The combined impact of increasing wildfire, insect outbreaks, and tree diseases are already causing widespread tree die-off and are virtually certain to cause additional forest mortality by the 2040s and long-term transformation of forest landscapes. Under higher emissions scenarios, extensive conversion of subalpine forests to other forest types is projected by the 2080s.

This finding is from chapter 21 of Climate Change Impacts in the United States: The Third National Climate Assessment.

Process for developing key messages: The authors and several dozen collaborators undertook a risk evaluation of the impacts of climate change in the Northwest that informed the development of the four key messages in this chapter (see also Ch. 26: Decision Support). This process considered the combination of impact likelihood and the consequences for the region’s economy, infrastructure, natural systems, human health, and the economically-important and climate sensitive regional agriculture sector (see Dalton et al. 2013a2135da9-c8b1-486f-9656-59d8a52b1975 for details). The qualitative comparative risk assessment underlying the key messages in the Northwest chapter was informed by the Northwest Regional Climate Risk Framing workshop (December 2, 2011, in Portland, OR). The workshop brought together stakeholders and scientists from a cross-section of sectors and jurisdictions within the region to discuss and rank the likelihood and consequences for key climate risks facing the Northwest region and previously identified in the Oregon Climate Change Adaptation Framework.7450bfd8-54cc-4c42-8058-7ae93f7692a5 The approach consisted of an initial qualitative likelihood assessment based on expert judgment and consequence ratings based on the conclusions of a group of experts and assessed for four categories: human health, economy, infrastructure, and natural systems.429802a3-633d-447c-874c-250ae4ee0003 This initial risk exercise was continued by the lead author team of the Northwest chapter, resulting in several white papers that were 1) condensed and synthesized into the Northwest chapter, and 2) expanded into a book-length report on Northwest impacts.a2135da9-c8b1-486f-9656-59d8a52b1975 The NCA Northwest chapter author team engaged in multiple technical discussions via regular teleconferences and two all-day meetings. These included careful review of the foundational technical input report429802a3-633d-447c-874c-250ae4ee0003 and approximately 80 additional technical inputs provided to the NCA by the public, as well additional published literature. They also drew heavily from two state climate assessment reports.219520b8-3d2e-40bf-8c39-fb51ded544d8 2ac1bce9-7e8e-41f5-a3ed-617646370b8c The author team identified potential regional impacts by 1) working forward from drivers of regional climate impacts (for example, changes in temperature, precipitation, sea level, ocean chemistry, and storms), and 2) working backward from affected regional sectors (for example, agriculture, natural systems, and energy). The team identified and ranked the relative consequences of each impact for the region’s economy, infrastructure, natural systems, and the health of Northwest residents. The likelihood of each impact was also qualitatively ranked, allowing identification of the impacts posing the highest risk, that is, likelihood × consequence, to the region as a whole. The key regionally consequential risks thus identified are those deriving from projected changes in streamflow timing (in particular, warming-related impacts in watersheds where snowmelt is an important contributor to flow); coastal consequences of the combined impact of sea level rise and other climate-related drivers; and changes in Northwest forest ecosystems. The Northwest chapter therefore focuses on the implications of these risks for Northwest water resources, key aquatic species, coastal systems, and forest ecosystems, as well as climate impacts on the regionally important, climate-sensitive agricultural sector. Each author produced a white paper synthesizing the findings in his/her sectoral area, and a number of key messages pertaining to climate impacts in that area. These syntheses were followed by expert deliberation of draft key messages by the authors wherein each key message was defended before the entire author team before this key message was selected for inclusion in the report. These discussions were supported by targeted consultation with additional experts by the lead author of each message, and they were based on criteria that help define “key vulnerabilities,” including likelihood of climate change and relative magnitude of its consequences for the region as a whole, including consequences for the region’s economy, human health, ecosystems, and infrastructure.429802a3-633d-447c-874c-250ae4ee0003 Though the risks evaluated were aggregated over the whole region, it was recognized that impacts, risks, and appropriate adaptive responses vary significantly in local settings. For all sectors, the focus on risks of importance to the region’s overall economy, ecology, built environment, and health is complemented, where space allows, by discussion of the local specificity of climate impacts, vulnerabilities and adaptive responses that results from the heterogeneity of Northwest physical conditions, ecosystems, human institutions and patterns of resource use.

Description of evidence base: Evidence that the area burned by fire has been high, relative to earlier in the century, since at least the 1980s is strong. Peer-reviewed papers based on federal fire databases (for example, National Interagency Fire Management Integrated Database (NIFMID)1970/1980-2011) and independent satellite data (Monitoring Trends in Burn Severity (MTBS), 1984-2011) indicate increases in area burned.93ae895c-ab04-4801-bb47-ac963c7311c1 4d24a997-855e-44e0-9693-2895851d9144 Evidence that the interannual variation in area burned was at least partially controlled by climate during the period 1980-2010 is also strong. Statistical analysis has shown that increased temperature (related to increased potential evapotranspiration, relative humidity, and longer fire seasons) and decreased precipitation (related to decreased actual evapotranspiration, decreased spring snowpack, and longer fire seasons) are moderate to strong (depending on forest type) correlates to the area and number of fires in the Pacific Northwest. Projections of area burned with climate change are documented in peer-reviewed literature, and different approaches (statistical modeling and dynamic global vegetation modeling) agree on the order of magnitude of those changes for Pacific Northwest forests, though the degree of increase depends on the climate change scenario and modeling approach. Evidence from aerial disease and detection surveys jointly coordinated by the U.S. Forest Service and state level governments supports the statement that the area of forest mortality caused by insect outbreaks (including the mountain pine beetle) and by tree diseases is increasing. Evidence that mountain pine beetle and spruce bark beetle outbreaks are climatically controlled is from a combination of laboratory experiments and mathematical modeling reported in peer-reviewed literature. Peer-reviewed future projections of climate have been used to develop projections of mountain pine beetle and spruce beetle habitat suitability based on these models, and show increases in the area of climatically suitable habitat (particularly at mid- to high elevations) by the mid-21st century, but subsequent (late 21st century) declines in suitable habitat, particularly at low- to mid-elevation. There is considerable spatial variability in the patterns of climatically suitable habitat. Evidence for long-term changes in the distribution of vegetation types and tree species comes from statistical species models, dynamic vegetation models, and other approaches and uses the correlation between observed climate and observed vegetation distributions to model future climatic suitability. These models agree broadly in their conclusions that future climates will be unsuitable for historically present species over significant areas of their ranges and that broader vegetation types will likely change, but the details depend greatly on climate change scenario, location within the region, and forest type. Evidence that subalpine forests are likely to undergo almost complete conversion to other vegetation types is moderately strong (relatively few studies, but good agreement) and comes from dynamic global vegetation models that include climate, statistical models that relate climate and biome distribution, and individual statistical species distribution models based on climatic variables. The fact that these three different approaches generally agree about the large decrease in area of subalpine forests despite different assumptions, degrees of “mechanistic” simulation, and levels of ecological hierarchy justifies the key message.

New information and remaining uncertainties: The key uncertainties are primarily the timing and magnitude of future projected changes in forests, rather than the direction (sign) of changes. The rate of expected change is affected by the rate of climate change – higher emissions scenarios have higher impacts earlier in studies that consider multiple scenarios. Most impacts analyses reported in the literature and synthesized here use emissions scenario A1B or A2. Projections of changes in the proportion of Northwest pine forests where mountain pine beetles are likeliest to survive and of potential conversion of subalpine forests used scenario A2. Statistical fire models do not include changes in vegetation that occur in the 21st century due to disturbance (such as fire, insects, and tree diseases) and other factors such as land-use change and fire suppression changes. As conditions depart from the period used for model training, projections of future fire become more uncertain, and by the latter 21st century (beyond about the 2060s to 2080s), statistical models may over-predict area burned. Despite this uncertainty, the projections from statistical models are broadly similar to those from dynamic global vegetation models (DGVMs), which explicitly simulate changes in future vegetation. A key difference is for forest ecosystems where fire has been rare since the mid 20th century, such as the Olympic Mountains and Oregon coast range, and statistical models are comparatively weak. In these systems, statistical fire models likely underestimate the future area burned, whereas DGVMs may capably simulate future events that are outside the range of the statistical model’s capability. In any case, an increase in forest area burned is nearly ubiquitous in these studies regardless of method, but the amount of increase and the degree to which it varies with forest type is less certain. However, fire risk in any particular location or at any particular time is beyond the capability of current model projections. In addition, the statistical model approaches to future fire cannot address fundamental changes in fire behavior due to novel extreme weather patterns, so conclusions about changes in fire severity are not necessarily warranted. Only a few insects have had sufficient study to understand their climatic linkages, and future insect outbreak damage from other insects, currently unstudied, could increase the estimate of future areas of forest mortality due to insects. Fire-insect interactions and diseases are poorly studied – the actual effects on future landscapes could be greater if diseases and interactions were considered more explicitly. For subalpine forests, what those forests become instead of subalpine forests is highly uncertain – different climate models used to drive the same dynamic global vegetation model agree about loss of subalpine forests, but disagree about what will replace them. In addition, statistical approaches that consider biome level and species level responses without the ecological process detail of DGVMs show similar losses, but do not agree on responses, which depend on climate scenarios. Because these statistical models simulate neither the regeneration of seedlings nor the role of disturbances, the future state of the system is merely correlative and based on the statistical relationship between climate and historical forest distribution.

Assessment of confidence based on evidence: The observed effects of climate on fires and insects combined with the agreement of future projections across modeling efforts warrants very high confidence that increased disturbance will increase forest mortality due to area burned by fire, and increases in insect outbreaks also have very high confidence until at least the 2040s in the Northwest. The timing and nature of the rates and the sources of mortality may change, but current estimates may be conservative for insect outbreaks due to the unstudied impacts of other insects. But in any case, the rate of projected forest disturbance suggests that changes will be driven by disturbance more than by gradual changes in forest cover or species composition. After mid-21st century, uncertainty about the interactions between disturbances and landscape response limits confidence to high because total area disturbed could begin to decline as most of the landscape becomes outside the range of historical conditions. The fact that different modeling approaches using a wide variety of climate scenarios indicate similar losses of subalpine forests justifies high confidence; however, comparatively little research that simulates ecological processes of both disturbance and regeneration as a function of climate, so there is low confidence on what will replace them.

References :

• Wildland Fire Summary and Statistics Annual Report 2011 (4d24a997)
• webpage National Monitoring Trends in Burn Severity (MTBS) Burned Area Boundaries Dataset (93ae895c)
• Climate Change in the Northwest: Implications for Our Landscapes, Waters, And Communities (a2135da9)
• Oregon Climate Assessment Report (2ac1bce9)
• A Workshop in Risk-Based Framing of Climate Impacts in the Northwest: Implementing the National Climate Assessment Risk-Based Approach (429802a3)
• The Washington Climate Change Impacts Assessment: Evaluating Washington's Future in a Changing Climate (219520b8)
• The Oregon Climate Change Adaptation Framework (7450bfd8)