U.S. Forest Service Assesses Postfire Threats with GIS
Soil Burn Classification Maps Enhanced
By Kevin P. Corbley
Located just south of Riverside, California, the Cleveland National Forest covers 460,000 acres of rugged mountain terrain that is considered a haven for raptors and other wildlife threatened by ever-encroaching development. In October 2007, this pine and shrub forest was ravaged by four separate fires during California's now infamous month of wildfires.
Before the embers had cooled, the U.S. Forest Service sent a Burned Area Emergency Response (BAER) team to map and assess the threats and risks to human safety, property, and natural and cultural resources on or adjacent to the national forest lands that burned. The assessment teams are composed of personnel with backgrounds in hydrology, soils, geology, and other natural and cultural resource management disciplines. Through the assessment process, a determination is made whether the level of risk to these values constitutes an emergency and if effective treatments can be implemented to reduce the potential threats from occurring.
"Wildfire devastation rarely ends when the flames are extinguished," explains Alan Gallegos, a Forest Service geologist who was part of a BAER team deployed into the area of the Cleveland National Forest burned by the Santiago Fire. The charred soil and barren landscape remain vulnerable to further damage from flash floods, landslides, and debris flows. Resources and infrastructure spared from the inferno face a new risk of destruction for years afterward unless immediate steps are taken to remediate the burn area.
Flames Can Char the Ground to a Hard Crust, Leading to Unimpeded Runoff
The 28,476-acre Santiago wildfire was of particular concern to Forest Service officials because it contained decades-old evidence of postfire phenomena known as debris flows. In these situations, the flames strip the ground of vegetation and char it to the point where a hard crust forms. This hydrophobic soil condition prevents the absorption of rainwater, creating a surface ideal for the unimpeded flow of runoff. As the water races across the landscape, it erodes the soil and deposits sediment into stream channels.
"If the flow gets into a drainage channel under the right conditions, the sediment is mobilized and turns into a debris flow that quickly snowballs and enlarges in volume," says Gallegos. "It becomes a slurry of mud and boulders that moves down the channel with great force."
This momentous flow has the potential to cause substantial damage where the mixture of rocks and sediment spills out of the channel at the bottom of a slope. Homes, buildings, roads, and automobiles that lie in the path can be wiped out by the force of the mud and rocks.
Building a Burn Zone GIS
Recovery efforts must be initiated as soon as possible after a fire to stabilize the soils. A challenge for the BAER team is to assess soil conditions and determine which areas are most susceptible to debris flows and other runoff hazards so that emergency response plans can be prioritized. The team typically begins its assessment using a Burned Area Reflectance Classification (BARC) map created by Forest Service remote-sensing analysts using multispectral satellite imagery and other data.
To speed the Santiago Fire assessment process, Gallegos deployed digital photomapping software called GPS-Photo Link from GeoSpatial Experts, an Esri Business Partner in Thornton, Colorado. The software enabled the team to quickly and precisely correlate ground photos with burn classes in the BARC map to create a soil burn severity map. Gallegos chose the GPS-Photo Link software on recommendation from a Forest Service remote-sensing analyst.
Before arriving at the Cleveland National Forest, the BAER team tapped into an existing enterprise GIS and clipped digital files to create a mini-GIS in ArcGIS Desktop pertaining to the area impacted by the Santiago Fire, which included surrounding private property. Digital layers relating to the topography, cultural features, infrastructure, and soil geology were included in the GIS, along with the initial BARC layer, and loaded onto a laptop computer as a reference for the team members.
Using the BARC map as a guide, the team members began making field observations on foot to evaluate how the fire changed soil conditions and the response to future precipitation. They calculated how much vegetation had been consumed and measured the thickness of branches and trunks on the shrubs and trees that had survived. Soils were categorized based on the thickness of their charred surface, water repellency, and amount of ash coverage. These observations were compared to the BARC map to ground truth the soil burn severity classes created from the satellite imagery.
"For a fire covering a few thousand acres, the BAER team might make its entire field assessment on foot, but there just wasn't time to do that for an area the size of the Santiago Fire," says Gallegos, explaining that Forest Service regulations required fast evaluation of the burn area.
As is typical procedure for large fires, the BAER team continued its assessment of the Santiago burn area from a helicopter. Relying on its field observations as a baseline, the crews flew over the remaining area and validated or amended soil burn severity classes in the BARC map. To support these visual aerial assessments, Gallegos took digital photographs from the helicopter, usually from an altitude more than 500 feet above the ground.
For this project, Gallegos and his team used a standard Pentax digital camera and a handheld Garmin Mark V GPS receiver. So that each photo could later be correlated with its location in the ArcGIS Desktop soil burn severity layer, Gallegos snapped a photo of the burn area and set a GPS point for each photo point. During processing, this information would allow the photomapping software to synchronize the camera and the receiver so that photographic images could be correlated.
This simple method of synchronizing the digital photos and GPS pointsmade possible by the photomapping softwarerepresented a huge time savings in the office. Prior to using the software, digital photos and GPS points had to be matched manually by reviewing the time stamp for each dataset and labeling GPS points to match the digital photos.
"We took 400 photos of the Santiago Fire area and processed them using ArcGIS Desktop in one day," says Gallegos, adding that it would have taken days to manually associate each photo with its correct soil burn severity class in the map layer.
Processing of the photos occurred at the end of the flights. Back at the local Forest Service office, Gallegos downloaded the photographs from the digital camera and the track logs from the GPS receiver into a laptop computer containing the map layer or the project area. Using time synchronization, the GPS-Photo Link software running on the computer automatically matched each photo with its precise GPS coordinates and stamped each image with a digital watermark containing the time, date, and location of acquisition.
Next, the photomapping software exported the photos to ArcGIS Desktop and saved them as attributes in the database. The software then created a thumbnail icon representing the location of each photograph and placed it in its proper position on the soil burn severity map in ArcGIS. A member of the BAER team could simply click the on-screen icon from any map layer to access and view the photograph illustrating damage conditions at that point on the ground.
This greatly accelerated the final phase of editing and validating the soil burn severity map. The team members clicked through the photographs on the laptop computer and visually determined whether the evidence in the photo supported or contradicted the soil burn severity assigned to that area during BARC map classification. In many cases, the photos compelled the team to amend the classification to more accurately reflect conditions on the ground.
Use of the photomapping software accelerated the soil burn severity map validation process by a matter of days, according to Gallegos, and made the resultant map far more accurate in terms of its classification of damage.
Putting the Assessment to Work
Once the soil burn severity map layer was completed, it was overlaid on the other geospatial layers in ArcGIS Desktop. Forest Service personnel relied on this information to determine which areas within the burn zone faced the greatest threats from flooding and debris flow hazards. Areas with severely burned soil, no vegetation, and moderate to steep slopes were considered the most vulnerable. The existence of downslope buildings, structures, or roads added to their priority rating.
About the Author
Kevin Corbley is a freelance writer who specializes in remote sensing, digital mapping, and GIS (tel.: 540-667-2511, e-mail: firstname.lastname@example.org).
For more information, contact Alan Gallegos, geologist, U.S. Forest Service, Sierra National Forest, Clovis, California (e-mail: email@example.com; tel.: 559-297-0706, ext. 4862).