Okanogan County, Washington

Okanogan County in Washington is mostly rural and has about 43,000 residents—a population density of about 8 residents per square mile (Okanogan County, 2023). The county has 6 cities (including Okanogan), 7 towns (including Conconully and Winthrop), 5 census-designated places (including Methow), and several other unincorporated communities (including Mazama). Agriculture and forestry are the county’s two major economic industries, although popular outdoor recreation tourist destinations, such as Methow Valley, also provide economic benefits (Okanogan County, 2023). A Tribal reservation is near the southeast corner of the county, and is about 700,000 acres (Okanogan County, 2023).

The county has a history of major fires and associated postfire effects. In 1894, Conconully had a major postfire flash flood and debris flow (Wilma, 2006). In July 2014, the Carlton Complex Fire burned more than 250,000 acres in Okanogan County, and destroyed more than 300 homes (Ahearn, 2018). Okanogan County sustained cascading postfire hazard consequences after the Carlton Complex Fire, which included floods and mudslides that damaged homes and other infrastructure (King, 2014; Land, 2014).

Cedar Creek Fire

On July 8, 2021, lightning strikes started the Cedar Creek Fire in Methow Valley, Wash.; the fire burned more than 55,500 acres southwest of Winthrop and Mazama (Contreras and Mickelson, 2021a; Forest Service, 2021). Most of the Cedar Creek Fire burn area is managed by the U.S. Department of Agriculture Forest Service (FS; 97.2 percent of the burned area, or 53,718 acres), and smaller amounts of land are managed by the Washington State Department of Natural Resources (WADNR; 1.6 percent of the burned area, or 870 acres) or are privately owned (1.2 percent of the burned area, or 647 acres; Contreras and Mickelson, 2021a).

After the fire, the Washington Geological Survey Wildfire-Associated Landslide Emergency Response Team (WALERT) found 3 debris flows and 2 hyperconcentrated flows in and around the burned area. Two of the debris flows occurred in adjacent drainage basins and deposited sediment in the irrigation ditch that carries water to the southwest side of the Methow Valley. The other debris flow deposited about 14,000 square yards of gravelly sand and woody debris across an agricultural field (Contreras and Mickelson, 2021a).

Modeling indicated low to high debris-flow hazards across the Cedar Creek Fire burned area. The WALERT team built upon the USGS assessment to identify the infrastructure and property at risk from hydrologic hazards, including the State Route 20 corridor, alluvial fans between Mazama and Winthrop, and the Wolf Creek Reclamation District (Contreras and Mickelson, 2021a). The WALERT team suggested that authorities place signs to warn the public of debris flows and flooding in camping areas, and that land managers support irrigation district managers to protect critical infrastructure and notify transportation network managers of the potential issues of blocked culverts, sedimentation, and road erosion (Contreras and Mickelson, 2021a).

Muckamuck Fire

The Muckamuck Fire was similarly ignited by lightning on August 4, 2021, and burned 13,314 acres near Conconully and Conconully Lake, Washington, before reaching 100 percent containment on October 15, 2021. The Muckamuck Fire burned area covered a more diverse land-ownership region as compared with the Cedar Creek Fire. The 13,312 acre burn area is managed by several entities, including the FS (56 percent, or 7,448 acres), private owners (26.3 percent, or 3,510 acres), the Bureau of Land Management (10.2 percent, or 1,366 acres), the WADNR (6 percent, or 796 acres), the Bureau of Reclamation (1.2 percent, or 156 acres), and the U.S. Fish and Wildlife Service (0.3 percent, or 36 acres; Contreras and Mickelson, 2021b).

Regarding hydrologic postfire hazards, the USGS debris-flow assessment models found low to high debris-flow hazards in drainages throughout the Muckamuck Fire burn area (Contreras and Mickelson, 2021b). Drawing from these assessments, WALERT identified several areas where debris flows could affect communities, property, and infrastructure. These areas included recreation cabins and drainage areas around Salmon Creek, one home on an alluvial fan near Mineral Hill Road, the town of Conconully, and alluvial fans along the north shore of Conconully Lake (Contreras and Mickelson, 2021b). The WALERT report (Contreras and Mickelson, 2021b) contains recommendations for landowners and managers to prepare for postfire hazard consequences and notes a need to inform residents of the potential for postfire flash floods and debris-flow hazards if their homes are built on alluvial fans or adjacent to streams flowing from the burn area. The report also recommends signs to warn the public of these hydrologic hazards at camping areas and parking lots. The WALERT team recommends advising transportation network managers and landowners on potential road erosion, blocked culvert issues, and the need to inspect culverts that drain from the burn area because of the increased likelihood of sediment transport during hydrologic events (Contreras and Mickelson, 2021b).

In late July 2022, WALERT geologists identified 3 debris flows, 1 hyperconcentrated flow, and 3 flood events in and around the burn area (Contreras and Mickelson, 2021a). The debris flow and flood waters combined in North Fork Salmon Creek and led to additional flooding downstream in Conconully on July 4, 2022. These hydrologic cascading consequences destroyed one cabin and deposited gravelly sand and woody debris on the town’s roads and against building and bridge foundations. The Governor of Washington requested assistance from the U.S. Army Corps of Engineers after the North Fork Salmon Creek and nearby side channels were clogged with debris and sediment. The U.S. Army Corps of Engineers responded by dredging and building debris-catch structures to increase the channel capacity of North Fork Salmon Creek and build the town’s resilience to future debris-flow events (Contreras and Mickelson, 2021a).

Stakeholder Identification

Stakeholders were selected based on their expertise related to each of the PHIRE project scientific areas: cultural resources, vegetation recovery, water quality, water quantity, aquatic-habitat quality, sediment transport, and debris flow. Stakeholders also shared their insight into other context-specific groups they recommended be represented, which included local emergency management officials and citizen group representatives.

Selected stakeholders were prioritized to ensure at least one participant represented each combination of scientific focus area and scale of decision making. For example, only one participant may be chosen that has expertise in water-quality concerns at the State level or vegetation-recovery concerns in a specific national forest. Participants were first contacted by email to set up 30-minute, one-on-one conversations that screened for general interest and familiarity with the topics. Formal email invitations were sent to the 20 selected individuals after the one-on-one conversations. If a stakeholder declined workshop invitations, an invitation was extended to another stakeholder who represented the same area of expertise and scale of decision making. Participants were a combination of scientists and local stakeholders that included public works employees, emergency management employees, and individuals representing community groups. The following is a list of the organizations the 20 workshop participants represented:

Scenario Building

Similar to the SSG’s methods, the PHIRE social science research team consulted with scientific experts to identify potential consequences and interventions of the potential hazard scenario (Machlis and McNutt, 2010). The identified scenario (or set of scenarios) reflects the consequences that scientists are concerned about and how consequences interact to shape the overall system in a coupled natural–human systems model (Machlis and McNutt, 2010).

The scenario for the Postfire Hazards Chains of Consequences Workshop was based on consultation with the USGS PHIRE project scientists and their expertise on consequences and the parameters affecting those consequences. Through initial one-on-one meetings, email correspondence, and a large collaborative brainstorming meeting, these scientists from many disciplines shared ideas for scenarios that would capture the wide range of potential consequences of concern in their fields. At the end of the collaborative meeting, the scientists agreed on the broad, yet probable, scenario of a high-intensity rainstorm following a period of drought in north-central Washington. Several scientists also suggested additional layers to this scenario, including a reburn of the 2021 Cedar Creek and Muckamuck Fire burn areas or another future fire burning a different area that affects the same watersheds as the 2021 fires.

The social science team originally planned two scenarios: (1) a high-intensity rainstorm affecting the burn area and (2) a subsequent fire affecting the same watershed and leading to partial reburn of the area. Based on feedback and discussion during the workshop, participants wanted to differentiate between short- and long-term consequences before discussing interventions on day two; therefore, the team adapted the original scenarios to reflect participant interest: (1) a high-intensity rainstorm event and (2) the longer-term (that is, more than 5 years after the rainstorm) consequences of the first scenario. Instead of considering the consequences of another fire in the area (that is, a full or partial reburn in the Muckamuck or Cedar Creek Fire areas), participants were prompted to consider the long-term postfire hazard consequences after the Muckamuck and Cedar Creek Fires (that is, consequences occurring more than 5 years after the initial high-intensity rainstorm scenario).

Identifying Consequences and Interventions

Participants began by brainstorming the first-order or top-level consequences that would begin the consequence chains in the presented scenario. The first-order consequences are broad and might include effects on people and their communities or on soil. Then, participants collectively brainstormed the cascading consequences during a virtual group discussion. Facilitators prompted participants to consider the widest range of possible consequences and the multidisciplinary cascading consequences across different consequence chains. Using the virtual group management software, participants then voted for the 10 consequences that were the most important to brainstorm interventions for. Each participant was given the opportunity to vote for as many as five consequences that they wanted to discuss interventions for. Once the 10 consequences that had the highest number of votes were selected, participants were prompted to brainstorm interventions individually. Participants were provided with definitions of the intervention description, effects, considerations, and two examples that illustrate the intervention identification process before they began (table 1). For each intervention, participants noted its description, effects, and any other considerations.

Applying the Chains of Consequences Method to the Postfire Hazard Context

Applying the chains of consequences method had short-term and potentially long-term effects on participants and their organizations by facilitating cross-boundary connections. This type of cross-boundary collaboration has been supported by conservation and environmental science literature because of the complex nature of environmental and social problems (Perz and others, 2010).

How participants chose to structure the consequence chains reflected their various disciplinary perspectives. At the beginning of the workshop, participants structured the consequence chains in an unexpected way. Based on previous chains of consequences workshop results, the workshop facilitators expected the participants to structure the chains using first-order consequences that reflected specific physical outcomes of the scenario that then precipitate into a series of consequences. However, the workshop participants decided to instead organize around effect categories, which still led to a rich discussion of the many consequences of the scenario and how the consequences in one category led to cascading chains across other categories. This decision demonstrates the flexibility of this workshop method to meet the needs and perspectives of stakeholder groups while still eliciting information of interest.

The PHIRE workshop applied the chain of consequence method to longer-term scenarios specific to postfire hazard consequences and intervention planning. Although this application differs from prior SSG efforts (DOI SSG, 2013, 20196; Hopkins and others, 2018), participants emphasized the value of discussing complex postfire hazard consequences and interventions during the workshop. When asked for feedback at the end of the workshop, one participant said that the workshop “increased [the participant’s] understanding with areas outside [the participant’s] expertise.” Other participants shared that the workshop was a “great way to interact with different agencies [and] disciplines” and that the “idea sharing was great.”

Mapping Cascading Consequences

Participants focused on and began discussing the first-order consequence categories—effects on the area’s community, hydrology, vegetation, and soil, which they defined through discussion as “fundamental issues for postfire.” Participants went on to brainstorm how the cascading consequences included multiple disciplines. This shift indicates the inherent interdisciplinary nature of these consequences and the stakeholders’ awareness of this interdisciplinarity.

In the consequence chains and discussion, participants focused primarily on human and community effects like personal safety and transportation access issues. Participants considered loss of life a potential cascading consequence of hydrologic events, specifically debris flow and flooding. They were also concerned with the loss of roads and other transportation issues caused by these hydrologic events. This focus on safety and related concerns is perhaps notable given that stakeholders were selected because of their interests and expertise in a variety of fields, including vegetation recovery, water quality, and cultural resources. These selections suggest that an emphasis on the prioritization of human life and safety concerns is deeply ingrained in the perspectives of professionals working in many disciplines associated with postfire hazards. This may also be because most participants live near or have other personal ties to the Okanogan County area.

When discussing community effects, a participant explained how agencies “did a great job communicating risk” around the July 4, 2022, debris-flow event, which provided communities with adequate warning ahead of time. Yet, despite this warning, some stakeholders explained the challenges of hazard mitigation with limited funds. They also emphasized the long-term mental health effects of these fires and postfire hazards, including the stress communities and individual residents face because they know what happened during the previous debris flows, and many others live in alluvial fans below the fire burn areas. For example, one participant explained that they have “whole communities and individual residences that, for the next ten years or more, as soon as they get raindrops on their house, their anxiety goes up because they know what happens upstream of them.” In other words, participants were concerned about the increased anxiety during every rainstorm following the fires since the July 4, 2022, debris flow.

In this discussion, participants also emphasized the potential loss of trust communities have in science information and the people who communicate the information as major problems. Participants also asserted that agencies must adapt to address these trust issues. Workshop participants tasked with issuing warnings also explained how they often grapple with informing the public while also ensuring communities do not tune out the warnings if they receive too many. Participants agreed that the main necessity to reduce loss of life and improve warning systems was more observational data. For example, items like rain gages could capture real-time data, which would give the communities specific information about how debris flows could affect them personally.

Using the consequence chains, participants demonstrated how vegetation and soil effects were considerably intertwined. They discussed seeds having difficulty establishing when there is no soil for them to grow in. Discussion also included the effects established vegetation has on stabilizing soil and reducing erosion. On the consequence chains, participants detailed how the loss of terrestrial habitat could cascade into effects on culturally important plant species and the loss of winter habitat for key game species. This loss could lead to reduced hunting opportunities (and revenue loss from fewer hunters), high costs from replanting key vegetation species, and diminished returns from traditional foods like Vaccinium L. (huckleberries).

Participants noted how debris flow cascaded into changes in (1) channel morphology, (2) channel scour, (3) stream confinement, and (4) loss of habitat connectivity. Debris flow also cascaded into a chain of consequences, including loss of life and blocking bridges and culverts, which could pond water, trigger more debris flows, and eventually cause flooding of roadways. Participants representing the town of Conconully and participants representing Okanogan County elaborated on the community effects of debris flow, specifically how irrigation systems and infrastructure have been affected since their July 4, 2022, debris flow.

Considering Interventions

The 10 consequences participants selected to identify interventions for focused primarily on human or community effects. Some participants were more concerned than others and made statements like, “I think we don’t do things soon enough to protect life and property that could be done.” Other participants expressed the opposite concern, and one participant explained that they “were struggling to connect to the consequence” because “loss of life from debris flow isn’t really something they hear about a lot.” Most participants discussed interventions for these human effects to limit the loss of life, damage to infrastructure, and impaired road access. Participants emphasized the need for longer-term (20 or more years) zoning and planning and fuel management, like prescribed burning, to effectively reduce loss of life, infrastructure, and property.

One participant emphasized the work that the NWS does to monitor weather, predict storms, and report hazards to communities (NWS, 2024). Discussing the NWS work started a larger discussion about the desire for increased streamgages and methods of getting information out to remote communities without cellular service. Rain gages were a primary concern among workshop participants. One participant stated that they “want to reiterate, getting more observation data is key,” elaborating, “Specifically, rain gages would be a feasible option to get more real-time data and so the public can see and use that on a personal level, so they can make the decisions they need to in the case of an emergency developing.” Another participant added to the discussion by suggesting, “I think it would be great to see, in addition to the State-level funding of rain gages, USGS work with the water science centers to fund [the] installation of rain gages and streamgages after fires and potentially [find an] opportunity to put aside funding to do that . . . [that was done] with great success in New Mexico and in the past it has been done in Washington with Natural Resources Conservation Service funding. Mission assignments and [the] Federal Emergency Management Agency can get involved as well. It is a quick way to get high-quality data to inform NWS and USGS research.”

When discussing the challenges of warning communities, one participant explained, “Getting the information out to the audience is web based these days and so many of the areas that we serve are off-grid and don’t have access to the internet. That’s one of the biggest challenges.” For remote communities, it is critical to have appropriate warning systems and communication to alert residents about impending hazards in advance.

To intervene in the consequence of washed-out roads, participants discussed the feasibility of soil stabilizers and designing future transportation systems that consider hydrological systems. One participant emphasized, “It's about scoping when it comes to feasibility. It’s spending your money where it matters. Scoping and focusing in on the areas through design and some kind of decision-maker working together.” Although disasters, such as fires and postfire hazards, offer potential opportunities for policy change to improve land-use planning and mitigation, studies have not found evidence of sufficient reinvestment in hazard-prone environments with government support after wildfires (Mockrin and others, 2016; Kramer and others, 2021).

To intervene in debris-flow hazards, participants suggested the installation of beaver dam analogs and other instream features. Other participants noted that this intervention is challenging because it involves intentionally adding debris to a stream, which potentially increases liability for downstream harm (such as modifying river flow and water availability). Several participants also discussed how vacating areas while natural recovery occurs could make hazards less disastrous. However, this option is not often possible because State highways and vital roadways often need to stay open, which requires repairs. They also suggested identifying areas at risk of hazards before rain events and establishing emergency response plans as quickly as possible.

Participants discussed other unintended consequences of potential interventions. For example, after suggesting mulching and chipping in burn areas as an easy and low-cost intervention to reduce loss of infrastructure and property, several participants explained how mulching can have unintended consequences when heavy rainfall washes it off and plugs culverts. Although this intervention helps reduce erosion by dispersing water droplets from rainfall, mulching could potentially contribute to more debris flows during extreme rainfall.

Reflecting on the Use and Applicability of U.S. Geological Survey Resources

Many reports discuss the need for advanced tools, information, and assessments to support postfire hazard decision making (for example, Federal Geographic Data Committee, 2019; Ekarius and others, 2020; Wildland Fire Mitigation and Management Commission, 2023). The Wildland Fire Leadership Council emphasizes the need for scientists to work with Federal, State, local, Tribal, and nongovernmental groups to jointly define the data and tool requirements and the strategic information needed for decision making (Wildland Fire Leadership Council, 2023).

The USGS has diverse capabilities across scientific disciplines and provides scientific data, tools, and resources to support postfire-hazard decision making. For example, the USGS can characterize burn severity by studying biomass consumed below and aboveground before and after a fire. Studying biomass is critical for postfire debris-flow and vegetation-trajectory likelihood assessments and Burn Area Emergency Response (BAER) plans postfire (Eidenshink and others, 2007). The USGS Landslide Hazard Program draws from this characterization of burn severity to assess and monitor postfire debris-flow risk. State water science centers also conduct postfire water-quantity and -quality monitoring and risk assessments for flooding, sedimentation, contaminants, and potential mitigation opportunities (USGS, 2023). This work involves deploying temporary water gages, sampling water quality, and advising on postfire mitigation options from a hydrology perspective (USGS, 2018). The USGS also provides geospatial and remote-sensing data through the National Geospatial Program (https://www.usgs.gov/programs/national-geospatial-program), National Land Imaging Program (https://www.usgs.gov/programs/national-land-imaging-program), and Earth Resources Observation and Science Center (https://www.usgs.gov/earth-resources-observation-and-science-eros-center). Data are also available at the interagency collaborative resource LANDFIRE-Landscape Fire and Resource Management Planning website (https://www.usgs.gov/mission-areas/core-science-systems/science/landfire-landscape-fire-and-resource-management-planning#overview). Specifically, LANDFIRE supports efforts to determine change in wildland fire risk by remapping remaining fuels (LANDFIRE, 2023). Researchers at the USGS also advise Burned Area Emergency Response teams on re-establishing desirable plant species and cover while minimizing fire-adapted invasive species postfire.

Throughout the Postfire Hazards Chains of Consequences Workshop, participants discussed available USGS resources for their postfire hazard decision making. Participants emphasized the tools, information, and assessments still needed to support their understanding of, and preparedness for, the cascading consequences of postfire hazards. For example, many participants discussed how the USGS Landslide Hazards Program Emergency Assessments of Postfire Debris Flow Hazards maps inform their postfire hazard decision making (https://usgs.maps.arcgis.com/apps/dashboards/c09fa874362e48a9afe79432f2efe6fe). After a fire, the USGS Landslide Hazards Program develops an Emergency Assessment of Postfire Debris-Flow Hazards. The Postfire Debris-Flow Hazards maps show the response of potential mountain front debris flow likelihood and volume after a simulated storm, but they do not identify or assess runout paths, inundation areas, or potential damage (https://www.usgs.gov/programs/landslide-hazards/science/emergency-assessment-post-fire-debris-flow-hazards). The NWS bases debris flow warnings and watches on these assessments. The Washington Geological Survey WALERT builds off the USGS debris-flow assessments to evaluate potential risk posed by landslides and debris flows (WADNR, 2023).

Adding to the discussion on rain gages (refer to the “Considering Interventions" section), participants also expressed the need for more models based on data specific to their region in Washington. Without accurate models for Washington, there is uncertainty about rainfall thresholds triggering debris flows and how long the debris-flow hazard exists.