Arizona Wildfires and Trajectory Models

Large-scale ridging continues across the central and eastern US through the middle of this week, with some storms firing around the northern (and Gulf Coast) periphery of this sprawling surface high.  With clear skies and general subsidence, near record-high temperatures are being reported across much of the central part of the country.

But, that's not what I want to talk about tonight.  When sprawling high pressure backs up against mountain ranges, the potential exists for some strong cross-barrier pressure gradients to develop.  This in turn can lead to strong winds which is bad news for people fighting wildfires.  Here's a look at the National Weather Service warning map for this evening.  Note how much of eastern Arizona, New Mexico, and the panhandles region are under red flag warnings (the pink colors).  Strong winds combined with dry air and vegetation with no rain in sight are making ideal conditions for wildfires.

Fig 1 -- NWS watches, warnings and advisories as if 0058Z, June 8, 2011.

As you've probably heard on the news by now, large wildfires are raging across eastern Arizona.  We can see the huge smoke plumes billowing from these fires on visible satellite imagery.

Fig 2 --  GOES-W visible satellite image from 00Z, June 8, 2011.  From the HOOT website.

It's impressive just how large of an area those smoke plumes are covering.  You can see that they blow off in a northeasterly direction, giving a general indication of the wind direction over that area.  Also on this visible satellite image we can see how the subsidence across the plains and the midwest has inhibited cloud formation, leaving clear skies.  Some showers and storms were occurring over Montana, and you can see the "marine push" of low clouds off of the Pacific Ocean that has moved into western Washington today.  There's a lot going on for a "quiet" weather day.

Returning to the wildfires, you can also see the smoke plumes showing up in the infrared satellite imagery.

Fig 3 -- GOES-W IR satellite image from 00Z, June 8, 2011.  From the HOOT website.

Infrared satellite imagery is usually shown in terms of a temperature scale, with the red, yellows and greens being colder and the blues and purples being warmer (I know...it's kind of reverse from how you think it should be...).  Notice how the smoke plumes show up as bright blues, which (according to the color scale on the left) are colder than the purples and dark blues that surround them.  Those purples and dark blues are showing roughly the surface temperature--with no other clouds above, the surface is the significant contributor to the outgoing infrared radiation measured by the satellite.  What's fascinating, then, is that the smoke plumes are cooler than the temperature at the surface.  Even though the source of the smoke plumes are fires that are much, much hotter than the ambient surface temperature, like pretty much everything else in the atmosphere the smoke cools as it rises.  The fact that we're seeing returns from the smoke plume that are cooler than the surface temperatures mean that the smoke has risen very far up into the atmosphere--and cooled down a lot.

So we see that the smoke is generally blowing off to the northeast.  Where will this smoke go?  We can use the wind information from our numerical weather models to run what is called a trajectory model to determine the path the smoke will most likely take over the next 48 hours (at least, according to our weather models).  One nice model that is commonly used for doing this is the NOAA HYSPLIT model.  It's very easy to use, and in fact it has an online interface here that allows you to enter in information to their form to do your very own trajectory model runs.  It's pretty cool.

Anyhow, I started a trajectory model at the latitude and longitude of a point in the middle of where all the fires were around noon CDT today.  I started air parcels at 100m, 500m and 1000m  above the ground at that location and asked the model to show the trajectories of these air parcels over the next 48 hours.  The result is this:

 

This HYSPLIT trajectory model uses the output of the NAM weather model, specifically the wind forecasts, to predict where a blob of air at the coordinates and heights I specified will move in the future.  I ran the model for 48 hours.  In the output plot, the red line shows the path of the air that started at 100m above ground, the blue line is for 500m above the ground and the green line is for 1000m above the ground.  In the top half of the image, you can see the direction that the winds are forecast to blow the air.  So, from this, we can infer that smoke from those fires will head to the northeast (which matches up nicely with what we saw in the satellite images).  By 48 hours, the air at 500m and 1000m looks to have made it all the way up to southeastern South Dakota.  However, in the bottom part of the plot it shows the height above the ground of each parcel of air as time moves forward.  Notice that by the time the model run ends 48 hours later, the air that started at 500m has moved way up to probably over 10000m---very far above the ground.  The air that started at 1000m, though, stays much closer to the ground, even if it still rises to over 2000m.  And finally the 100m air pretty much runs into the ground and gets lost in the Texas panhandle somewhere.

So what can we take from these trajectory models?  Areas immediately to the east and northeast of the fire can expect a lot of air quality issues from smoke.  The further you go out along the path, though, the more spread out and diffuse the smoke becomes.  However, I still would not be surprised to see reports of hazy conditions in Nebraska over the next few days as the smoke from fires moves northeast.