Sunday, 1 August 2010

Paris and Plume Measurements

We’ve had a busy couple of weeks in Halifax, and now the BORTAS campaign for this year is in full swing. As well as the lidar we have a large array of other scientific measurement equipment set up here on the roof of the physics building at Dalhousie University, including a wind monitoring radar and instruments that measure particle sizes and ozone levels directly. I don’t have much to do with those instruments but we were all kept busy last week preparing and learning to operate the ‘Portable Atmospheric Research Interferometric Spectrometer for the Infrared’. PARIS, as she is known, is an instrument which measures atmospheric trace gases and should give us a better idea of what the chemical constituents of smoke plumes actually are. We also had to build a shelter for PARIS, a small shed that’s been named FRANCE. I’m not sure what that stands for, suggestions on a postcard, or by comment, please...

Portable Atmospheric Research Interferometric Spectrometer for the Infrared (PARIS-IR), operating on the roof of the physics building.

On the boreal biomass burning side of things: over the past week we’ve been getting a reasonably regular procession of biomass burning plumes passing overhead, as predicted by the GEOS-5 CO profile forecasts that Mark Parrington has been producing, and we’ve managed to measure the aerosols that we’d expected to be associated with that type of plume using the lidar. For example here’s the forecast and the lidar plot from measurements which took place on Wednesday and Thursday.

The CO from boreal biomass burning forecast profile over Halifax, showing high concentrations of CO at around 2km from 20 to 40 hours after the plot began (which was 1200 UTC on the 20th). This time corresponded to 0800 to 0400 UTC, and the lidar measurements from that time clearly show a similar shaped aerosol feature at roughly the same height and time. CO plot courtesy of Mark Parrington, University of Edinburgh.

You’ll notice that compared to the plot from the 9th July post, the aerosol layer here is much thicker, higher and more concentrated than the meagre one from that day. You’ll also notice how nicely it ties up with the CO forecast above. The smoke plume from a forest fire would be expected to carry high concentrations of carbon monoxide and high loads of aerosols, so it’s nice to see experimental data that seems to confirm this, and which also suggests that the CO forecasts are accurate.

We find out the locations of forest fires throughout North America from a number of sources. In particular Natural Resources Canada produce a daily plot showing where high concentrations of forest fires are occurring within Canada (, whilst FIRMS (Fire Information for Resource Management System,, of the University of Maryland, provide similar forest fire position plots for all of Northern North America as Google Earth files. Both services use satellite infrared measurements to find the fires.  As you will see from the plots below, we shouldn’t be running out of smoke plumes soon, we just hope that the wind brings them our way!

The Natural Resources Canada fire hotspots map for 22nd July , and the FIRMS google earth plot of fire locations during the 48 hours previous to the 22nd of July.

The google earth image above has been overlaid with a 72 hour Hysplit back trajectory matrix. The trajectories all end in a grid cantered on Halifax, with height and time relating to the aerosol layer discussed above. The trajectories seem to suggest that Halifax was at the focus of a region of converging air at that time, so the smoke plume could have come from anywhere in Northern and Western Canada, though it is tempting to assume that the fires visible to the west of Hudson Bay (which had also been burning for several days beforehand) were probably the source.

 Finally, if you take another look at the CO forecast plot above, you’ll notice an enormous concentration of CO due above Halifax around 60 hours after the plot began. Assuming the forecast is correct that plume should be overhead right about now, and that’s the reason why I’m sitting in the lab writing this blog at 02:30 on a Friday morning, hoping and waiting for the clouds to clear so that I can get the lidar fired up and start making some measurements!