Julie A. Winkler. Michigan State University

Transcription

1 Julie A. Winkler Michigan State University

2 Outline National Climate Assessment Process and Structure Organization of the Midwest Technical Input Team (MTIT) Climate Projections for the Midwest Goals of the MTIT whitepaper on climate projections General terminology Examples of projected future changes Other uncertainty sources (structural uncertainty of climate impact models) Brief overview of major findings of MTIT sectoral reports

3 Overview of the National Climate Assessment and the MTIT Report

4 NCA Process and Structure NCA Development and Advisory Committee (NCADAC) Regional, Sectoral, and Cross-Cutting Author Teams Produce short chapters intended for public Regional and Sectoral Technical Input Teams Support the activities of the chapter authors Develop more detailed support documents written at a technical level Incorporate both peer-reviewed and gray literature

5 The Midwest Technical Input Team (MTIT) Report Responsibility for the MTIT report Great Lakes Sciences and Assessments Center (GLISA) National Laboratory for Agriculture and the Environment MTIT report was published this summer by Island Press, titled Climate Change in the Midwest: A Synthesis Report (pdf available free of charge). Sectors included in the MTIT report generally mirror those included in the National Assessment report with the addition of two background whitepapers. Whitepaper authors were recruited from experts in the region. Attempted to identify a diverse author group. Each whitepaper was reviewed by two experts in the field, and underwent revision to address reviewer comments. EPA also provided an agency review of the majority of the whitepapers.

6 Sectoral Whitepapers #1 Agriculture in the Midwest Jerry Hatfield National Laboratory for Agriculture and the Environment Impacts and Adaptation in the Biodiversity and Ecosystems Sector Kimberly Hall The Nature Conservancy Great Lakes Nearshore and Coastal Systems Scudder D. Mackey Habitat Solutions NA

7 Two Background Whitepapers Historical Climate and Climate Trends in the Midwestern USA Jeff Andresen Michigan State University Steve Hilberg Illinois State Water Survey Ken Kunkel National Climatic Data Center Climate Projections for the Midwest: Availability, Interpretation and Synthesis Julie Winkler Michigan State University Raymond Arritt Iowa State University Sara Pryor Indiana University

8 Sectoral Whitepapers #2 Climate Change and Energy Janice A. Beecher Michigan State University Jason A. Kalmbach Michigan State University Climate Change Vulnerabilities within the Forestry Sector for the Midwestern United States Stephen D. Handler Christopher W. Swanston Patricia R. Butler Leslie A. Brandt Maria K. Janowiak Matthew D. Powers P. Danielle Dutto Outdoor Recreation and Tourism Sarah Nicholls Michigan State University Northern Institute of Applied Climate Science USDA Forest Service Northern Research Station USDA Forest Service Eastern Region Michigan Technological University

9 Sectoral Whitepapers #3 Climate Change Impacts on Transportation in the Midwest John Posey. Water Resources Brent Lofgren Great Lakes Environmental Research Laboratory Andrew Gronewold Health Jonathan Patz University of Wisconsin Madison East-West Gateway Council of Governments Great Lakes Environmental Research Laboratory

10 Projections of Future Climate for the Midwest

11 Regional climate trends Mean temperatures have increased since Annual precipitation across the Midwest generally decreased from the late 1800s through the dust bowl years of the mid 1930s, followed by a generally increasing trend that continues to present. Increases in temperature and precipitation have not been consistent across season or time of day. Growing season length has increased across the region during the past several decades. A reduction in the amount and duration of ice cover on lakes across the Midwest, including the Great Lakes, has been observed.

12 Review commonly-used approaches to develop local/regional climate projections and highlight strengths and limitations. Provide readers with a basic understanding of climate projections to aid in an informed and nuanced interpretation of the substantial literature on potential climate impacts in the Midwest region. Summarize by climate variable potential future changes in the Midwest. Incorporated a wide range of climate projections available for the region Expanded upon Climate of the Midwest U.S. (Kunkel et al. 2013) prepared for the National Climate Assessment Development and Advisory Committee Table of downscaled climate projections available for the region

13 Downscaling Infer higher spatial or temporal resolution Dynamical downscaling Use of numerical models such as regional climate models Statistical downscaling Empirical-dynamical downscaling Surface variable is related to a circulation and/or free atmosphere variable Disaggregation downscaling Infer finer-scale values from coarse-scale spatial or temporal field of a particular variable (e.g., temperature) Figure 1. Illustration of the spatial scales of climate projections, as developed using dynamical, empirical-dynamical, and disaggregation downscaling methods applied to GCM simulations. Note that multiple downscaling steps can be applied. SOURCE: Winkler et al., 2011.

14 North American Regional Climate Change Assessment Program Regional climate model simulations driven both by reanalysis fields and by GCM results Available for historical and mid-century time slices Table 1: Available NARCCAP simulations. Regional Climate Models (RCMs) Global Climate Models (GCMs) GFDL CGCM3 HADCM3 CCSM NCEP CRCM X X X ECP2 X X X HRM3 X X X MM5I X X X RCM3 X X X WRFG X X X ECPC WRFP SOURCE: X X

15 An ensemble is a suite of climate projections An attempt to capture multiple sources of uncertainty Figure 3. Development of an ensemble of climate projections. The dashed line indicates uncertainty sources that are infrequently considered. Source: Winkler et al. 2011b.

16 An ensemble is considered to provide a lower bound of the maximum range of uncertainty (Stainforth et al., 2007) Ensemble of opportunity Adding an additional projection to an ensemble may not provide an improved estimate of the overall uncertainty The degree to which consensus among projections can be interpreted as increased confidence in a future change is unclear Interpreting ensemble means Ensemble members are usually equally weighted. An ensemble mean can be misleading. Source: IPCC, 2001

17 Focused on temperature, precipitation and wind variables

18 Figure 7. Temperature and precipitation changes over North America from the MMD-A1B simulations. Annual mean, DJF and JJA temperature change between 1980 to 1999 and 2080 to 2099, averaged over 21 models. OURCE: Christensen et al Annual increase (ensemble mean) of approximately 5.5 F in Midwest by Larger increase in summer (~8 F) over the western portion of Midwest In winter a southwest to northeast gradient is evident with largest increase in northeast (~9 F)

19 Figure 1. Multi-model mean annual differences in temperature ( F) between the 3 future periods and , from the 15 CMIP3 model simulations. SOURCE: Kunkel et al Figure 2. Multi-model mean annual and seasonal differences in temperature ( F) between and , from the 9 NARCCAP regional climate model simulations. SOURCE: Kunkel et al. 2012

20 Some differences evident between statisticallydownscaled projections compared to GCM or RCM output. For example, WICCI scenarios (A1B emissions scenario) Annual warming of 4-9 F by mid-century Warming in summer of 3-8 F by mid-century (largest changes in northern Wisconsin) [contradictions NARCCAP spatial pattern] Warming in winter of 5-11 F by mid-century (with largest increases in northwestern Wisconsin) Projected Change in Summer Average Temperature ( F) from 1980 to 2055 Projected Change in Annual Average Temperature ( F) from 1980 to 2055 Source: WICCI Projected Change in Winter Average Temperature ( F) from 1980 to 2055

21 Hayhoe et al. (2010a,b) downscaled scenarios from 3 CMIP3 models for Great Lakes region For early period ( ) larger projected changes in winter For mid-century, larger projected changes in summer in southern portion of Midwest (Indiana, Illinois) and in winter in northern portion (Wisconsin, Minnesota) Fig. 5. Projected increase in (a) winter (Dec Jan Feb) and (b) summer (Jun Jul Aug) average temperature as simulated under the SRES A1fi (higher) and B1 (lower) emissions scenarios by the average of 3 AOGCMs for nearterm ( ), midcentury ( ), and end-of-century ( ). Temperature projections are in units of degrees Celsius relative to the average and have been statistically downscaled to a spatial resolution of one-eighth degree. Source: Hayhoe et al. 2010a

22 NARCCAP scenarios suggest to considerable spatial variability 25 day average increase in southern portion of Midwest Fewer than 5 days in northern portion of Midwest Similar in magnitude to Pileus Project scenarios (even though estimated from older GCM simulations Figure 5. Spatial distribution of the NARCCAP multi-model mean change in the number of days with a maximum temperature greater than 95 F between and (top). Climatology of the number of days with a maximum temperature greater than 95 F (bottom). SOURCE: Kunkel et al Source: pileus.msu.edu. Projected Change in the Frequency of 90 F Days Per Year from 1980 to 2055 SOURCE: WICCI

23 Figure 7. Spatial distribution of the NARCCAP multi-model mean change in the annual maximum number of consecutive days with a maximum temperature greater than 95 F between and (top). Climatology of the annual maximum number of consecutive days with a maximum temperature greater than 95 F (bottom). NARCCAP Annual maximum number of consecutive days per year with 95 F will increase from 15 days in extreme southern portion of Midwest to less than 5 days in the northern portion. Hayhoe et al. (2010a,b) frequency of heat waves similar to 1995 Chicago heat wave expected to range from every other year (low greenhouse gas emissions) to three times period year (high greenhouse gas emissions scenario)

24 Figure 9. Spatial distribution of the NARCCAP multi-model mean change in the length of the freeze-free season between and (top). Climatology of the length of the freeze-free season (bottom). NARCCAP Fairly uniform increase across region of days by mid century Pileus Project Somewhat smaller projected increase of approximately 15 days in Michigan

25 Projected changes in freeze risk are highly uncertain Source: pileus.msu.edu.

26 Large degree of uncertainty in Midwest precipitation projections evident from initial NCA 2000 report and later reports. Ensemble mean of CMIP3 models for end-of-century suggests: Increase in annual and winter precipitation for much of the Midwest, except for western portion Little change or a small decrease in summer Over 90% of the 21 models project an increase in precipitation in Michigan, northern Ohio, Indiana, and Illinois. Approximately half of the 21 GCMs projected an increase in precipitation in the Midwest by the end of the 21 st century and the other half projecting a decrease or no change. Figure 7. Temperature and precipitation changes over North America from the MMD-A1B simulations. Top row: annual mean, DJF and JJA precipitation change (in percent) between 1980 to 1999 and 2080 to 2099, averaged over 21 models. Bottom row: number of models out of 21 that project increases in precipitation. SOURCE: Christensen et al

27 Figure 12. Multi-model mean annual differences in precipitation (%) between the 3 future periods and , from the 15 CMIP3 model simulations. SOURCE: Kunkel et al Figure 13. Multi-model mean annual and seasonal differences in precipitation (%) between and , from the 9 NARCCAP regional climate model simulations. SOURCE: Kunkel et al. 2012

28 Downscaled Scenarios WICCI scenarios Approximate 25% increase in winter and early spring (i.e., March) precipitation Little confidence in precipitation change for other seasons. Hayhoe et al. 2010a Increase in precipitation in Great Lakes region in winter and spring, especially in the southern portion of area (Illinois, Indiana, Ohio) Projected Change in Winter Average Precipitation (inches) from 1980 to SOURCE: WICCI Fig. 7. Projected change in (a) spring (Mar Apr May) and (b) summer (Jun Jul Aug) average precipitation as simulated under the SRES A1fi (higher) and B1 (lower) emissions scenarios by the average of the subset of 3 AOGCMs used for the impact analyses presented in this volume. Precipitation projections are in units of percentage change relative to the average and have been statistically downscaled to a spatial resolution of one-eighth degree. SOURCE: Hayhoe et al. 2010a

29 Figure 17. Spatial distribution of the NARCCAP multi-model mean change in the number of days with precipitation exceeding 1 inch between and (top). Climatology of the number of days with precipitation exceeding 1 inch (bottom). SOURCE: Kunkel et al Projected Change in the Frequency of 2" Precipitation Events (days/decade) from 1980 to 2055 SOURCE: WICCI

30 Evaluation of NARCCAP simulations for midcentury displays weak consistence in the climate change signal. Current suite of climate projections suggests little change in wind resources or wind extremes to mid-century or longer. Figure 8. Difference in the fifty-year return period sustained wind speed (U 50yr ) over the Midwestern US for vs The frames show the different AOGCM-RCM combinations. The magnitude of change is only shown for grid cells where the value for the future period lies beyond the 95% confidence intervals on the control period. Note; none of the grid cells behind the legend in frame (b) exhibited significant changes. SOURCE: Pryor and Barthelmie (2012b).

31 There is no single best climate model or downscaling approach. There is greater confidence in projected temperature change than precipitation change. In spite of confidence in future warmer temperatures, change in freeze risk remains uncertain. The degree of uncertainty surrounding precipitation change remains high. There is little confidence in the sign (positive or negative) of change in mean precipitation for the warm season. There is somewhat greater confidence in projections of increases in the frequency and intensity of extreme warm season precipitation events. The use of a multimodel mean of a projected change may be misleading, particularly for projected changes in precipitation. Some wind extremes occur at scales below those captured by global and regional climate models or involve processes that are not well understood, but the current suite of climate projections suggests little change in wind resources or wind extremes to the middle of the current century.

32 Name/ Reference CMIP3 GCM archive (Meehl et al. 2007) Coverage/Resolution/ Variables/Period Global Spatial resolution varies by GCM Archived at monthly time step, but finer time steps available for most models Ensemble Size Downscaling Procedures Availability Over 20 GCMs (AR4 era) 3 emissions scenarios (SRES A2, A1B, B1) Not downscaled Graphical summaries available in IPCC AR4 Working Group I report. Time series of monthly precipitation and mean temperature available from the Program for Climate Model Diagnosis and Interpretation ( ipcc.php) Bias Corrected and Downscaled WCRP CMIP3 Climate Projections (Maurer et al. 2007) Global 1/8 o lat/lon resolution Mid century and late century time slices 16 GCMs (IPCC AR4 era) 3 emissions scenarios (SRES A2, A1B, B1) Disaggregation (BCSD) method. Gridded temperature and precipitations observations were upscaled to a 2 resolution and GCM projections were regridded to this resolution. Quantile mapping was used to calculate change factors which were then downscaled using a simple inverse distance approach and applied to the original finely gridded observed dataset. Monthly time series available through Climate Wizard and at _projections

33 The structure of response/impact models has only recently been recognized as an important source of uncertainty for climate assessments. Examples Great Lakes water levels wheat production

34 Sectoral Assessments

35 Agriculture Main Findings Shifts in the timing of precipitation will affect field preparation time in spring. Observations for Iowa, for example, display a decrease from 1976 to 2010 in workable field days between April and mid-may. For C3 crops, reduction in evapotranspiration caused by increasing CO2 may diminish with increasing temperatures.. Overall impacts on perennial corps are uncertain because of the uncertainty in chilling requirements and the potential exposure to normal freezing conditions relatively later in terms of crop development. Furthermore, adaptation options may be fewer for perennial crops. Agriculture is a fluid system and continual adaptation is taking place to adjust to changing climate conditions. Producers have readily adopted changes which entail changes in planting date and maturity selections. Capitalintensive adaptations may be more difficult to implement. Adaptation strategies will need to include practices which protect the soil from erosion events while at the same time increase the soil organic matter content. It is difficult to evaluate how crop insurance payments will change in the future. Yield (kg ha -1 ) M idw est C orn G rain P roduction Year M ichigan Iow a Annual corn grain yields for Iowa and Michigan from 1866 through 2011 (Source: USDA- NASS).

36 Biodiversity -- Helping Species Adapt to Climate Change Increase connectivity and soften management In the Midwest there are many barriers to species movement, including natural features (e.g., Great Lakes) and vast responses of land that may be inhospitable due to current land use (e.g., agricultural land). Continue to pro-actively address the threat of invasives Shifting some of our conservation attention from species to stages (i.e., consistent landscape-scale units of variation) Moving from a focus on species to a focus on landscapes Increasing green infrastructure (e.g., forests and wetlands) to handle stormwater Restore functional ecosystems in watersheds dominated by agriculture Moving toward smarter conservation that is more agile and able to quickly shift strategies

37 Coastal Main Findings Interannual water level variability would support and maintain coastal wetland biodiversity and associated fish and wildlife habitats. Extreme precipitation events in winter and spring would increase nutrient and sediment loadings into the Great Lakes. Increased storm magnitude and frequency coupled with warmer surface water temperature would reduce ice cover, increase wave power, and reduce winter ice shore protection which will increase the risk for coastal flooding and result in accelerated erosion. Extended periods of lower water levels would offer potential new habitat for submergent aquatic vegetation and new coastal wetland communities. But exposed lakebed areas may be vulnerable to expansion by invasive wetland plant species. Increased surface water temperatures would cause gradual ecotonal shifts in aquatic species distributions from cold-waterspecies to warm-water species in intermediate to shallow water nearshore and coastal areas of the Great Lakes.

38 Energy Future Issues and Considerations Because electricity is an on-demand service and supply and demand must be balanced on a real-time basis, changes to demand have a direct and immediate bearing on supply. Extreme weather events would place further burdens on the supply of electricity. Climate change would also influence the performance of generation equipment. Higher temperatures result in decreased efficiency in combustion turbines that are primarily used to generate electricity in the Midwest region. Most energy production processes, traditional and alternative, are water intensive. Unpredictable water conditions is an operational challenge. Given variability in water supply, even relatively water-rich regions are not immune from these effects. The Midwest region might be relatively disadvantaged in terms of wind and solar energy resources, which would argue for expanding development of bioenergy resources. State renewable portfolio standards (RPSs) have become a centerpiece of climate policy.

39 Forestry Key Vulnerabilities Climate change will amplify many existing stressors to forest ecosystems, such as invasive species, insect pests and pathogens, and disturbance regimes (very likely). Climate change will result in ecosystem shifts and conversions (likely). Many tree species will have insufficient migration rates to keep pace with climate change (likely). Climate change will amplify existing stressors to urban forests (very likely). Forests will be less able to provide a consistent supply of some forest products (likely). Climate change impacts on forests will impair the ability of many forested watersheds to produce reliable supplies of clean water (possible). Climate change will result in a widespread decline in carbon storage in forest ecosystems across the region (very unlikely). Many contemporary and iconic forms of recreation within forest ecosystems will change in extent and timing due to climate change (very likely). Climate change will alter many traditional and modern cultural connections to forest ecosystems (likely). USGS

40 Outdoor Recreation and Tourism-- Summary Climate variability and change can have both direct and indirect impacts on outdoor recreation and tourism. Direct implications refer to climatic changes that impact the feasibility of, or satisfaction with, outdoor recreation and tourism activities. Indirect implications result from projected changes in the natural environment. Climate variability and change can impact both the supply of outdoor recreation and tourism resources and settings and the demand for outdoor recreation and tourism activities and experiences. Anticipating the reaction of participants to climate variability and change is complicated. Tourism Climatic Index over the United States for January and July in the 1970s, 2050s, and 2080s.

41 Transportation Summary of Key Impacts Medium confidence There is a rising risk of disruption of Mississippi River navigation. Flooding impacts are already significant and have grown in recent decades. There is a risking risk of temporary flooding of roads and rails due to riverine flooding and ponding. Underlying assumption is that the frequency of intense precipitation events is, and will continue to, increasing. There is a rising risk of disruption to Great Lakes navigation due to variability in water levels. Warmer temperatures will increase rail and expansion joint stress and decrease pavement life. Warmer temperature will create more difficult conditions for construction labor. Low confidence Warmer air temperatures and increased frequency of extreme weather and strong winds may disrupt air traffic. Faster stream currents caused by an increase in extreme precipitation events may result in increasing severity of bridge scour.

42 Water Resources Summary In general, precipitation has been increasing in the Midwest and this trend is projected to continue. Precipitation increases are particularly pronounced when looking at the winter season and when looking at the few largest rain events of the year, and this is expected to continue. Methods of calculating evapotranspiration (ET) under changed climate are the subject of emerging research, showing that widely-used methods based on temperature as a proxy for potential ET exaggerate projected increases in ET. When incorporated into further simulations, this leads to excessive reductions in streamflow and lake levels. Simulations using a more energy-based approach to ET give more mixed results in terms of changes in streamflow and lake levels, and often show increases.

43 Health Summary Four health impacts areas in the Midwestern region of concern relate to: urban heat waves air pollution water quality and waterborne diseases vectorborne diseases While some capacity to adapt is evident for the region, aging infrastructure poses concomitant risk, especially in the case of municipal water systems. Health benefits accruing from greenhouse gas mitigation can be large, as shown by a green transportation scenario. Therefore, such health benefits must be included in any assessments and policy discussions related to energy production or transportation planning.

44 MTIT Report available at: kex.org/virtua l- library/climat e-changemidwestsynthesisreportnationalclimateassessmentphp