July 02, 2013

The Yarnell Hill Fire: The Meteorological Origins

This morning I took a look at the meteorology associated with the Yarnell Hill fire in Arizona on Sunday, and the more I dug into it, the more disturbed I got.  You will see why as I explain.

From what I can glean from news reports, the fire blew up around 4-5 PM Sunday (June 30th).   A nearby observation site (RAWS station) was located about 5 miles away.  The observations, shown below, indicates a profound wind shift from south to north around 5 PM associated with a sudden increase of wind gusts to just over 40 mph.  Solar radiation dropped rapidly at the same time, indicating a sudden increase in cloudiness.

The origin of this sudden increase in wind is clear:  outflow from a line of convection (thunderstorms) that had developed during the preceding hours and which was moving to the southwest.  Here are some satellite images for the hours preceding and during the terrible accident (the circle indicates the location of the fire).  First image (20 UTC, 1 PM MDT, no daylight savings time there) shows the convective line to the northeast.

By 2230 UTC (3:30 PM MST) the convective line was approaching the fire and clouds had spread over the location.

A little over an hour later (2345 UTC) one can clearly see the development of a cumulus tower directly over the fire.   This is call pyrocumulus.  The heat from the fire can cause a tall cumulus cloud to form directly over the fire.

The Flagstaff National Weather Service radar clearly showed the approaching convection.  Here is the radar at 2:58 PM.  You can see the arc of red/yellow/green colors approaching the fire from the NE.
There is often an outflow of cooler air moving away from convection...the leading edge is known as a gust front (see figure).  Downdraft air from thunderstorms spread out as it hits the surface, producing strong winds.  It appears that there was such strong outflow from this convection that caused the

winds to shift rapidly from southerly to northerly and to increase suddenly in speed (to 43 mph at the nearby station).   The vertical sounding at Flagstaff, Arizona at 0000 UTC July 1 (5 PM on Sunday) showed the potential for strong, downdraft winds, with a moist layer at midlevels (the temperature and dew point close together between 650 and 300 hPa) and dry air (big separation between temperature and dew point) near the surface (see graphic).  As rain falls into dry air, there is strong evaporation and cooling, that produces negatively buoyant (descending) air parcels that accelerate towards the surface.  When they hit the surface they spread out, producing intense horizontal winds.

A measure of the potential for strong downdrafts and gust fronts is something called downdraft convective available potential energy (DCAPE).  The sounding at Flagstaff has values of around 1600 J per Kg, which is very high (anything above 1000 can produce strong downdrafts).

The existence of the strong convective outflow winds is confirmed by an amazing video of the area from 4 to 4:20 PM (click on image to view, cam viewing north).  You will see strong winds picking up, an explosion of the fire, and then smoke pushing down towards the cam.  You can see a fire line explode along the crest.


So it is apparent what occurred ..first the winds were from the south, followed by a rapid shift of 180 degrees, sudden increase of winds to over 40 mph, and the fire blew up and reversed direction. 

Numerical model forecasts of this event were quite good.   NOAA runs a High Resolution Rapid Refresh (HRRR) system that makes high-resolution (3-km grid spacing) forecasts every hour, using many types of observational data to initialize it.   The forecasts from this system, started at 1900 UTC  (noon MST) and was available by 3 PM (you can access them here).   I put a red oval in the first picture to show the location of the fire, and the plot shows maximum wind speeds (knots) over the past hour..  The model simulated the convection fairly well as well as the winds it produced.  Here is the forecast for 2 PM...you can see the strong winds (red/purple colors) to the  northeast of the fire.

The forecast for 4 PM shows the winds reaching the fire site.

The University of Arizona WRF forecasting system also indicated the potential for strong convection-related winds. (see graphic, click to expand)

You can see why I find this disaster so unsettling.   Hours before the incident it was clear there was a real threat...satellite and radar showed developing convection to the north that was moving south towards the fire.  High-resolution numerical models showed a threat.  Were there any meteorologists working the fire?   If not, why not?   This terrible tragedy needs to be reviewed carefully.

A number of media outlets called the strong winds unpredictable and random.  This is not correct, as shown by the information I provided above. 


  1. I think asking "Why not?" is definitely the right question. And if the proper resources were assisting with the fire, did they raise an alarm? And was it ignored?

    It's hard to see something like this happen and see that it was so completely preventable if the proper resources had been applied.

    Thanks for the research on this!

  2. Great report Cliff, the best I have read yet.

    I was involved in the formation of Prescott's very first hotshot crew back in the early 70's. It started out as a Type II crew then became a Type 1 hotshot crew in 1973 -all under the USFS. It was mainly composed of college students that were active rock climbers, mountaineers and environmentalists. It was co-ed, it fought fires safely all over the West, including that big one in Twisp years ago, it even did some helitac work although it was not a helitac team per se. The Granite Mountain Hotshots started in 2002 under the City of Prescott Fire Department, it was an inter-agency team.

    One thing that is driving me crazy about the way the media is reporting this is exactly what you point out - that somehow the weather was all a big surprise.

    What people do not seem to realize is that the fire itself creates its own dynamics, this is well known and should be anticipated. Planning an attack on a wildfire starts out with a series of "IF" questions, the biggest of which is: "What IF what we think turns out to be wrong? If we go to position X, what do we do if the winds reverse or change? What do we do if things go to hell?"

    Over the years, a series of rules and methods have been developed that account for even the most extreme conditions.

    It is a well known pattern of tragedies that no single decision is the cause, it is a series of mistakes, miscalculations and misjudgements taken together that are the problem. This time it will be no different.

    Losing an entire hotshot crew means some bad thing happened at the response co-ordination level, weather communications being just one.

  3. Meteorologist was available, but has to be -- and was not -- requested by the fire manager, per this:

    (posted by dbostrom

  4. "Were there any meteorologists working the fire? If not, why not?"

    Cliff, you do understand that the fire occurred in Arizona and that the state government there is dominated by one political party whose emphasis has been tax cuts vice public service, right?

    And you are aware the federal budget cuts often called "sequestration" chopped 5 percent from the U.S. Forest Service's Fire and Aviation Management Budget in fiscal 2013, right?

    The answer to your question is "free market solutions".

  5. This is a very interesting read. Thank you for your research and post.

    Hopefully this information,along with your questions,will change the protocol for dealing with these fires.

  6. Here is the url to the National Interagency Fire Center website, and "the rules" -the "10 Standard Fire Orders and 18 Watch Out Situations" -if anyone is interested.


    The media has language like "Prescott has one of the 110 hotshot crews" -or some such. Actually, Prescott has 2 hotshot crews, the The Prescott Hotshots which has been a part of the US Forest Service since 1972, and the team that was lost, the Granite Mountain Hotshots which is part of the Prescott Fire Department since 2002. Both are part of the National Interagency Hotshots under NICC coordination.

  7. Cliff, Great analysis. The 2013 Wildfire Safety refresher had two modules where they talked about the weather www.nifc.gov/wfstar/

    Tim, Excellent link and also this one www.floridaforestservice.com/wildfire/safety_wf_tragedy.html

    Both of these lists are printed on the Fireline Handbook.

    The class room portion of the S-130/S-190 Wildfire Fire Fighter training are on-line.
    (current "red card")

  8. Thank you for this insightful research. I hope it might help prevent the tragic loss of such brave men. We are likely to see more of this type of events unless more rigor is applied

  9. Animations of the GOES visible images (along with a few other satellite image examples) are availabel on the CIMSS Satellite Blog: http://cimss.ssec.wisc.edu/goes/blog/archives/13341

  10. BoingBoing posted a link to a piece on climatewire that looked into this. It seems there were some meteorological staff in the mix but not enough, not in time, etc. It looks like the climax of this came at a shift change or other hiccup in staff awareness. But as you say, there were some predictable elements to this.

  11. Ma ybe its easier for the managers in this situation to believe it was all random and unpredictable. Otherwise how do you justify the loss of so many young, wonderful people.

  12. The storm wind change was no suprise at this time of year. Speaking as a layman who grew up close to Prescott and follows Arizona meteorology closely, what we saw was like clockwork for the unstable winds to change direction from N to S. During this monsoon season, storms began after that 60° dew point mark over AZ. thick forest after a strong bar(s) of High fronts. Every H storm front in July will fall south from the mogollon rim that is a 1000 to 2500 foot drop for aprox 500 miles in length. That storm begans a cold mass and falls fastly toward the heat, which naturally moves S. Swarming cool weather down towards Prescott area almost every day at this time of year. Of course sometime variates in strength about 48 hours by the rise of dew point as energy builds again. So, Prescott valley had the fire blazing(hot air) and the mogollon rim had a Large cold mass falling quickly as expected but with the fire winds are going to change to the S. right away. The wind may shift any direction with the fire and the cold mass ,although a local would expect to see the winds go more south with normal June temps. I believe the hot shot comand should have known weather would be unstable by looking up at the large structured cumulus clouds push that Cold Air mass that naturally falls towards the heat below 1500 feet. Again, locals expect this storm, and the Yarnell hill fire exaserbates moisture towards heat and every local would tell you there would be strong winds and maybe rain coming an hour ahead of the front moving S. That fire had a bullseye on it expecting S to SW winds to blow with verga rain or rain with extreme hail, with the fire that was unlikely to see rain or hail, but strong winds and unstable rain at about 1500 to 1800 hours (mst) will very likely be seen.

  13. "Were there any meteorologists working the fire? If not, why not?"

    Hi Cliff, decent analysis but the above question needs clarification especially in light of the way off the mark and misguided comment from RentonRebel. Hank Roberts was on the right track. An IMET (Incident Meteorologist) is generally ordered when a Type 1 or 2 fire fighting team is ordered to manage a fire. These teams are ordered when the fire has reached a certain complexity level that requires the greater resources these management teams offer over the local fire departments, Forests, etc.

    Once a Type 1 or 2 management team is ordered, the team decides if an IMET is needed and ordered up if so. Until then, the firefighting team works with the local NWS office to get the needed weather information. This is done through Spot forecast requests and briefings done over the phone. So, weather support is there from the beginning via the local NWS office or a NWS IMET when one is ordered.

    Weather support was ongoing with this fire as well and we will all learn the details of this tragic incident once the investigation team (now assembled) finishes its analysis.

    One thing the fire fighting does very well (beside fighting the fires!) is to learn from these types of incidents to try to prevent future injuries or worse. From someone having been on the ground, one thing that everyone needs to be aware of is this is a VERY dangerous job and that fact that there aren't more "incidents" with the extreme fire behavior we have is a testament to the skills of the fire fighters and the excellent safety procedures already in place.

  14. Along similar lines, my thoughts on the recent wildfire and tornado disasters: http://www.buzzbernard.com/does-a-hurricane-disaster-loom-why-i%E2%80%99m-worried/

    (Didn't we learn anything from Mann Gulch and Storm King Mountain?)

  15. This exact wind phenomenon happened yesterday here in Cottonwood... saw the roll cloud form just like the image. I thought an F-0 tornado was forming. Was a really scary event. Trees looked as if they were going to snap. Winds were probably 40mph. Very accurate analysis, man.

    My sympathy for the fallen and their families. I shed tears for them.

  16. @ WXMAN42,

    You did not explain why you thought my comment was misguided. Maybe it was just a misunderstanding.

    What we knew before the fire was:
    a) The frequency and severity of fires has increased by almost 50% since 2000
    b) The federal and state budgets have not nearly kept pace with this increase with the amount of resources required to fight these fires and keep the population, property and Fire Fighters prepared and safe.
    c) The sequestration which resulted from the Republican party's policy to no longer authorize the payment of the debt it voted to incur has impacted the U.S. Forest Service’s Fire and Aviation Management Budget in fiscal 2013 with 5% chopped out.

    d) As Cliff meticulously pointed out, the technology existed to prevent the deadly "surprise" on the Fire Fighters.

    I agree that lessons can and will be learned from this disaster.

    My point is that we need not wait for death to learn lessons.

    Meteorological research and resources are a public good which MUST be financed by the public it supports.

    Yes, the process exists where if requested an IMET is assigned. It was not requested in this case.

    The ability to assist fire fighters in fires large AND small exists without dependency upon special requests. This ability is specifically not employed because of political budget constraints.

    The policy to wait until death necessitates change processes is a political budget decision, one which is demanded by one specific political party and ineffectively repudiated by the other.

  17. @

    Adding a citation here to buttress my point.

    Re CBS News about the House Republican budget authored by Paul Ryan, GOP VP nominee 2012 - March 11, 2011, 4:17 PM

    The budget, which proposed about $60 billion in budget cuts, would slash funding for the National Weather Service and the National Oceanic and Atmospheric Administration (NOAA).

    The Republican's proposed "continuing resolution" to fund the government, which was defeated in the Senate this week, aimed to cut $1.2 billion - or 21 percent - of President Obama's proposed budget for NOAA, ClimateProgress.org reports.

  18. Thank you for your analysis Dr. Mass. I am an NWS Incident Meteorologist based in Anchorage, AK, and have been serving on wildfire and other incidents since 1990, throughout the US and Australia. The reason there was no Incident Meteorologist at the Yarnell Hill fire when it blew up was that it was still a "Type 2" or "Type 3" fire, not large or complex enough for the "Type 1" rating at that point to require a larger Incident Management Team. Who almost always will request an Incident Meteorologist.

    No matter what we do as well, as Incident Meteorologists, there are some realities to face. For instance, while serving as an IMET on a fire near Stanley, ID in 2006 I forecast isolated dry thunderstorms to occur in the afternoon around the fire area. The region around Stanley is outside of WSR-88D radar coverage from either Boise or Pocatello. All I had to go on for my weather watching was lightning detection and satellite.

    At 4 pm that day (9/02/2006) isolated thunderstorms developed over one of the fires in the complex, and a strong downburst wind pushed the fire toward 34 crewmembers. I was not able to warn for this, lightning detection just showed 1-2 strikes after the fact some miles away. The fire was 20 miles from camp where I was stationed, and not even visible from it. Yet all 34 firefighters were able to run through their pre-defined escape route to their safety zone, and made it. They were out of radio contact with our base for 90 minutes while this was occurring! You can bet we all were petrified at that time.

    So it is the training and quick action of the firefighters following their standard orders, in addition to our weather support, that keeps everyone safe. Thank you, Michael Richmond

  19. High Base Thunderstorm rolled in (as predicted in morning fire weather discussion) blew the fire in all directions, and brought Mother Nature her tragedy. It's as simple as that. As Michael said you can't have IMETS on every fire. But you can utilise the local office for spot fire weather forecast and the like.

    There are a couple of things bothering me at this time. Live and dead fuel moistures at the time of the event were at historic lows. Fuel loading was at historic highs. As Malcolm Gladwell would say; "this is an outlier". This type of weather is typical for this burning period. What was not typical was the state of the fuels. On any other year a red flag watch or warning would not be issued for this geographical area for the threat of thunderstorms. But on Sunday with the georgrahical area under extreme burning conditions maybe it least a red flag watch could have been posted. I don't want to armchair the weather office but it needs to be looked at.You don't want to be pulling the flag up unless it warrants it. South Canyon was issued a red flag warning in 1994. Chris Cuoco' message never made it to the fireline. That was the tragedy.

    There have been countless burnovers and close calls from thunderstorms. Hell,the dude fire did the same thing. Predictable is preventable. You you can count on a fire being unpreditable under such conditions. That's the thing. In the end it is predictable.

    You can bet on the lack of acting on intuition which brought flam on there backs. Just like Mann Gulch and South Canyon. Assumptions of the person above you in rank or in other agencies who knows what they are doing and not wanting to question their actions. We will see.

  20. @RentonRebel
    To explain and hopefully address my thoughts better I will try to clarify. Your initial comment appeared to be an attempt to take a tragedy and turn it into a political statement without knowing what the whole process is, how the system works and is set up, what actually happened on the ground on that day or how fiscal policy affects/doesn’t affect the situation. Unfortunately there are a lot of assumptions being made by folks that don’t know what did or didn’t happen on the ground and that is a foolish place to go.

    Even Cliff, whom I respect for all he has done with the NWS locally (Langley Hill radar) and nationally (pushing NCEP modeling efforts) and knows a lot about the inner workings of the NWS doesn’t understand how the NWS supports the firefighting agencies, either through the local weather office or through the IMET program. I say that based on the question he asked in the last paragraph “Were there any meteorologists working the fire? If not, why not?” If he understood the process and the working relationship between agencies, that question would not have been asked.

    Everyone is assuming that the outflow was a “surprise.” At this juncture I’m not saying it wasn’t but this assumption is based on what? Your perusal of the information disseminated between agencies or the logs of all the briefings that might have taken place between the various agencies at work or amongst the fire fighters themselves? If folks 1500 miles from the scene can sit back and “see” what happened clearly why assume that the local folks that deal with this type of weather every season didn’t “see” it coming? A quick and dirty primer on the IMET program can be found in this article along with some information on the weather support that had taken place that day. It’s a good and worthwhile read. http://www.climatecentral.org/news/incident-meteorologists-on-the-front-lines-of-wildfires-16189

    There are so many other factors that could have led to this tragedy such as underestimating the change in fire behavior (historically low fuel moistures, high fuel loadings, etc.) with the wind shift, safety zone issues, communication problems, extreme topography complicating the escape, etc. Without the proper information which none of us have, we can’t sit back and make blanket assumptions and statements about what happened.

    One must also be careful about making blanket statements and assumptions about how fiscal policy affects or doesn’t affect the situation in the field. Local forecast offices are at the ready 24/7 to support the teams as needed and IMET support is there when needed as well, so “the ability to assist fire fighters in fires large AND small” exists and is in place and has not been impacted by the fiscal issues of late.

    It would be best to let the investigation team do their work and discover what actually happened. Then, based on what they discover work toward solutions that will help prevent a similar tragedy in the future. Jumping to uninformed assumptions and conclusions is much more harmful than helpful.

    Hopefully I have explained my thoughts and not inflamed which is not my intent.

  21. Although your piece brings up very good points for the future of firefighting my gripe is that the BLM just stood around watching this fire at 8-10 acres on the first day and did nothing to put it out. Many friends who share Yarnell as home are now homeless and 19 valiant firefighters have perished.

  22. This fire started Friday afternoon! Lightning struck on friday.

  23. @Gelyn

    According to early reports Arizona State Forestry took over the fire the next morning after the lightning came through. 2 crews,3 engines,2 air tankers and one egg beater where assigned by 10:00 am. Unless you have more solid info please share. If it was truly a BLM show for the first 24 hours, had command of initial attack and "willingly" held back resources through out the next burning period-----they have another South Canyon on their hands. Be careful with your words in the early stages of the investigation.

  24. The Wildfire Lessons Learned Center ( http://www.wildfirelessons.net/Home.aspx )has a searchable database (Incident Reviews) of wildfire incidents going back many decades. The reviews of past incidents (in this case "entrapment" or "burnover") are very detailed and contain critical assessments and recommendations.

    Unfortunately, that database contains far too many similar incidents of entrapment due to sudden, but predictable, wind shifts.

    As mentioned the Dude Fire (Walk Moore Canyon in Payson, Arizona 1990) was very similar in cause and result with six fire fighters trapped and killed by a sudden wind shift due to convective activity. As a result of that incident many changes were instituted to diminish the potential for entrapment due to sudden, and somewhat predictable, wind shifts.

    The Australian Rural Firefighting organization produced a video in 2000 (youtube http://www.youtube.com/watch?v=uMGuiv2SYeg) that documents the dangers due to sudden wind shifts. They detail many tragic events that have occurred during wildland fires in Australia. My understanding is that video is a required training item for all US wildland fire fighters.

    It is hard for me to understand how, with the strict standards set for US wildland fire fighters, especially LCES, and the modern understanding of the effect of rapidly moving convective activity on wildland fires, especially in steep terrain with many closed drainages, how the Granite Mountain Hotshots were assigned the task that led them to their entrapment and how spot forecasts were not demanded by their superintendent, and how Yarnell Incident Command did not order their withdrawal when the convective activity moved so rapidly toward them.

    I am sure the Lessons Learned Center will do their usual good job of determining where the breakdown occurred in forecasting, or communicating, or on the ground decision making.

    My condolences to all involved.

  25. @WXMAN42

    Is there some reason that a direct comparison couldn't be made between this fire and the Dude Ck Fire (http://www.fireleadership.gov/toolbox/staffride/downloads/lsr11/lsr11_info_summary.pdf)? Sincerely, it appears that virtually the same situation occurred again, yet nobody had learned from the last one, so it was repeated.

    The USDA Forest Service has to understand that when forecasted weather conditions show the risk of extreme fire behavior, it is going to be awfully hard to explain why people were exposed to it merely to protect property, despite the increased risk. No fire on extreme fire days is routine and noone should send crews in assuming it will be routine.

    Occupational Health & Safety requires the provision of a safe work place. Any fire on extreme fire days is by definition, not a safe workplace. By all means, take risks if lives are at stake, even some risks if property is saveable, but on no account should risks be taken to save property or material that is shown by modeling to be unsaveable given the potential conditions.

    The rest of the world has adapted to reality and imposed virtually those same restrictions upon firefighter managers (notably the UK, Canada & Australia). It is well past time that the US did the same. Anyone who willfully ignores the risk and flouts it, should be held to account, like any other manager, in the criminal courts.

  26. i Cliff - I read through your material and it fit exactly what occurred in Colorado Springs last year. I happened to be on I-25 just north of the Springs when the fire jumped the mountain and came down into town. There was a storm at that moment causing airflow from west to east where as before the storm the airflow was south to north. The south to north airflow gave confidence to the people in the Springs that the fire would not affect them. The sudden shift in winds caught everybody off guard. Thankfully, no firefighters lost their lives. But, an unfortunate elderly couple were trapped and died in their home. I hope that the meteorological community becomes aware of the critical role they have to play in predicting forest fire activity. Thanks for your efforts.


Please make sure your comments are civil. Name calling and personal attacks are not appropriate.

Rain without Clouds, the Upcoming Cooling, and Strong Leeside Winds: All in My New Podcast

The radar image this morning at 5:30 AM showed rain...some heavy... offshore. As shown in the satellite image at the same time, much of that...