What makes a Red Flag Warning?

We're getting into one time of year (another is in the mid to late summer) when we start to see a fair number of Red Flag Warnings crop up across the country.

Fig 1 -- US Warnings, watches and advisories as of 1918Z, Feb 19, 2011.

In the warnings and advisories map above, the dark pink areas in the trans-Pecos region of western Texas and southern New Mexico and also in southern Virginia and northern North Carolina are areas with red flag warnings.  The light pink and blue colors across the northern plains and upper midwest are associated with a potential winter storm moving through in the next few days.

During my time in Oklahoma, I saw a lot of red flag warnings come and go for the state, particularly on the high plains to the west.  But I never figured out just what exactly that unique warning meant.  I knew it had something to do with fire danger, but what conditions were needed to merit a red flag warning?

It turns out there are three criteria, all of which make sense in the context of fire danger:

1. Relative humidity at the surface that is less than or equal to 15%.
2. "20 foot winds" of 20 mph or more and/or gusts to 35 mph.
3. National Fire Danger Rating System danger level of "high" or greater.

These conditions must occur or be forecast to occur for three hours or more for there to be a red flag warning issued.  So let's look at these three criteria in the context of the weather right now.

Surprisingly, very few people seem to generate relative humidity maps at the surface.  As meteorologists we tend to be far more concerned with dewpoints at the surface as opposed to relative humidity.  Usually this is because relative humidity is "relative" to the temperature (hence its name).  Thus it makes a somewhat poor variable for actually looking at how much moisture is in the air.

With that said, I had to go to Intellicast to find a relative humidity map at the surface for today:

Fig 2 -- Surface relative humidity as of 2:00 PM EST.  From Intellicast.

Conveniently, we see areas of very low relative humidity right where the red flag warnings have been issued.  With a 10% RH area in Virginia and North Carolina, it's clear to see why the 15% or less criterion was met there.  However, it's more difficult to see in west Texas.  This Intellicast map shows an isolated pocket of 70% RH right near that area.  So why are we looking at a red flag warning if this criteria doesn't seem to be met?

Let's check this morning's sounding from El Paso:

Fig 3 -- This morning's 12Z sounding from El Paso, TX (EPZ) on Feb. 19, 2011.  From the HOOT website.

Note how in this morning's sounding from El Paso, the morning cold air near the surface is very visible.  Radiative cooling overnight means the surface cooled down a lot---look how cold the temperature at the very bottom of the profile is compared to the air just a little bit above.  Remember what we said about relative humidity--it's "relative" to the temperature.  In some ways (though not explicitly), the relative humidity measures how close that temperature profile above is to the moisture profile.  If the temperature at the surface warms up (which it does throughout the day) but the amount of moisture stays the same, the temperature and moisture profiles will separate more and the relative humidity will go down.  Plus, the air above the surface is also dry.  Strong winds of 20+ knots shown in the sounding (along with somewhat steep lapse rates above the surface cold layer) mean that there is probably a fair amount of mixing going on.  Through this mixing process, the dry air aloft will "mix down" to the surface and dry things out even more. Though the surface layer is already pretty dry--the dewpoint in the above sounding is below freezing at the surface.

So we have reason to believe that as things are warming up today the relative humidity will drop and the surface layer relative humidity will fall.  So we can extrapolate that the first criterion will be satisfied in west Texas if it hasn't been already.

What about the next one--the winds at "20 feet" need to be at least 20 mph or gusting to 35 mph. How do we check that?

There's an odd disconnect between this particular criterion and what we actually measure.  In typical weather stations, winds are measured at 2 meters and 10 meters above the ground--or, at about 6.5 feet and 33 feet.  The meteorological community likes to keep things in the international standard units, using meters and hectopascals and degrees Celsius.  But there are those occasional areas where we have to defer to the public's use of Imperial Units here in the US.  Hence at the surface, we use miles per hour and degrees Fahrenheit and whatnot.  I'm guessing this "feet" specification comes from the fact that these red flag warnings are designed for widespread public use, so 20 feet is easier to understand. (But why not 30 feet?  Why not the 6.5 feet?  These are things I need to look up...)

So here are the wind observations as of early this afternoon:

Fig 4 -- Surface observations for the CONUS at 19Z, Feb. 19, 2011.  From the HOOT website.

Of course, in meteorology at the surface we often try to cater to the marine crowd since they are heavy users of wind forecasts.  As such, our wind barbs are all in knots.  The 20 mph threshold we are looking for corresponds to about 17.4 knots.  So we'd be looking for winds in at least the 15-20 knot range.

It's somewhat of a hard sell in both regions based on these observations--also note these are 10 m or 33 foot winds, which tend to be slightly higher than what we'd see at 20 feet.  In west Texas there is a big hole in this observation map--to the east in west central Texas there are somewhat weaker winds around 10 knots.  However, we do see winds in the 20 knot range throughout central New Mexico.  On the El Paso sounding above we also saw winds at 20 knots out of the southwest a little bit above the surface.  With vertical mixing going on, gusts to or above 20 knots are definitely possible.  So we can assume that the winds will meet the criterion in west Texas.

In Virginia and North Carolina it's a bit harder to tell.  There are some 20 knot wind observations in the foothills of the Applachians.  Also note that the general flow is out of the west--downslope out of the mountains.  This could contribute to some adiabatic warming which would drop the RH values even further--just a side note there.  But anyhow, through much of the central part of these states, winds are only showing sustained values at 10 knots.  Let's check this morning's sounding from Greensboro, NC.

Fig 5 -- Sounding from 12Z this morning at Greensboro, NC (GSO) from Feb. 19, 2011.  From the HOOT website.

We can see the very dry lower part of the atmosphere.  But note the wind structure.  While the winds at the surface are weak, immediately off the ground they jump to 30 knots out of the northwest--then 50 knots at the next level up.  These are some very strong winds very close to the surface.  Even with the not-so-steep lapse rates shown here, as the surface warms during the day we should see some vertical mixing develop.  This mixing won't have to extend very far to start tapping into those high winds just above the surface and transporting some strong wind gusts down to the surface.  As such, with strong winds so close to the surface, I'd say that at least a forecast of high enough winds to meet our red flag criterion at the surface seems justified.

So what about the final condition?  The National Fire Danger Rating System describes itself as a bunch of models that try to forecast fire danger not only due to the weather, but also due to topography and the current quality of the fuel material on the surface.  For instance, lots of dry vegetation in the area would increase the fire danger and so on.  They have a somewhat rudimentary, but still very educational website here that details what they do and how they do it.  They also publish these fire danger maps based on current observations and on model forecasts.  Here is their forecast map for today:

Fig 6 -- Forecast fire danger map from the NFDRS for Feb. 19, 2011.

It's interesting how they develop these maps.  It appears that each reporting station has a "circle of influence" where they kind of assume that the entire area around the reporting station has the same fire risk.  Where these circles overlap, there is some kind of algorithm that blends the forecasts from the stations that overlap.  Anyhow, there is a 5-level scale as shown on the legend of the map--low, moderate, high, very high and extreme.  For a red flag warning, all there needs to be is a "high" risk--so anything yellow, orange or red on the map.  We do indeed see that west Texas and southern New Mexico are at the "high" level of risk.  Central Virginia and North Carolina actually get up into the "very high" risk.  So, based on these maps, that final criterion is indeed met in both those areas.

There's a look at what goes into a red flag warning.  Red flag warnings usually result in burning bans or cautionary statements that any sort of wildfire that starts will have a strong chance of escalating quickly.  Be cautious about any sort of open flames when under a red flag warning...

Looking at a the potential for a strong storm later next week in the Pacific Northwest and with warmer temperatures in the central US, all that snow from the blizzards is going to start to melt.  Flooding concerns are about to get huge--so start paying attention if you live in a flood prone area...