SW Oregon Douglas Fire Study

New study from Oregon State. From the press release:

Researchers in the College of Forestry at Oregon State University used satellite imagery and local data to analyze the factors driving differences of severity in the fire, which burned about 50,000 acres north of Grants Pass. Located in the Klamath Mountains ecoregion, the area is dominated by Douglas fir, ponderosa pine and white fir and is a mix of private and federal ownership and state-owned O&C (Oregon & California Railroad) lands.
While daily weather was the most significant driver of fire severity, the researchers found that other factors such as ownership, forest age and topography were also critical. Intensively managed private forestlands tended to burn with greater severity than older state and federal forests. The findings are important because they point to the need for collaboration among landowners, both public and private, to reduce the wildfire risk across the region. 
Some caveats: This was one fire in a unique region. The fires started in older federal forests and wind drove them across property lines.
And as the authors note, “There is strong scientific agreement that fire suppression has increased the probability of high severity fire in many fire-prone landscapes (Miller et al. 2009, Calkin et al. 2015, Reilly et al. 2017), and thinning as well as the reintroduction of fire as an ecosystem process are critical to reducing fire severity and promoting ecosystem resilience and adaptive capacity (Agee and Skinner 2005, Raymond and Peterson 2005, Earles et al. 2014, Krofcheck et al. 2017).”


24 thoughts on “SW Oregon Douglas Fire Study”

  1. I haven’t seen the references, but there seems to be conflicting statements here. In italics: “Intensively managed private forestlands tended to burn with greater severity than older public forests.” Below: “thinning is critical to reducing fire severity.” Seems more explanation is needed. Jim Bailey

    • Of course, if you pretend that the Forest Service still does lots of huge clearcuts, then there might be more concern needed. Here in California, the USFS hasn’t done clearcuts since 1993. Yes, we need much more thinning in the Sierra Nevada.

      • It is no surprise that a wind-driven crown fire can race through a plantation and kill every tree, thus the high fire severity. The same goes for natural openings, such as those created by wildfire or in “ecological forestry” projects designed to create early seral habitat — wherever “spatially homogenized fuels” are found — though the authors do not address this, and focus on “intensive plantation forestry.”

      • Larry, they are still clearcutting on private lands and those landowners are putting public resources at risk of severe fire. Combine this with the increased water demand and reduced summer stream flows caused by dense tree farms and …. so much for good neighbors.

          • If private lands used longer rotations it would have numerous co-benefits:
            * stands would spend a smaller fraction of their existence in the high-fuel-hazard configuration
            * increased landscape carbon storage
            * increased water quality
            * a smaller fraction of the landscape would be producing hydrologic extremes, i.e., peak flows (caused by clearcuts roads) and low summer flows (caused by dense tree farms)
            * enhanced habitat for a wider variety of terrestrial and aquatic species
            * improved scenic values

            My other suggestions would be to increase stream buffers, retain a few more trees and snags, and relax reforestation and “free to grow” requirements, which would reduce the need for chemical spraying and allow more diverse early seral conditions to persist.

        • “Putting public resources at risk”….. the private land owners pay for a higher level of protection then is provided for federal lands locally. With a 24-48 hour response for additional federal fire resources, there’s not a lot of question why fires spread from federal to private.
          Lack of fire suppression on Federal lands greatly contributed to the losses of privates resources, on this complex. This was the same problem this last season with Chetco Bar, spread off of Federal and destroyed private resources, accounting for more than half of the private acres burned in the State. This was from on the ground observation during both instances sited.
          Maybe we can have examples of private plantations putting federal resources at a higher risk?

    • The simple explanation is that this study compared fire effects in private ownerships dominated by regen logging and dense tree farms (which tends to have more high severity fire effects) versus public forests with more mature & old-growth forests and some aging plantations, both thinned and unthinned (which tends to have lower fire severity). There’s no contradiction here. I find it most interesting that weather is the dominant factor, followed by stand age, followed by topography, while “pre‐fire forest biomass were not an important predictor of fire severity.” This study turns on its head, the old notions about “time since fire” increasing fuel hazard. It’s the opposite!

      Finally, the effects of thinning are likely very complex with some tendencies toward increasing fire hazard and some tendencies toward reducing fire hazard, but agency planning almost entirely fails to recognize these nuances.

      • One “nuance” that some eco’s gloss over is the 100,000,000+ dead trees (in California), especially on un-managed public lands. Pretending that dead old growth in such numbers is perfectly ‘natural’ is not supported by forest densities and annual rainfall amounts.

      • That’s kind of a broad statement, 2nd. Having reviewed many fuel treatment EAs and EISs (and response to appeal points) I think most of the ones I’ve seen have been very cognizant of the pros and cons. That’s what fuel specialists do. If you have specific examples, we can take a look.

    • Yes, it’s important to be a good neighbor. How is that related to this?

      Also, thinking that you might be an academic, what do you think is a good way to give positive feedback on the way the paper is written. In FS world, or other venues, you might write to their boss, but that maybe doesn’t fit with academic culture?

      • The “Good Neighbor” reference relates to the fact that this study shows that private industrial forest owners maintain a hazardous fuel condition across a large portion of the holdings, thus putting public resources at risk. Too often the good neighbor language is used counterfactually as a club against public lands that are less actively managed.

        • You say that it is “counterfactual” but public lands are often not managed to help reduce fire risks. I would argue that this study does actually show what you stated ” private industrial land maintains a hazardous fuel condition across a large portion”.
          For one thing private industrial land varies a great deal in how it’s managed.
          if it’s an age thing to say it’s a large portion, you would have to know the acres in each age class (and really we are talking private land in SW Oregon, not private land more generally.
          That they “maintain a hazardous fuel condition” , based on this study, does not take into account the authors own limitations as well as the statements of others on this blog, which call into question exactly what can be said based on these data.

  2. I had a lot of questions about this study.. like “intensive management isn’t one thing.. it’s a variety of things”, and what did they measure to indicate “severity” and how can you tell that from satellites?. They have a whole section called limitations. .. here are just a couple, but I really like the way the authors deal with the kinds of questions most people would have about their research. Kudos to them!

    “Fourth and perhaps most fundamentally important, we relied on pre-fire 463 biomass and stand age as proxies for fuel, in part because Landsat and other passive optical 464 sensors have limited sensitivity to vertical and below-canopy vegetation structure (Lu, 2006).
    465 Accurate and spatially complete quantitative information of forest surface and canopy fuels were not available for the Douglas Complex. More broadly, there are significant limitations 466 to spatial 467 predictions of forest structure and fuels using GNN and other methods that rely on passive 468 optical imagery such as Landsat (Keane et al. 2001, Pierce et al. 2009, Zald et al. 2014), which is 469 why we relied on the more accurately predicted age and pre-fire biomass variables as proxies.
    470 Surface and ladder fuels are the most important contributors to fire behavior in general (Agee 471 and Skinner 2005), and surface fuels have been found to be positively correlated to fire severity 472 in plantations within the geographic vicinity of the Douglas Complex (Weatherspoon and 473 Skinner 1995). Yet correlations between biomass and fuel loadings can be highly variable due to 474 site conditions and disturbance history (i.e. mature forests with frequent surface fires may have
    475 high live biomass but low surface fuel loads, while dense young forests that have regenerated 476 after a stand replacing wildfire will have low live biomass but potentially high surface fuel 477 loading as branches and snags fall). Therefore, GNN predicted pre-fire biomass may represent 478 the total fuel loading, but not the available surface and ladder fuels that have the potential to burn 479 during a specific fire, and this is supported by the low importance of pre-fire biomass as a
    480 predictor of fire severity in our study. Furthermore, it is important to recognize that in addition 481 to total surface and ladder fuels, the spatial continuity of these fuels strongly influences fire 482 behavior (Rothermel 1972, Pimont et al. 2011). Fifth, while private industrial and BLM forests
    483 in our study area had very different forest conditions due to contrasting management regimes, 484 ownership alone misses management activities (e.g. site preparation, stocking density, competing 485 vegetation control, partial thinning, etc.) that can influence fuels and fire behavior.”

    And I really, really, really like where they compare their findings with other findings and hypothesize reasons for the differences:

    Management driven changes in fuel spatial patterns at tree and stand scales could also reconcile 417 differences in prior studies that have found increases (Thompson et al. 2007, Odion et al. 2004) 418 and decreases (Prichard and Kennedy 2014) in fire severity with intensive forest management.
    419 The two studies that observed an increase in fire severity with intensive forest management were 420 conducted in the Klamath ecoregion of Southwestern Oregon and northwestern California, the same ecoregion as this study. In contrast, Prichard and Kennedy (2014) 421 examined the Tripod 422 Complex in north-central Washington State, where harvests mostly occurred in low to mid – 423 elevation forests dominated by ponderosa pine, Douglas-fir, lodgepole pine (Pinus contorta var.
    424 latifolia), western larch (Larix occidentalis), and Engelmann spruce (Picea engelmannii). These 425 forests have lower productivity compared to those studied in the Klamath ecoregion, with more 426 open canopies and longer time periods to reach canopy closure after harvest, which likely results 427 in more heterogeneous within stand fuel spatial patterns. Furthermore, forest clearcut units were 428 relatively small in the Tripod Complex (mean 53 ha, Prichard and Kennedy 2014), and while 429 these harvest units were spatially clustered, they were not large contiguous blocks as found in the 430 O&C Lands. Lastly, it is unclear if the harvest units evaluated by Prichard and Kennedy (2014) 431 experienced the full distribution of fire weather or topographic conditions compared to 432 unharvested units, as our study does, which may confound their conclusions and our 433 understanding of the relative importance of the factors driving fire behavior and effects.

    I’m not sure I would call anything on the east side “intensive plantation forestry”. What seems to be going on here is that plantations of certain ages managed in certain ways, in certain sized blocks, (perhaps, depends on groundtruthing) had generally more severe burning in this fire. I bet that there are folks even on this blog familiar with that country, who could hypothesize why that might be the case.

    • There were several factors related to this complex and the severity of damage.
      This study, while acknowledging the inability to get good information under the canopy, tends to focus on and indicate that the fire severity was a combination of weather and plantations.
      The plantations were where the fire was able to be aggressively attacked and where retardant was most effective. The older stands in this area were characterized by dense understory, ranging from 10-40’ in height. Many areas had live oak in the understory, which readily torches out sending flames into the crowns. On more than one occasion during this complex, crews were chased out of these older stands due to crown fires caused by the ladder fuels. Successful fire breaks in these stands were 50-100’+ wide (visible on satellite) The fire moved quickly thru the plantations, resulting in the use of indirect attack and burn outs. While this is common on fed fires, it is not common in this area on private lands where we usually use direct attack backed by air support.
      Additionally, there were hundreds of acres “burned out” for fuel reduction credit on federal lands. These burn outs were burned off of contingency lines and were kept at low to moderate intensity. In several instances they were done after the initial line had been secured and mopped up 50-100’ in. This was not done on private land, which further adds to the misconception viewed from satellite imagery of severity on federal vs private plantations.

    • The study authors were pretty clear that since they had no way of directly measuring pre-fire fuels loads and fuel configurations, they used stand age (and ownership) as a proxy.

      • The study authors also were clear that they utilized July 7 of 2014 imagery to assess fire effects. What is not mentioned is that by September 1 of 2013, there were at least 15 separate logging operations underway on the private lands not only salvaging the dead merchantable timber but also clearing out marginally merchantable young stands & brush with as little as 25% mortality, deemed to be damaged to the extent they were not viable for longer term management. By July of 2014 several thousand acres of forestland had been cleared, some already replanted. In contrast the only tree removal on BLM lands was along roadsides with safety issues related to private lands management. Perhaps someone can explain how the imagery taken after this work was done could be a valid comparison with the neighboring un-treated BLM stands.

  3. 2ndLaw wants to talk about being a good neighbor. What is missing in the report as well as comments so far (Forester353 appears to know much about the Douglas Complex) is how the multiple fires started eventually converging into few but larger fires. I can’t speak to the southern part of the burn, that mostly on Medford BLM & private, but the northern portion, Roseburg BLM & private was the result of 26 individual lightning strikes out of several hundreds that night, causing ignitions. According to the immediate responders (company crews, Douglas Forest Protection & BLM staff) at first light, 25 of those 26 were on BLM ground. The private ground fire & some of the BLM fires were quickly controlled. However multiple roads into BLM stands had been decommissioned, making immediate direct attack impossible on other BLM fires. Due to heavy ground fuels, ladder fuels & highly defective old trees (many of the old stands in the Cow Creek area run over 50% defect) combined with restricted direct attack response those fires fueled the firestorm in the first hours of the fire. As Forester points out, the young plantations, which burn fast & consume most of the fuels, became the sacrificial ground. In contrast, fires burning in the old forest with 300+ tons of fuel per acre would die down late into the night, only to once again blow up as the heat, dry air & winds of mid afternoon fanned the flames.

  4. Thanks much, Forester 353 and Javier. A few more questions.. I looked up the definition of fire severity and found this..https://www.fs.fed.us/postfirevegcondition/documents/publications/keeley_ijwf_2009.pdf

    “Fire severity , and the related term burn severity , have created considerable confusion because of recent changes in their usage. Some authors have justified this by contending that fire severity is defined broadly as ecosystem impacts from fire and thus is open to individual interpretation. However, empirical studies have defined fire severity operationally as the loss of or change in organic matter aboveground and belowground, although the precise metric varies with management needs. Confusion arises because fire or burn severity is sometimes defined so that it also includes ecosystem responses. Ecosystem responses include soil erosion, vegetation regeneration, restoration of community structure, faunal recolonization, and a plethora of related response variables. Although some ecosystem responses are correlated with measures of fire or burn severity, many important ecosystem processes have either not been demonstrated to be predicted by severity indices or have been shown in some vegetation types to be unrelated to severity. This is a critical issue because fire or burn severity are readily measurable parameters, both on the ground and with remote sensing, yet ecosystem responses are of most interest to resource managers.”
    So (1) is this the current definition…?
    (2) How does the definition (I couldn’t find a specific definition in the paper, but I could have missed it) relate to how the authors measured it in this study “We quantified fire severity using the Relative differenced Normalized Burn Ratio (RdNBR), a satellite imagery based metric of pre- to post-fire change.”

    It seems like on the face of it, losing trees can be bad from the soil perspective (without roots, soil can wash downhill) but also frying the soil from burning with heavy surface fuel loadings could also be bad (even if some trees are left alive). Maybe this is not true in SW Oregon with different conditions and soils. Anyway, does measuring “fire severity” take both those potential negative impacts into account? Or is there even a standard way to define and measure fire severity?

    Also could someone post a photo of the two general kinds of stands, plantations and the older natural so those of us not from SW Oregon can get an image of what they are like? I last saw SW Oregon plantations 40 years ago when they were seedlings, and I’m sure they look different now.

  5. I visited the Douglas fire complex maybe a year after or so to look at purchasing some trees, but by then they were getting pretty wormy. This was on BLM ground. There was almost total mortality. These were 80 year old plus stands of timber. Whole watersheds were reduced to ground zero. I don’t think you could burn much more severely. I had several interesting discussions with people who were on the fire.
    One comment by someone who was on the fire when it started was, “we needed more resources and less bureaucracy.” Another comment by someone who spent weeks on the fire was that the leaders of the federal teams of fire fighters, were not responsive to the local knowledge or weather of the area and inflexible in their plans of operation. One person also who spent weeks on the fire commented how the fire fighters were boasting of their ability to lay down eight miles of burn out in a day. He also mentioned the use of dropping exploding ping pong balls from helicopters. He figured this was how we managed our forests, by fire.
    I also visited the Checto Bar fire this weekend. I would say the forest service has been successful in reducing the forest in this corner of Oregon by 2/3 with their fires each summer. It is now a much hotter, drier place.
    How can we consider the FS to be a good stewards of the land or a good neighbor when they let a 1/4 acre lighting fire turn into 190,000 acres wildfire.

    • On the GMUG they did a plan amendment on fire use.. I wonder whether SW Oregon forests did those?
      Seems to me that the key measurement should be whether the fire use was within prescription. I don’t think that acres burned, regardless of prescription, would be a particularly useful metric. Prescriptions (should) take into account all the things that should be considered in advance and by people who know the country.

      Fire Use Plan Amendment – 2007

      2007 Fire Use Forest Plan Amendment modifies fire management policies which will allow fire managers to manage certain lightning caused fires for resource benefits on the GMUG National Forest, if prescriptive conditions are met.

      Decision Notice and Finding of No Significant Impact- Fire Use Plan Amendment – January 2007 (PDF, 55 kb)

      Map – Considerations for Suitable Wildland Fire Use (PDF,673 kb)
      Map – Lands Generally Suitable for Wildland Fire Use (PDF, 626 kb)

      Environmental Assessment for Forest Plan Amendment – 2007 (PDF, 218 kb)
      Appendix A Specific Amendments for Fire Use (PDF, 92 kb)
      Appendix B Fire Managemnet Plans and Land Management Plans (PDF, 13 kb)
      Appendix C Amendment Siginificance (PDF, 26 kb)
      Appendix D Fire Use Planning Process (PDF, 16 kb)
      Appendix E Fire Regime Descriptions (PDF, 40 kb)


Leave a Comment