Before the smoke clears…

by Jerome V. De Graff

It was mid-day on August 9, 2013 and I was standing next to my Forest Service pick-up having a déjà vu moment.  Before me were the steep slopes of the Tuolumne River Canyon blackened only days before by the advancing flames of the Rim wildfire.  I had seen an identical scene during the Stanislaus Complex Fire twenty-six years earlier.   At 145,980 acres, the Stanislaus Complex still ranks as the fourteenth largest wildfire in California history.  The Rim Fire would end up burning 257,314 acres to become the third largest.  Of the twenty largest wildfires in California between 1932 and 2013, three took place during the first fifty-five years.  The remaining seventeen roared across the landscape during the last twenty-six years.  Times had changed.

Figure 1 – The canyon of the Tuolumne River looking east (upstream) from a point on Lumsden road east of Groveland, CA. Smoke from actively burning area of the Rim Fire obscure the more distant features.

I was once again serving as a geologist on a Burned Area Emergency Response (BAER) team; my thirty-sixth one in as many years working for the U.S. Forest Service.  BAER teams conduct a rapid assessment to identify threats to public safety, property, and the environment attributable to fire’s effect on the landscape of a national forest.  BAER teams are composed of engineers, geologists, soil scientists, hydrologists, and archeologists to name a few of the specialties involved.   The speed associated with making this assessment stems from the short time available to not just identify where these threats exists but also to fund, design, and install measures to mitigate them.  The mitigation work typically must be completed before the first damaging rainfall event.  Because many wildfires take place in late summer and early fall, there are only a number of weeks available to accomplish the process from threat identification to mitigation measure completion.

My job as a BAER team geologist had not changed much between the time of the 1987 Stanislaus Complex and 2013 Rim fires—but the technology and science had.  My role was to identify where increased hazard from debris flow or rockfall activity might result from fire’s impact on vegetation and soil and to recommend mitigating actions where the hazard threatened values-at-risk (people, structures, roads, etc.).

Fire never burns evenly across a natural landscape.  The distribution and types of vegetation present on the landscape is responsible for this variability because vegetation is the fuel sustaining the fire.  Another variable is topography.  The orientation of ridges and canyons to the prevailing winds blowing during the course of the wildfire changes the fire’s behavior.  Large fires create localized winds and surface heating which can interact in unique ways within the landscape.  So the burned area is a mosaic of unburned to intensely burned vegetation that needs to be mapped as a starting point for a BAER team assessment.  This is known as a soil burn severity map.

During the Stanislaus Complex Fire, I helped make our soil burn severity map using the technology of the time—the sharp-eyed, trained observer.  Armed with topographic maps and a felt-tipped maker, you took a seat in a helicopter that would give you a clear view.  It was also a good thing if you were largely immune to motion sickness because the next few hours would be spent circling around and around over the fire.  As you passed over the burned area, you would draw large polygons and assign a burn intensity based on how the vegetation looked from about a thousand feet up. Later, spot checking of these polygons by soil scientists would modify the initial polygon boundaries or the assigned intensity rating.

For the Rim Fire, the initial soil burn severity map arrived byte by byte as it was transmitted from the Forest Service’s Remote Sensing Applications Center (RSAC) in Salt Lake City.  Called a BARC (burned area reflectance classification) map, this digital map was produced by algorithms which interpreted ground reflectance differences between satellite imagery taken before and after the wildfire.  Again, spot checking identified locations confirming or altering soil burn severity ratings for the final map polygons.  While the time needed to produce a soil burn severity map has not changed much between the Stanislaus Complex and Rim fires, the soil burn severity mosaic across the landscape is now represented at a much greater resolution.

A companion change in technology contributing to current soil burn severity maps is the now pervasive use of geographic information systems (GIS) technology.  I was working on the Rim Fire with one of the GIS specialists I had worked with on the Stanislaus Complex Fire.  He told me the primary GIS contribution in 1987 was compiling a table showing the number of acres burned for different land ownership categories.  The working area for the Rim BAER team (which is always referred to as the BAER Den) was festooned with different maps on the wall, stacks of them on the back table, and multiple maps tucked in the pockets or packs of different specialists going into the field.  Some maps showed soil burn severity on a topographic base while others showed it with only the road network or with the boundaries of affected watersheds.

Figure 2 – With the soil burn severity map taped to the wall, BAER team specialists use their laptops to model erosion, water flow and other changes due to the effects of the fire on the landscape.

While visual display possibilities for GIS maps seem endless, it is their application in assessment models that is the true advance.  Hydrologists incorporate them into models for predicting changes in stream flow.  Soil scientists are able to predict additional volumes of soil being eroded from the slopes.  My assessment work benefits from GIS in two ways.  The first is having the U.S. Geological Survey modeling debris flows; their probability, likely debris flow volume, and combined hazard for every watershed affected by the Rim Fire.   Remarkably, this is done from their offices in Golden, Colorado.  In close coordination, the BAER team provides soil burn severity, fire perimeter and other GIS-based data to them via our Internet links.  Using their empirically-based models and additional data, they generate and transmit the maps needed to show the different levels of debris flow hazard.  The BAER geologists then focus on identifying what is at risk and how that risk might be mitigated.

The modeling of debris flow hazard attributable to the wildfire’s effect on the landscape is possible because of a fundamental change in our understanding of this process.  At the time of the Stanislaus Complex Fire, greater debris flow hazard was understood to require having burned area where a predisposition to generate debris flows and high soil burn severity coincided.  Past debris flow activity would have resulted from infiltration-triggered slope failures.  Most of my assessment effort in 1987 involved examining high burn severity areas for any evidence of past debris flow activity upslope from values-at-risk.  If it was present, a threat existed; if it was not present, no threat existed.  This proved to be somewhat inaccurate when the rains came because debris flows did not happen where several hundred thousand dollars in mitigating structures were built and did happen elsewhere where nothing had been done.

Research since 1981 has demonstrated that debris flows after wildfires are generated primarily by runoff-dominated erosion due to surface overland flow.  The debris flow modeling done for the Rim Fire was based on factors identified and empirically quantified as influencing this process based on examination of post-fire debris flows throughout the western United States.

During assessment of the Stanislaus Complex Fire, I spent no time looking for areas of increased rockfall hazard.  Recognition of this wildfire-related hazard is another recent development and is particularly important as a public safety issue for vehicles using roads or hikers walking trails within a burned area.  Using the physical and fire-related characteristics associated with burned slopes where rockfall occurred on roads in other wildfires, a GIS-generated map of the Rim Fire highlighted roads passing below or across slopes with similar characteristics.  The map showed that only about 77 kilometers of road within the burned area required a detailed assessment out of the 748 kilometers of road present.  Given the rapid assessment requirement of the BAER process, this was a valuable insight for focusing our efforts where they were most needed.

BAER assessment of large wildfires like the Stanislaus Complex and Rim fires requires bringing in additional specialists to supplement those from the local offices.  With more team members, including some who are not familiar with the local area, there is a greater-than-usual need to compare field notes to ensure a thorough assessment of existing threats.  On the Stanislaus Complex, this involved a lot of time at the BAER den peering at maps in small groups comparing information which we had recorded in our field notebooks.  This is not a particularly efficient exercise at the end of a 10 to 12-hour day in the field.  Dedicating time earlier in a day would be at the expense of time when field assessment could be done.

This process of sharing collected information among specialists working on the same problems and with those assigned to related hazard issues was facilitated on the Rim Fire by the use of electronic note-taking.  There were two major advantages to this use of electronic note-taking devices.  First, there was a higher confidence that an observation was correctly located.  Because these units have a built-in Global Positioning System (GPS), any note is automatically tied to its respective location.  This compensates for those team members being unfamiliar with the local area and provides for more precise and uniform location information for all observations.  Second, the note-taking devices can be uploaded to computers.  This enables the information to be e-mailed to other team members, incorporated into GIS products, and inserted into other documentation.  An interesting variant to this electronic note-taking involved the increased prevalence of smart phones among members of the team.  Information compiled daily by the planning unit of the firefighting organization is provided to all the firefighting units through shift plans.  QR coded icons on the cover page of each shift plan permitted downloading maps and other updated information needed for safe field operation.   This mechanism enabled BAER team members with smart phones to access this information as easily as the firefighting crews.

Figure 3 – Electronic note taking devices link field notes to GPS coordinates for later sharing and data manipulation. The desire for large, more easily read displays is always tempered by how big a device you want to carry in the field. Geologist Kellen Takanaka enters a note while geotechnical engineer Jonathan Berry looks for the next sample location.

Figure 4 – The cover for the Rim Fire shift plan issued on Friday, September 13, 2013. The QR icons provide connection to current versions of four maps covering the fire area.

It was sunny the morning of September 20^th with only the slightest hint of smoke still in the air.  I would soon start the two-hour drive to my home after eleven straight days of working 10 or more hours per day on the Rim Fire.  In the BAER den, I gathered some remaining gear, turned in forms and said my goodbyes with the realization that this would be my last BAER team assignment—retirement was only a matter of months in my future.  I looked around the BAER den before I walked out thinking about the changes since the Stanislaus Complex BAER team 26 years earlier.  Then we had struggled with having enough wall space to hang maps or tables to work on our reports and notes.  On this assignment, it had been a struggle to locate a convenient outlet to plug in your laptop computer or to find the password for the local WIFI connection.  Tables which would have been covered with piles of maps and papers in 1987 are now overflowing with laptops with a rat’s nest of cords underneath them.  As I closed the door and headed for my truck, I realized technology and science will continue to change for future BAER assessment.  But the people—dedicated specialists trying to implement good science in a hurry to avoid impending hazards—will continue to be the most important part of the assessment effort.

Figure 5 – The sign posted on the door to the room serving as the work area for the BAER team at one of the area motels.

Jerome De Graff is a native of upstate New York and received his Bachelor’s degree from The State University of New York at Geneseo. He began working for the Forest Service after receiving his Masters degree in Geology from Utah State University. For over 36 years, Jerry served as an environmental or engineering geologist on National Forests in Utah and California until his retirement in Feb. 2014. Jerry drew upon his participation in Burned Area Emergency Rehabilitation (BAER) teams from 1981 through 2013 for this blog post. In addition to being a geologist on BAER teams, he also has served as team leader on fires in California. Because of his experiences, the US Agency for International Development sent him as part of a 3-person team to interact with fellow professionals in Bulgaria and Greece in 2001. In 2002, he returned to Bulgaria as the team leader and co-instructor for conducting two BAER training sessions. He continues his geology career as an adjunct faculty member for California State University.