Forecasts and Planning Tools

Climate Change Scenarios

Introduction

Although climate can vary naturally and will continue to do so in the future, human inputs of greenhouse gases are almost certain to cause continued warming of the planet. This warming has potentially significant implications for the Pacific Northwest (PNW) that warrant consideration in resource planning and management.

Estimates of future carbon dioxide (CO2) concentrations range from a doubling of pre-industrial values to an increase by a factor of 3.5 by 2100. Using these projections of future greenhouse gases, numerous research centers around the world employ numerical models of Earth's climate system to project future global climate. The Climate Impacts Group (CIG) recently examined a select subset of these global simulation models and produced updated scenarios of future climate for the PNW (see Table 1). A summary of the new scenarios is provided below.

Future Northwest Climate

Temperature

All scenarios evaluated by the CIG project a warmer climate for the PNW in the 21st century. More specifically:

• Climate models project a future rate of warming of approximately 0.5ºF (0.3ºC) per decade for the PNW through at least 2050 with average warming of 1.9ºF (1.1ºC) by the 2020s and 2.9 ºF (1.6ºC) by the 2040s, relative to 1970-1999 average temperature (for full range, see Table 1). For comparison, observed warming in the PNW during the 20th century was approximately 0.2ºF (0.1ºC) per decade.
• Temperatures are projected to increase across all seasons, although most models project the largest temperature increases in summer (June-August). This result stands in contrast to previous research findings at the CIG and elsewhere that winter warming would exceed summer warming.
• The warming projected for the PNW for the 21st century is substantially greater than the 1.5ºF (0.8ºC) increase in average temperature observed in the PNW during the 20th century. Additionally, average temperature could increase beyond the year-to-year variability observed in the PNW during the 20th century as early as the 2020s (Figure 1). This means that species or systems that respond primarily to changes in temperature are likely to continually face new conditions as a result of climate change.

Table 1: Changes in PNW climate from ten climate models for the 2020s and 2040s. All changes are benchmarked to average temperature and precipitation for 1970-1999. (More summary data)

	Temperature change			Precipitation Change
	Annual	Oct-Mar	Apr-Sept	Annual	Oct-Mar	Apr-Sept
2020s
Low	+ 0.7ºF (0.4ºC)	+ 0.4ºF (0.2ºC)	+ 0.8ºF (0.5ºC)	- 4%	- 3%	- 12%
Average	+ 1.9ºF (1.1ºC)	+ 1.7ºF (0.9ºC)	+ 2.1ºF (1.2ºC)	+ 2%	+ 4%	- 2%
High	+ 3.2ºF (1.8ºC)	+ 2.6ºF (1.5ºC)	+ 3.8ºF (2.1ºC)	+ 7%	+ 12%	+ 5%
2040s
Low	+ 1.4ºF (0.8ºC)	+ 1.1ºF (0.6ºC)	+ 1.4ºF (0.8ºC)	- 4%	- 1%	- 14%
Average	+ 2.9ºF (1.6ºC)	+ 2.5ºF (1.4ºC)	+ 3.3ºF (1.8ºC)	+ 2%	+ 5%	- 4%
High	+ 4.6ºF (2.6ºC)	+ 4.1ºF (2.3ºC)	+ 5.4ºF (3.0ºC)	+ 9%	+ 17%	+ 6%

Precipitation

Modest changes in regional precipitation are expected through mid-century, although changes in precipitation are less certain than changes in temperature. More specifically:

• The majority of models project slight decreases in summer (June-August) precipitation and slight increases in winter (December-February) precipitation, but little change (0-10%) in the annual mean by mid-century.
• Winter precipitation changes are largest in December-February, but still within the range of year-to-year variability observed during the 20th century (Figure 1).
• A larger percentage of overall winter precipitation is expected to fall as rain rather than snow due to warmer winter temperatures.
• Any changes in summer precipitation will be small given how little rain the region currently receives during summer.

Some models project increases in precipitation while others project decreases (Table 1). The divergence in model projections results from the fact that precipitation is affected by complex yet sometimes subtle changes in large-scale atmospheric circulation patterns which, in turn, are influenced by many imperfectly understood processes (e.g., ocean currents, tropical circulation, interactions between vegetation and the atmosphere).

It is also important to note that natural year-to-year and decade-to-decade fluctuations in precipitation are likely to be more noticeable than longer term trends associated with climate change. Thus, species or systems that respond primarily to changes in precipitation are likely to have already experienced the range of variability expected in the 21st century. Systems that are tuned to precipitation and temperature, however, are likely to find the conditions of the 21st century different from what they have previously experienced.

How will climate change affect particular climate events in the PNW?

Because many key aspects of climate (e.g., windstorms, heat waves) are not well simulated by models, the CIG cannot speculate how they may change in the future. However, droughts may become more common due to the effects of warmer temperatures and reduced winter snowpack on late summer streamflows.

How will ENSO and PDO be affected by projected climate change?

Changes in the behavior of climate patterns like the El Niño/Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO), and the Arctic Oscillation (AO) are projected by most models, but models typically aren't able to reproduce the variance observed in those climate patterns during the 20th century. As a result, there is no conclusive evidence as to how climate patterns such as ENSO, PDO, and the AO may change in the future.

Comparing the 2005 Scenarios with Previous PNW Scenarios

How do the 2005 climate change scenarios differ from the CIG's previous scenarios? The new climate change scenarios released in October 2005 show smaller temperature increases through mid-century than previous PNW climate change scenarios (Table 2). Precipitation changes are generally also smaller for this period. These differences are largely due to the following:

• Consideration of more models and climate scenarios. The CIG's previous climate change projections were derived using eight models forced by one greenhouse gas and sulfate aerosol emission scenario that assumed a 1% per year growth in CO2 emissions through the 21st century, a rate notably higher than the observed rate of 0.4% per year. The new climate change scenarios are derived from an evaluation of ten global climate models driven by two scenarios (A2 and B1) of future greenhouse gas and sulfate aerosol concentrations.
• Improved baseline for comparing future change. Previously, the baseline to which future changes were compared was calculated from long "control" climate simulations in which CO2 was held fixed. Each modeling group chose a different control value of CO2, which meant that the baselines included different degrees of warming. The new method for establishing the baseline calculates changes relative to the average climate of 1970-2000 as simulated by the climate models.
• Using 30 year means for comparing change. Precipitation changes were previously reported as ten-year averages, which meant that substantial decade-to-decade variability in simulated precipitation was included in the estimated precipitation change. Using 30 year means sifts through the variability to focus on longer-term change. Consequently, the projected changes in precipitation are smaller for the 2020s than previously reported.

The overall implication of the 2005 scenarios is that temperature changes previously projected for the 2020s and 2040s will likely occur a decade or more later in the 21st century than previously projected. As a result, climate impacts that depend on these temperature changes (such as reductions in snowpack) could occur later in the century than initially projected. A general assessment of how the 2005 climate change scenarios alter the results of previous hydrologic simulations is available in Lettenmaier et al. 2005. More detailed information on how the new scenarios may affect PNW resources will become available as the CIG incorporates the new scenarios into future climate impacts assessments.

Table 2: A comparison of the CIG's 2005 climate change projections with previous scenario projections from four (of eight) climate models as used in Snover et al. 2003 and Hamlet 2001. (More summary data)

	Annual Temperature Change		Annual Precipitation Change
	Previous Scenarios	2005 Scenarios	Previous Scenarios	2005 Scenarios
2020s
Low	+ 2.5ºF (1.4ºC)	+ 0.7ºF (0.4ºC)	+ 2%	- 4%
Average	+ 3.1ºF (1.7ºC)	+ 1.9ºF (1.1ºC)	+ 6%	+ 2%
High	+ 3.8ºF (2.1ºC)	+ 3.2ºF (1.8ºC)	+ 14%	+ 7%
2040s
Low	+ 3.1ºF (1.7ºC)	+ 1.4ºF (0.8ºC)	- 3%	- 4%
Average	+ 4.1ºF (2.3ºC)	+ 2.9ºF (1.6ºC)	+ 4%	+ 2%
High	+ 5.2ºF (2.9ºC)	+ 4.6ºF (2.6ºC)	+ 9%	+ 9%

Climate Change Streamflow Scenarios

To assist water resources planners and decision makers in assessing their vulnerability to projected future climate change, the CIG has developed a climate change streamflow scenarios tool for converting regional historical streamflow records into projected future streamflows for the 2020s and the 2040s. The streamflow scenarios, which are free to the public, can be directly incorporated into existing planning methods (such as critical period analysis) in place of the historic streamflow record.

Please note that the data provided in the climate change streamflow scenarios tool are currently based on previous CIG scenarios. Given the slower rate of warming and drier precipitation projections associated with the 2005 scenarios, it is likely that any temperature-related effects for the 2020s and 2040s (e.g., decreases in snowpack) would occur later in the century. Any precipitation-related effects projected for the 2020s in the previous scenarios (e.g., increases in annual streamflow volume) are probably overestimated. For more information on how a particular data set may be affected by the new scenarios, please contact the CIG.

Planning for Climate Change

Decisions made today can shape future vulnerability to a variety of stresses, including climate change. An examination of the possible impacts of future climate changes provides valuable information that can be used to inform planning in the PNW. The CIG provides numerous resources, including climate change scenarios for changes in temperature and precipitation, the climate change streamflow scenarios tool, and consultancy-based climate impacts studies, to support the inclusion of climate change information in PNW resource management and planning.

The availability of different global climate models and forcing scenarios (e.g., A2 and B1) allows the CIG to evaluate a range of possible future climate conditions for the PNW. Determining which scenario range should be used for more detailed modeling studies (e.g., evaluating climate change impacts on a specific water supply) will depend in part on the sensitivity of a resource to variations in climate and the risks associated with those changes. For example, the implications of drought for a river serving as the primary water source for a dense metropolitan area may be much greater than for a similarly sized river with few demands placed on it.

When relatively little is at stake, resource managers may want to choose climate change scenarios with the least amount of warming (i.e., best case scenarios) for evaluating specific climate impacts. When there is more at stake, or when climate impacts could have irreversible ecosystem consequences, resource managers may want to consider warmer (i.e., worst-case) scenarios. The appropriateness of any one model and/or climate scenario for assessing climate impacts should be evaluated on a case-by-case basis depending on the nature of the study.

What about uncertainty? Continued research on the global climate system and PNW environment will continue to expand our understanding of climate change impacts. The absence of "perfect information" should not, however, prevent planning for climate change. Good decisions can be made in spite of the uncertainty associated with projected changes, just as good decisions are made in spite of uncertainty about other factors, such as future economic conditions or rates of population growth. Careful consideration of the range of projected climate impacts, combined with an analysis of a resource's vulnerability to these impacts, will support prudent approaches to planning.

Caveats and Cautions

Regional signal. Global climate models are not designed to simulate regional climate. Important regional features like mountain ranges and estuaries are missing. The pattern of changes in temperature, however, is expected to be (and has been in the past) fairly similar across the whole region.

Climate variables. Models simulate observed patterns of temperature better than observed sea-level pressure (a representation of atmospheric circulation and common weather features), and they simulate sea-level pressure better than precipitation. Consequently, we have highest confidence in projections of temperature change, less confidence in projections of changes in atmospheric circulation, and lowest confidence in simulations of precipitation. Other details of climate that are badly simulated in present climate, such as changes in the frequency or intensity of storms, are probably unreliable in future climates as well.

Contrasts within the region. Simulations with a regional model (a climate model with very high spatial resolution) suggest a few important respects in which climate change may differ from the projections of the global models. For example, warming may proceed more quickly at higher elevations than in the lowlands owing to snow-albedo feedback: when snow cover is reduced, it enhances absorption of solar radiation and warms the surface more (Leung et al. 2003).

Last updated February 14, 2006