Weather conditions during the tragedy at Yarnell Hill, and where do we go from here

Radar at 5 pm MDT, June 30, 2013 The pointer is at Yarnell, Arizona.

We first wrote about the thunderstorm that may have contributed to the June 30 deaths of the 19 firefighters on the Yarnell Hill Fire, June 30 at 6:58 MST. But since that information is buried in a long article about the fire, and a little more information has emerged, we are summarizing the facts about the weather conditions that tragic Sunday afternoon. We also have some suggestions for providing firefighters with weather warning information that could save lives.

The formal investigation into the deaths, what caused them and any lessons to be learned, is just beginning. The results will probably not be known for many months. We will leave it up to the investigators to determine why it happened and what decisions were made before and during the incident. But the facts about the weather that day have already been recorded in various ways.

Carrie Dennett, a Fire Information Officer with the Arizona State Forestry Division, said the Arizona Dispatch Center received a call at 4:47 p.m. MST, June 30, that firefighters on the Yarnell Hill Fire had deployed fire shelters. The Dispatch Center was not in direct communication with firefighters on the ground at the fire. The information would typically have been relayed from the local Yarnell Hill Fire organization up through lower level dispatch offices.

Radar at 5 pm MDT, June 30, 2013 The pointer is at Yarnell, Arizona.
Radar at 4 p.m. MST, June 30, 2013 The pointer is at Yarnell, Arizona. WeatherUnderground.

The radar map above from WeatherUnderground at 4 p.m. MST June 30 shows a large thunderstorm cell north and northeast of the fire at Yarnell, Arizona. The pointer is at Yarnell. At 10:55 a.m., according to data from the MODIS satellite, the fire was approximately a mile or so north and northeast of the town, between the town and the approaching thunderstorm. At that time and until after 4 p.m., the wind blowing from the south-southwest and the southwest, most likely would have caused the primary spread of the fire to be toward the northeast.

Animations of weather satellite images that afternoon HERE and HERE, show the development and movement of the thunderstorm into the fire area. The red “X” and circle mark the approximate location of the fire.

Before and after 4 p.m. the cell was moving toward the southwest, and may have produced strong winds that changed the wind direction by 180 degrees (see below) and may have been the reason the fire moved into Yarnell. It also could have caught firefighters by surprise.

From 10 a.m. until 4 p.m. local time at the Stanton RAWS weather station four miles south of the fire, the wind was from the south-southwest or southwest, but between 4:01 p.m. and 5:01 p.m. it began blowing from the north-northeast at 22 to 26 mph gusting up to 43 mph — a 180-degree change in the wind direction.

Data from Stanton RAWS weather station, near Yarnell, AZ
Data from the Stanton RAWS weather station, four miles south of Yarnell, AZ, June 30, 2013

We were told by a spokesperson for the fire that the location of the firefighters when they died was between the fire and Yarnell, which would put them north or northeast of the town. Michelle Lee of the Arizona Republic told us that they were about one quarter mile southwest of Glen Ilah estates, “in the mountains”. Glen Ilah is on the southwest side of Yarnell. Anyone in that area between 4 and 5 p.m., who previously had the wind at their backs for seven hours with the fire moving away from them, may have suddenly and unexpectedly found the fire heading toward them at a rapid rate, pushed by winds gusting over 40 mph. Wind direction changes like this can be caused by strong outflowing winds from a thunderstorm in the dissipating stage.

At 5:01 p.m. the temperature was 95 degrees and the relative humidity was 17 percent. That, coupled with sustained winds of 26 mph with gusts over 40 mph, could have caused the rate of spread to increase to the point where it would have been impossible for any firefighting resources, in the air or on the ground, to implement any kind of effective fire suppression action, especially at the head of the fire which was moving rapidly toward Yarnell.

The Weather Forecast

A Type 3 Incident Management Team transitioned to assume command of the fire at 10 a.m. on June 30. That morning the fire had burned less than 1,000 acres. Incident Meteorologists from the National Weather Service, IMETs, are frequently assigned to large wildfires. While they are on site they can provide a great deal of very detailed weather information to firefighters, monitoring the conditions closely and in constant communication with the IMTeam.

It would have been unusual for an IMET to be assigned to a fire of that size on June 30, and with a small Type 3 IMTeam. The next morning, however, an IMET did receive an order to respond to the fire along with a much larger Type 1 IMTeam.

But even without an on-scene IMET, firefighters have the capability to request from the National Weather Service a special “spot forecast” for a fire, and at least two were provided before the fatalities occurred, one at 8:33 p.m. MST on June 29, and another at 9:45 a.m. June 30. The latter one, according to the information at the top of the forecast, was requested at 9:39 a.m. MST and produced six minutes later, a remarkably short turnaround.

The spot forecast for the morning of the incident predicted “isolated thunderstorm activity”, and, “Partly cloudy. A slight chance of showers and thunderstorms in the afternoon”. The predicted wind was “east winds around 5 mph…becoming southwest with gusts up to 20 mph in the afternoon”.

It is extremely difficult or even impossible for meteorologists to predict exactly where and when thunderstorms will occur, especially if they are “isolated”, or few in number. It is also difficult to forecast the exact wind speed and direction under any conditions.

Where do we go from here?

Firefighters need to know when there is a sudden change in weather conditions that can threaten their safety. Since it will never be possible to have an IMET at every fire, there needs to be a concerted interagency effort involving the National Weather Service and the *land management agencies, state, local, and federal, to develop a system to provide firefighters with the situational awareness information they must have to reduce the probability that they will be surprised by a life-threatening change in the weather.

Firefighter’s Emergency Situational Awareness Device —  FESAD

We may need new developments in hardware for firefighters so that they can receive weather warning data directly from the source without going through layers of bureaucracy. The military probably already has something like this, but I envision a device with a 7-inch display that could receive satellite transmissions anywhere there is a clear view of the sky. Each firefighter would not need to have a Firefighter’s Emergency Situational Awareness Device, a FESAD — just provide one for every fire. It could be based on a satellite phone, would have a GPS receiver (like in most smart phones), and be capable of receiving text and images. A deluxe version might also function as a satellite telephone, so that the firefighter could call the weather forecaster to get more information. The person sending the data to the fire would not have to know the phone number of the receiver, but would draw a box on a map and any devices within that box would receive the data, without bothering others that would not be affected by the warning. When the FESAD arrives at a fire, the user could have it send a message registering its location so that warnings for that area could be received. This would also alert the on-duty Remote-IMET to add it to their watch list.

Alert who?

Remote-IMET (R-IMET)

The warning data could be sent to the fire by an NWS forecaster always on duty with the primary responsibility to remotely monitor weather conditions near multiple fires. (The NWS already does this for the entire nation for severe storm, tornado, and hurricane warnings.) For fires, let’s call them Remote-IMETS. Some of the Geographic Area Coordination Centers already employ full time fire meteorologists. With the right software they could perform this function.

An R-IMET would have the technology available to instantly transmit text and graphical data about an emerging dangerous condition directly to those on the ground without relying on telephones to filter the information down through various layers. An R-IMET would not have time to handle all of the IMET duties normally performed by an on-scene IMET, but would concentrate primarily on issuing urgent warnings that could affect the safety of firefighters.

What if …. there had been a FESAD and R-IMET available for the Yarnell Hill Fire 

FESADs and R-IMETs could save lives.

The iron is hot

There is a great deal of national interest right now in the deaths of the Granite Mountain 19. If any new initiatives that would cost money are going to be implemented, our Fire Leaders must strike now while the iron is hot. There may be a small window of two to four weeks during which there will be some sympathy nationwide and citizens and lawmakers are open to enhancing the safety of wildland firefighters, even if it involves spending money.

An example of a rapidly closing window of opportunity is the movement for controlling gun violence after a gunman fatally shot twenty children and six adult staff members at Sandy Hook Elementary School in Newtown, Connecticut. If the administration and lawmakers had taken action quickly, meaningful legislation would have had a much better chance of passing. Instead, they waited months, and achieved little.

*A side note

It pains me to describe the federal fire departments as “land management agencies”. Organizations whose primary mission is to grow trees, clean campground toilets, or manage visitors, do not have at the top of their To-Do List, “Provide real-time weather warning information for firefighters”. We need a National Fire Service.

Typos, let us know HERE, and specify which article. Please read the commenting rules before you post a comment.

Author: Bill Gabbert

After working full time in wildland fire for 33 years, he continues to learn, and strives to be a Student of Fire.

21 thoughts on “Weather conditions during the tragedy at Yarnell Hill, and where do we go from here”

  1. Mr Morgan explains it very well

    My mentors who are now airline drivers and former flight instructors…….flying glass cockpits and some that are not…have always instilled SA into me…every flight…even today.

    Now with respect to all this technology….why does not the fire world start paying attention to the latest Korean 777 crash? Apparently there is some disconnect with pilots and automated flight systems and how to use them…

    I remember aircraft (airliners) had to be hand flown to the ground and in the last 20 + yrs…Autoland systems where the pilot is a mere “flight manager” overseeing automated systems…….see where this is going?

    All the technology in the world is NOT going to save everyone’s fourth point of contact…..SA does not come from auto mated systems in all cases…although there are many here that seem to think that more technology may have saved the 19 lives…possibly

    YES……. let us allllll look at why or if Mann Gulch, Storm King or Yarnell have similarities. I imagine two had no WUI implications and one did.

    Now let us all look at the topography of all three and two ( Yarnell and Mann Gulch , which may have some similarities)

    Then maybe look at alllllllllll the history and READ all the fire books (Pyne, et al) and try to convince me or plenty of us that we have truly learned our lessons through “Lessons Learned”. FLA’s, “LMA Staff Rides” etc and all the the other SMS, safety briefs, etc etc etc etc and tell me more about automated systems?

    Til we get or SA in order and start looking at some impossible topography and keep our heads to the sky, to actually start looking at thunderstorm development…..we will see more of these unfortunate incidents such as this and MAFFS 7 last year when thunderstorm development may or may not have been a contributing factor.

    Time for me to break out Ag Handbook 360 (Fire Weather) again……and turn off my computer!!!!

  2. Has any one looked into the similarities between Mann Gulch, Storm King Mountain, and Yarnell? Especially wind shear, fire whirls, and blowups?

  3. First and foremost I morn the loss of this brave crew.

    It does look like weather played a large part in this tragedy, but I will reserve judgment till it has been carefully looked at in detail by people with a lot better knowledge and information then myself.

    On technology, years ago my mentor and high hour flight instructor told me that, “We have the technology but our common sense has yet to catch up with it”.

    Bill is very correct in that we have to have a National (Wildfire) Fire Service to get everyone under one roof and in a cohesive organization with common goals, training, equipment, policy, management and a number of other things. It would be more efficient, cost effective and safer then what we have. And in down time they could work for the good of the resources. And it needs to be free of Homeland Security/FEMA, too big and cumbersome.

  4. It is very premature and completely unacceptable to be trying to figure this out or assign blame at this point! They could have had all the information and made all the correct decisions and simply have been in the wrong place in a dangerous environment. Fire is going to scorch and the only way to prevent it is to not be there in the first place. There should be no speculative sideline guessing, it could lead to confusion and misdirection about what really happened. While a thorough investigation is critical, the fact is that fire fighting is dangerous and deaths are unavoidable even if protocol is followed to the letter. I hate to see any hint that there was a blame to be placed when you have no facts at all to support assertions.

    1. @Joseph: thanks.

      For about 15 years now I have been waiting for some fire guy “spokesman” to quit with the agency line (aka “talking points”) and just stand up and say, “You know what? Sometimes s*** happens.”

  5. The RAWS station at Iron Springs, north near Prescott, has a more direct recording of the t-storm effects that afternoon. Look at the change from 1406 to 1506:
    QISA3 IRON SPRINGS 34.593611 -112.511389 5385 ft RAWS

    Data provided by: Bureau of Land Management & USDA Forest Service
    MesoWest Disclaimer

    English Units Variable Descriptions

    6,30,2013, 0,06,MST, 80.0,27,6.0,11.0,261,2,0.0,,3.20,,,,11,270,12.40,50.8
    6,30,2013, 1,06,MST, 78.0,22,6.0,12.0,228,2,0.0,,3.20,,,,12,263,12.40,45.1
    6,30,2013, 2,06,MST, 71.0,32,3.0,8.0,187,2,0.0,,3.20,,,,8,228,12.40,46.6
    6,30,2013, 3,06,MST, 73.0,32,3.0,5.0,226,2,0.0,,3.20,,,,5,224,12.40,48.3
    6,30,2013, 4,06,MST, 69.0,37,2.0,5.0,204,2,0.0,,3.20,,,,5,255,12.40,47.8
    6,30,2013, 5,06,MST, 69.0,34,2.0,7.0,327,2,0.0,,3.20,,,,7,314,12.40,46.1
    6,30,2013, 6,06,MST, 69.0,35,2.0,8.0,187,2,21.0,,3.20,,,,8,309,12.40,46.7
    6,30,2013, 7,06,MST, 81.0,29,4.0,6.0,204,2,183.0,,3.20,,,,6,197,12.80,53.1
    6,30,2013, 8,06,MST, 89.0,22,6.0,10.0,253,2,322.0,,3.20,,,,10,243,13.10,53.9
    6,30,2013, 9,06,MST, 91.0,18,7.0,11.0,69,2,558.0,,3.20,,,,11,248,13.20,51.3
    6,30,2013, 10,06,MST, 95.0,17,8.0,12.0,243,2,736.0,,3.20,,,,12,244,13.20,53.2
    6,30,2013, 11,06,MST, 96.0,14,7.0,16.0,255,2,871.0,,3.20,,,,16,240,13.20,50.0
    6,30,2013, 12,06,MST, 99.0,13,7.0,27.0,138,2,954.0,,3.20,,,,27,172,13.30,50.7
    6,30,2013, 13,06,MST, 96.0,12,8.0,19.0,284,2,752.0,,3.20,,,,19,254,12.80,46.9
    6,30,2013, 14,06,MST, 96.0,14,10.0,22.0,228,2,730.0,,3.20,,,,22,251,12.80,50.0
    6,30,2013, 15,06,MST, 79.0,17,15.0,25.0,93,2,256.0,,3.25,,,,25,82,12.80,40.9
    6,30,2013, 16,06,MST, 68.0,47,7.0,29.0,250,2,8.0,,4.19,,,,29,57,12.60,51.7
    6,30,2013, 17,06,MST, 81.0,51,20.0,30.0,263,2,9.0,,4.19,,,,30,281,12.50,65.2
    6,30,2013, 18,06,MST, 80.0,33,14.0,30.0,271,2,50.0,,4.19,,,,30,268,12.50,54.9
    6,30,2013, 19,06,MST, 80.0,29,7.0,14.0,282,2,86.0,,4.19,,,,14,268,12.60,52.2
    6,30,2013, 20,06,MST, 77.0,33,4.0,10.0,329,1,13.0,,4.19,,,,10,336,12.50,52.3
    6,30,2013, 21,06,MST, 69.0,54,6.0,8.0,298,2,0.0,,4.19,,,,8,307,12.50,55.5
    6,30,2013, 22,06,MST, 67.0,60,2.0,8.0,206,2,0.0,,4.19,,,,8,308,12.40,55.9
    6,30,2013, 23,06,MST, 68.0,59,3.0,5.0,288,2,0.0,,4.19,,,,5,314,12.40,56.4

  6. What about the NOAA Weather Radio broadcasts? I receive these on my scanner almost anywhere that I go. Would they not be available and useful in similar situations?

    1. NOAA radio broadcasts are more generalized for an area. They don’t give specific information about individual storm cells or the timing of passing fronts at a specific location that cause the shifting wind issues that tend to lead to these kinds of tragedies. I used to program NOAA into our radios, but really, Fire Weather Forecasts for your specific Zone gives you that information, and you can look back and forth, without having to concentrate so acutely on listening to make sure you catch all the information.

  7. Just re-read the original post, and noticed the Side Note. You are right Bill, many times, that is the crux of the problem. I remember a screaming match I had with the Facilities Unit Leader and Camp Manager on a fire managed by a Type 1 Team from down south. He was not thinking about firefighter well being, he was managing his buildings. I had all sorts of heads prairie dogging out of doors in the school we were using as a Camp! They admitted to being wrong, but our Strike Team still had to deal with our pile of gear before we went to sleep.

  8. And yet it’s a human endeavor. Every crew has a spotter. Somebody is flying around the fire calling in every air drop, staying out of the smoke. Anyone living around Prescott knows the thunderstorms are fast, violent, unpredictable. Perhaps this tragedy is less about information and awareness, and more about trying to hold a line. A firefighter is not conditioned to get out of the way and watch the town burn….

    1. “A firefighter is not conditioned to get out of the way and watch the town burn”
      A very strong statement about the culture of the the wildland community and maybe more so with Hotshots.
      Instead the nation will watch them on tuesday.
      No town is worth any of this. Is this our decade reminder of that.

  9. Once again thanks Bill for for all of the informative information. These are great discussion points and there are tools at our disposal for these kinds of situations. The problem is we get so caught up in these tools and we loose site of the most important tool of them all, which is experience and the old saying look up, look down, and look around. It’s all about SA and we all loose it from time to time.

    1. I agree with Mr.Zach technology today is amazing & we as firefighters should take advantage of all of it but unfortunately with most tragetys is to look for ansers. Thunderstorms are a huge part of wildland fire ( how most fires start) this storm was no different than other storms. It is unfortunate that this crew found themselfs in a bad place with not enough time to react… I believe they did loose their SA.. We should not assign blame but try and learn. I do not believe a weather adviosary system should ever replace experience or good SA…

  10. IMO, just a weather radar app on a smart phone could make a significant difference. We are all trained to understand weather. Anymore, we all carry our phones, mine fits in my radio (chest harness), even with my BK and a second radio in it. I’m not sure that adding another device is the way to go. We just need to become better at using the available technology. Then, trickle that technology down to all levels of Incident Management. The Comm Unit that serves our local Type 3 team has portable Incident repeaters, I suspect they have the expertise to set up a portable cell phone repeater. If there are areas not covered by cell signal, maybe we should be fixing those coverage gaps in current technology. When we teach extrication, or structural Search and Rescue, we teach it’s more efficient to use the means of egress that people ordinarily use to move around. The principle is the same, use things people are used to, and they will use those tools more effectively than something that is only used for a 2 to 4 months of the year.

    1. Rather than add yet another job to the Comm Section, why not make arrangements ahead of time with cellphone companies to provide local repeaters as needed, much as departments now make mutual aid arrangements before they are needed.

  11. There was ample cell phone signal to text a picture from the fire which means there was ample cell phone signal to check the current weather conditions from your favorite wx website. Where’s your SA?

  12. Bill; as always,nice job putting together pieces of the puzzle. Keep up the good work. Our hearts and prayers to those who’s lives have been impacted by this tragic event.

  13. For years I have received severe weather warnings sent to me by e-mail. I received the following on July 1:

    Dear GovDelivery Weather Alert Subscriber,
    We are writing to inform you that this notification service will be discontinued effective July 31, 2013.
    Earlier in the year, it was announced that GovDelivery would continue providing this weather alerts after the National Weather Service discontinued the service.
    Due to the substantial costs of providing a messaging service at this scale with high reliability, GovDelivery cannot continue the free service.
    We invite you to visit to sign up for other weather alerting services.
    Please contact with any questions about weather information from the National Weather Service.
    GovDelivery’s policy is to hold any end user data for up to 24 months in the event that a government client chooses to reactive its account. If you wish to delete your information at this time, you can do so now: Delete Account
    Thank you,
    GovDelivery Customer Support Team

  14. Several of the portable aviation oriented GPS receivers have real-time weather & weather radar available via XM satellite radio. No cell-phone coverage needed. They’re about the size of GPS units we use in our cars/trucks & can be used for driving as well. WX coverage subscription from XM currently about $100/mo and provides coverage nationwide.

  15. When I submit Spot Forecast requests for planned prescribed burns, I often include my cell number and ask to be notified if significant weather (t-storms, outflow boundaries, etc) are exptected to impact the burn location. NWS staff in Miami have contacted me with that type of information and it has been very helpful.


Comments are closed.