters that occasionally buzz over the city.  We have a microphone on the roof and we can hear them approaching from a fair distance, and we can shut off the lidar manually before the radar even detects them.

 The LIDAR running at night.

The data retrieval software also needs to be started every 60 minutes, and to produce the plots requires a little skill and patience. If you look carefully at the plot below (taken monday night, current plots can be found at http://aolab.phys.dal.ca/data/current/) you’ll see black bands at 5 and 5.5 km. These are where I’d set the normalisation limits. The software has to be told how much light to expect from clear, aerosol free air, and we do that by selecting a region and telling it that in that region the air is free of aerosols. Clearly if their are aerosols in that region the plots will look wrong, so selecting the right normalisation region is important. Usually you’d choose a region high up in the atmosphere but for that shift I was having to continually lower the region, to bring it under the cloud which had been steadily sinking all evening.

This plot demonstrates some interesting features. First, and most obviously, you can see the bottom edge of the cloud. It had fallen from about 9km at 2100 (GMT, 6pm local time) to 6km when I took the screenshot at 0230.. Occasionally, like between 2200 and 2300, the cloud was thin enough so that we were still getting some readings from above it, but for most of the evening the bottom of the cloud had been as far as our laser has reached. The second thing is the band of bright colours down low, in the first 700m or so. This is thin cloud or fog, it wasn’t thick enough to stop the beam, but it did provide enough reflections to show up quite significantly.

The third thing you might notice, with a keener eye, is the swirling blue tinted patterns between the two cloud layers. These are aerosols, which is exactly what we’re out here to look for. I’ll hand over to Rob who ran some further analysis, to try and find out where that aerosol came from...
Lidar profile above Halifax, and 5 days backward trajectory for air that was at 2000 m elevation at 2200 h on monday night, produced by the NOAA HYSPLIT model. Click on the figure to enlarge it.

Running the NOAA HYSPLIT model at the altitude at which the aerosols were being detected by lidar can give us an insight into source regions of these aerosols. The model uses what it knows about wind strengths and directions in the past few days to calculate what paths a given parcel of air has taken to be where it is now. As you can see from the plot, the back trajectories suggest that the aerosols could have originated somewhere in the US midwest, before making their way across Ontario and Quebec. These are not heavily industrialised regions so the aerosols are probably not anthropogenic. The actual source and type of these aerosols is a matter for further analysis. 

Back in Edinburgh, Mark Parrington is using forecast output from the NASA GEOS-5 system to predict the trajectories of biomass burning plumes, and when they pass over Halifax. The forecast CO profiles over Halifax (http://xweb.geos.ed.ac.uk/~mparring/BORTAS/Halifax/), plotted in a similar form to our lidar plots, highlight where Carbon Monoxide is strongly concentrated with the measured aerosol. CO is produced in a number of different processes, both industrial and natural, but in particular CO would be expected to coexist with aerosols from Biomass burning, so correlations between his plots and our own are very exciting.

CO forecast profiles over Halifax, plots courtesy of Mark Parrington, University of Edinburgh. Click to enlarge.

A last word from Eddy:
I should have mentioned in my last post that I’m here thanks to a grant awarded by the Nuffield Foundation, which I’m very grateful for, and very sorry not to acknowledge it last post.

1 comment:

  1. Smoke is over Lake Huron at the moment.

    Check out the Smog Blog http://alg.umbc.edu/usaq/ later this evening (Saturday July 11)

    ReplyDelete