Bryan JonesCUNY Institute for Demographic ResearchNational Center for Atmospheric Research

Projecting Future Exposure to Climate Related Hazards: Extreme

Heat

Outline

1) Project overview2) The North American Regional

Climate Change Assessment Program (NARCCAP)

3) Spatially explicit future population scenarios

4) Projecting and decomposing exposure to extreme heat

5) Conclusions and ongoing work

Developing New Models to Understand Human Vulnerability to Climate-Related Hazards at Multiple

ScalesClimate Scenarios

Spatial Population Scenarios

EconomicGeographicDemographic

Vulnerability/Exposure

1. Identify the multi-level drivers of spatial population change.

2. Construct a theoretically consistent modeling framework for producing spatial population scenarios.

3. Assess exposure and vulnerability to climate-related hazards under alternative scenarios.

Q: How do climate and demographic change combine to drive future patterns of exposure and vulnerability to climate hazards?

NARCCAP

High resolution (1/2°) climate change simulations for:

• Analysis of climate model uncertainty• Impacts, adaptation, and vulnerability research

NARCCAP

General Circulation M odel (GCM ) Acronym SponsorCom m unity Clim ate System M odel v3 [4] CCSM National Center for Atm ospheric ResearchCanadian Global Clim ate M odel v3 [5] CGCM 3 Canadian Center for Clim ate M odeling and AnalysisHadley Centre Clim ate M odel v3 [6] HadCM 3 Hadley CentreGeophysical Fluid Dynam ics Laboratory Global Clim ate M odel [7] GFDL National Oceanic and Atm ospheric Adm inistration

Regional Clim ate M odel (RCM ) Acronym M odeling GroupCanadian Regional Clim ate M odel CRCM OURANOS / UQAMExperim ental Clim ate Prediction Center Regional Spectral M odel ECP2 UC San Diego / ScrippsHadley Centre Regional M odel v3 HRM 3 Hadley CentrePSU / NCAR M esoscale M odel M M 5I Iowa State UniversityRegional Clim ate M odel v3 RCM 3 UC Santa CruzW eather Research and Forecasting M odel W RFG Pacific Northwest National Lab

• Four Atmosphere-Ocean General Circulation Models (GCMs)

• Downscaled using six Regional Climate Models (RCMs)• Eleven GCM-RCM combinations

NARCCAP Example: Temp & Precip ProjectionsWinter Spring

Summer Autumn

NCAR/CUNY Spatial Population Downscaling ModelResearch Goal:

To develop an improved methodology for constructing large-scale, plausible future spatial population scenarios which may be calibrated to reflect alternative regional patterns of development for use in the scenario-based assessment of global change.

Important Characteristics:

• Capture observed geographic patterns• Flexible framework• Theoretically grounded

Data Considerations:

• Resolution• Projectable• Consistency

Existing Large-Scale Projection Methods

•Proportional scaling (Gaffin et al., 2004; Bengtsson et al., 2006; van Vuuren et al., 2007)

•Trend extrapolation (Balk et al., 2005; Hachadoorian et al., 2011)

•Hybrid (Asadoorian, 2005; Nam and Reilly, 2012)•Gravity-based (Grübler et al., 2007)•Smart Interpolation (e.g., EPA, 2010)

EPA A2-Scenario (2100)IIASA A2-Scenario (2100)

Gravity-Based Models

Boston

New York

Albany

Boston

New York

Albany

Hartford

• Flows – 2 directions• Spatial interaction• Migration and transportation

• Influence – 1 direction

• Spatial allocation• Accessibility or attractiveness

Gravity

Potential

2ij

jiij D

PPI

n

j ij

ji D

Pv1

2

•Population agglomeration is a proxy for the socio-economic characteristics of “attractive” places

•Spatial choice informed by accessibility

Projected US Population: IIASA

A2 Scenario

Assumptions:

Population Potential SurfaceNew Population DistributionPopulation Distribution

NCAR/CUNY Spatial Population Downscaling Model

•Allocation is limited by a geospatial mask indicating land not suitable for development.

m

j

djiii

ijePlav1

Traditional population potential model:

Modified potential model:

Spatial Mask

Adjustment Factor

Population

ParameterDistanceParameter

Population

Distance

• Urban and rural populations coexist within grid cells.

• Parameters (α and β) are estimated from historical data for both urban and rural distributions.

• Potential is calculated for each cell over a window of 100km.

NCAR/CUNY Spatial Population Downscaling Model

m

j ij

ji D

Pv1

2

m

j

djiii

ijePlav1

Base-Year Distributi

on

Calculate Urban

Potential

UrbanParamete

rs

Spatial Mask

Adjustment

Factor

Historical Data

Elevation

SlopeSurface Water

Protected Status

Allocate Projected

Rural Population

Change

Allocate Projected

Urban Population Change

New Population Distributi

on

RuralParamete

rsCalculate

Rural Potential

Final Population Distribution

t(100)

NCAR/CUNY Spatial Population Downscaling Model

National/Regional Population Projection

Northeast ParametersSouth ParametersMountain Parameters

Alternative Parameters: Northeast Census Region

SRES Scenarios

A2 - USA• Medium/fast population growth

• Emphasis on economic growth

• Rapid urban expansion

B2 - USA• Slower population growth

• Emphasis on local problem solving

• Environmental protection and social equity are valued

• Slower urbanizationSource: Nakicenovic et al., 2000

SRES Population Scenarios: United States

NCAR A2-Scenario (2100) NCAR B2-Scenario (2100)

Base-year (2000)

Results: NCAR A2 and B2 Scenarios

Source: IPCC, 2012

Exposure Sensitivity

Potential Im pact

Adaptive Capacity

Vulnerability

Projecting Exposure to Extreme HeatExposure and Vulnerability

• Barnett et al. (2010) find that the strong correlation between temperature measures lead to similar predictive

ability.• Gasparrini et al. (2012) find that excess mortality related to extreme heat events can be effectively described as the independent effect of individual days’ temperature rather

than as a function of multi-day heat waves.

Extreme Heat

> 35°C

Exposure to Heat Extremes in the United States

Exposure Sensitivity

Potential Im pact

Adaptive Capacity

Vulnerability

“A2-South” Population Scenario

Calculating Exposure

Projected Change in Exposure

2.3 billion person days

• Exposure is projected to increase anywhere from 4 to 6 times observed levels

9.8 billion person days

Spatial Variation in Projected Exposure

Decomposing Change in Exposure

Additional model runs:

1) Constant population (Climate Effect)

2) Constant climate (Population Effect)

3) Constant climate, constant spatial distribution (National)

4) Constant climate, constant spatial distribution (Divisional)

National-level Decomposition

Divisional-level Decomposition

Decomposing the Population Effect

• Aggregate growth• Broad redistribution (migration)

• Local Redistribution (urban/spatial structure)

57%34%9%

Including higher resolution estimates of the urban heat island effect may increase importance of

local change.

Decomposing the Population Effect

•A four to six-fold increase in exposure to temperatures above 35°C is projected over the next 40 years.

•Population and climate change both contribute significantly.

•There is significant regional variation in both exposure and its drivers.

•Aggregate population growth and broad population redistribution drive the population effect.

Conclusions and ongoing work

•Global exposure•Representative Concentration Pathways (RCPs) and Shared Socioeconomic Pathways (SSPs) combinations•CESM climate projections (29-member ensemble)•SSP-based spatial population scenarios

•Extreme-heat mortality•Mortality response to exposure (alternative definitions of extreme heat)•Impact of population structure, socio-economic status, and mitigation strategy•Probabilistic mortality projections based on adaptation assumptions

•Continued refinements to the spatial population model

ResourcesNARCCAP: http://www.narccap.ucar.edu/index.html

RCP Database: http://tntcat.iiasa.ac.at/RcpDb/dsd?Action=htmlpage&page=welcome

SSP Database: https://secure.iiasa.ac.at/web-apps/ene/SspDb/dsd?Action=htmlpage&page=about

IPCC 5th Assessment Report: http://www.ipcc.ch/report/ar5/index.shtml

Climate and Socioeconomic Scenarios: http://climate4impact.eu/impactportal/downscaling/downscalingdocs.jsp?q=Scenarios

http://www.ipcc.ch/pdf/supporting-material/expert-meeting-ts-scenarios.pdf

http://www2.cgd.ucar.edu/sites/default/files/iconics/Boulder-Workshop-Report.pdf