I have been testing AOD (aerosol optical depth, a measure of light extinction due to aerosols in the atmosphere) devices this week. As such, I've been comparing results with nearby AERONET sites (https://aeronet.gsfc.nasa.gov/). I was looking at the
NEON_CVALLA site and noticed this weird spike in AOD in the afternoon yesterday (October 17
th). I checked the other sites and saw spikes at the
NEON_RMNP site,
Table Mountain site, and high values at the
Digital Globe calibration site. These are all Level 1 products, so they may not pass quality checks, but I still thought it was strange and warranted further investigation.
I checked HMS, and I noticed some fires near Fort Collins, which I assumed were probably prescribed fires since they aren't there today. Zooming in, I more clearly saw that there was indeed some smoke produced from the fires.
 |
| HMS Smoke and Fire product for October 17th, 2017 over the whole western US (left) and zoomed in over northern Colorado (right). Data from: satepsanone.nesdis.noaa.gov. |
I double-checked this against MODIS. There was no smoke noted from the Terra instrument (which has a morning overpass), but there was smoke visible from the Aqua instrument (which has an afternoon overpass).
 |
| MODIS True Color Images from Terra (left) and Aqua (right) for October 17th, 2017. |
I checked if there was any impact on our surface air quality, and there was a slight increase in PM
2.5. Coarse PM (2.5-10) also increased. The timing seems earlier than what was noted by the satellites, but the satellites may have not picked it up in the morning and the smoke may have not made it to the AERONET sites until later in the day. There is also a diurnal cycle to the PM time series where concentrations are higher in the morning and night, and low during the day. This could be from the boundary layer trapping concentrations near the surface. Then, as the boundary layer grew throughout the day, concentrations decreased. It's hard to untangle all of this and directly attribute both the PM changes in Fort Collins and the AOD spikes at AERONET sites further downwind to this smoke, but it seems to be at least a plausible source.
 |
| PM2.5 and Coarse PM concentrations measured in Fort Collins, CO for October 16-18th, 2017. |
Another double-check that this was indeed a prescribed fire sent me to InciWeb (where all fire incidents are recorded), where I found that the smoke was likely from the Pingree Hill Prescribed Burn (
https://inciweb.nwcg.gov/incident/4727/) as its location generally matched the location from the satellite data. They burned 95 acres yesterday.
Was it a cool night on the 16th? That's a pretty big PM spike in the morning... a combination of biofuel and low boundary layer? Although the PM10 seems high for biofuel. Windy? DUST?!
ReplyDeleteI definitely think the boundary layer is at play, but not sure of the actual source of the spikes on the morning of the 17th. It only got down to about 35-40F overnight (slightly warmer than the past few nights), but it still could be fireplaces/wood stoves. The wind was pretty calm (https://ccc.atmos.colostate.edu/~autowx/fclwx_plot_display.php), which is why they chose to burn that day. The coarse PM could be construction dust since the monitor is on campus. Greeley Hospital site also saw a spike in PM2.5, although two hours later. So it is definitely a bit confusing. I think the peak on the evening of the 17th likely has some smoke from the fire, but the morning peak on the 17th is too early for the burn. However, I did tweet the Canyon Lakes Forest Service to find out when they started the ignition (they didn't start tweeting out pictures and info until around 11am on the 17th).
ReplyDeleteHmmm, interesting. This is where having a distributed PM network would really help I guess, to figure out how isolated the enhancement is.
Delete