1

Challenges associated with projecting urbanization-induced

heat-related mortality

David M. Hondula1,2, Matei Georgescu2, Robert C. Balling, Jr.2

1 Center for Policy Informatics2 School of Geographical Sciences and Urban Planning

2nd International UGEC Conference

Taipei, Taiwan November 2014

2

Maricopa County, Arizona

• Part of 12th largest metro area in U.S. (2012 pop: 3.9 M)

• Mean monthly high temperatures:– June: 40ºC/104ºF– July: 41ºC/106ºF– August: 40ºC/104ºF

3

2006 2007 2008 2009 2010 2011 20120

20

40

60

80

100

120

He

at-

Re

late

d D

ea

ths

Heat-health impacts in Maricopa County

HeatAZ.com (Maricopa County Dept. of Public Health)

4

Urbanization: A key climate forcing

Georgescu et al. 2013, Nature Climate Change

SunCorrHi Scenario

5

Urbanization: A key climate forcing

Georgescu et al. 2013, Nature Climate Change

SunCorrHi vs. B1 2040-2060

SunCorrLow vs. A2 2040-2060

6

Contrasting observed diurnal temperature trends and the “oasis effect”

Georgescu et al. 2013, Nature Climate Change; Georgescu et al. 2011, JGR

Urban site

Rural site

7

Study Framework

MaxTemp

Mean Temp

MinTemp

Moving averages

SunCorrHi

SunCorrLow

SunCorrAdapt

Urb

aniz

atio

n s

cen

ario

Exposure variable

Projections of future heat-related mortality

8

Urbanization Effects on Temperature

“Max” “Min” “Mean”

Hondula et al. 2014, Science of the Total Environment

9

Temperature-Mortality Model

Hondula et al. 2014, Science of the Total Environment

10

Temperature-Mortality Model

Hondula et al. 2014, Science of the Total Environment

11

Calculation of Excess Mortality

0 20 40 60 80Day of Summer

48

49

5

0D

aily

Mo

rtal

ity

June 7Observed temp. (°C) 43.5

Baseline mortality 50.04

RR Mortality at given temperature

1.017

Predicted excess mortality

0.862

Hondula et al. 2014, Science of the Total Environment

12

Synthetic Future Temperature Time Series

Hondula et al. 2014, Science of the Total Environment

13

Projections of urbanization-induced heat-related mortality for greater Phoenix

Hondula et al. 2014

Science of the Total Environment

Max

Mean

Min

+205 deaths/yr

-46 deaths/yr

14

Projections of urbanization-induced heat-related mortality for greater Phoenix

Hondula et al. 2014

Science of the Total Environment

Max

Mean

Min

15

Conclusions

• Urbanization may contribute to future changes in heat-related mortality

• Projections highly sensitive to urbanization scenario and exposure variable choice

• Urban “oasis” effect in arid and semi-arid climates a potential mitigator of daytime heat stress

• Preference for low growth or adaptation strategies depends on key exposure variable

16

Acknowledgements

• ASU Piper Health Policy Informatics Initiative

• NSF EAR-1204774

• Office of Vital Statistics, Arizona Department of Health Services

• ICB local and scientific committees

17

Challenges associated with projecting urbanization-induced

heat-related mortality

David M. Hondula1,2, Matei Georgescu2, Robert C. Balling, Jr.2

1 Center for Policy Informatics2 School of Geographical Sciences and Urban Planning

2nd International UGEC Conference

Taipei, Taiwan November 2014

18

Projections of urbanization-induced heat-related mortality for greater Phoenix

Max Mean Min

Hondula et al. 2014, Science of the Total Environment

19

Urbanization: A key climate forcing

Georgescu et al. 2013, Nature Climate Change

SunCorrHi Scenario

SunCorrLow

SunCorrAdapt

20

Urbanization Scenarios

Georgescu et al. 2012, Envi. Res. Letters

21

Projections of urbanization-induced heat-related mortality for greater Phoenix

Hondula et al. 2014, Science of the Total Environment

22

Projections of urbanization-induced heat-related mortality for greater Phoenix

Moving Average FilterExposure Variable Max Min Mean Max Min Mean Max Min MeanSunCorr Urbanization Scenario Hi Hi Hi Lo Lo Lo Adapt Adapt AdaptPresent-Day Heat-Related Deaths (HRD) per 10 years 487.4 670.3 797.9 487.4 670.3 797.9 487.4 670.3 797.9Future HRD per 10 years

Run1 261.3 2530.1 1943.0 378.2 1391.1 1316.9 42.8 2312.6 1288.4Run2 348.4 2735.0 2244.7 403.4 1404.6 1343.8 46.7 2439.2 1444.5Run3 264.5 2759.5 2163.1 327.3 1392.7 1253.8 24.7 2317.8 1269.8Run4 HRD 392.9 2871.4 2402.6 440.8 1498.6 1454.4 26.5 2454.0 1449.0Mean of Runs1-4 311.6 2721.0 2187.5 391.3 1408.0 1335.8 31.2 2378.3 1364.2Absolute Change in HRD per 10 years

Run1 -226.2 1859.8 1145.1 -109.3 720.8 519.0 -444.7 1642.3 490.5Run2 -139.1 2064.7 1446.8 -84.0 734.3 545.9 -440.8 1768.9 646.6Run3 -222.9 2089.2 1365.2 -160.2 722.4 455.9 -462.7 1647.5 471.9Run4 -94.5 2201.1 1604.7 -46.6 828.3 656.5 -460.9 1783.7 651.1Mean of Runs1-4 -175.9 2050.7 1389.6 -96.2 737.7 537.9 -456.2 1708.0 566.3Relative Change in HRD

Run1 -46.4% 277.5% 143.5% -22.4% 107.5% 65.0% -91.2% 245.0% 61.5%Run2 -28.5% 308.0% 181.3% -17.2% 109.5% 68.4% -90.4% 263.9% 81.0%Run3 -45.7% 311.7% 171.1% -32.9% 107.8% 57.1% -94.9% 245.8% 59.1%Run4 -19.4% 328.4% 201.1% -9.6% 123.6% 82.3% -94.6% 266.1% 81.6%Mean of Runs1-4 -36.1% 305.9% 174.2% -19.7% 110.1% 67.4% -93.6% 254.8% 71.0%

MA1

Hondula et al. 2014, Science of the Total Environment