Urban Sprawl and UHI in Dallas and Minneapolis
Urban Sprawl and UHI in Dallas and Minneapolis Matthew Welshans, MGIS Student, Penn State University April 11, 2014 – Association of American Geographers Annual Meeting
Project Summary • • • Define Urban Heat Island (UHI) and Urban Sprawl Explore Data Used in Project Methodology for Project Results Conclusions and Next Steps
Urban Heat Island Definition Image Source: US EPA (2012)
Why is Urban Heat Island a Concern? Kai Hendry (Flickr) Dr. Edwin Ewing/CDC Carrie Sloan (Flickr)
Urban Sprawl – NE of Dallas
The Problem • Urban Heat Island is affected by the growth of metropolitan areas – Size of heat island – Increase in temperature difference between rural/urban areas • What is the correlation between increased urban sprawl and the change in urban heat island?
Study Areas Dallas-Ft. Worth-Arlington, TX MSA • • • 12 counties in northeast Texas 2010 Population: 6, 426, 214 9, 286 square miles (~690/sq mi) Minneapolis-St. Paul, MN/WI MSA • • • 11 counties in southeast Minnesota and 2 in western Wisconsin 2010 Population: 3, 759, 978 6, 364 square miles (~590/sq mi)
Data Sets – Land Use/Land Cover Data (2001, 2006, 2011 Draft) • National Land Cover Database (Landsat 7) • Split into 15 land cover categories • Percent Impervious Surface (%IS) calculated per each pixel – Temperature Data • Derived from ASTER Imagery from the MODIS Satellite • Three swaths per study area were chosen based within 2 years of the LULC Data.
Why ASTER For Temperature Data? LANDSAT 7 ETM+ ASTER Satellite Landsat 7 (1999) Terra EOS Satellite (1999) Resolution Visible/NIR (4 bands): 30 m TIR (1 band): 60 m Visible/NIR (3 Bands): 15 m TIR (5 bands): 90 m From ASTER User Handbook Version 2 (2002)
Deriving Temperatures from ASTER • Temperature calculated using Gillepsie et al (1998)’s Temperature Emissivity Separation (TES) Method for each image. – Atmospheric Scattering effects filtered out – Max and min pixel emissivity calculated – Surface temperature ± 1. 5°C calculated using Planck’s Law
Methodology • Split each study area into eastern and western sections • Sampled each swath extent with ~10, 000 points • Averaged temperatures in each land cover category • Averaged temperatures based on 10 -percent intervals in percent impervious surface (IS) • Calculated average Urban (>15% IS) and Rural (<15% IS) to produce UHI calculation
Results – Minneapolis (West) UHI 6 2001 2. 28 C -0. 41 C 2. 68 C 5 2004 3. 23 C -0. 56 C 3. 79 C 4 2011 3. 17 C -0. 78 C 3. 95 C 3 8/6/2001 2 8/30/2004 9/10/2011 1 0 010 10 -2 0 20 -3 0 30 -4 0 40 -5 0 50 -6 0 60 -7 0 70 -8 0 80 -9 90 0 -1 00 Departure from Average (°C) 7 -1 -2 Percent Impervious Surface
Minneapolis (West)
Results – Dallas UHI 6 2001 1. 59 C -0. 30 C 1. 89 C 5 2005 1. 71 C -0. 63 C 2. 34 C 4 2013 1. 22 C -0. 57 C 1. 78 C 3 5/18/2001 2 3/10/2005 3/16/2013 1 0 010 10 -2 0 20 -3 0 30 -4 0 40 -5 0 50 -6 0 60 -7 0 70 -8 0 80 -9 90 0 -1 00 Departure from Average (°C) 7 -1 -2 Percent Impervious Surface
Dallas
Collin County, TX Pop 2000 Pop 2010 491, 675 782, 351
Why The Difference? • Daytime Surface Albedo (reflectivity) – Higher in cleared areas versus water, wetlands, and forest – Proportional to surface temperature – Differs depending on time of year
Why The Difference? Minneapolis (West) Water Wetlands 2001 5, 30% 8, 79% Minneapolis (West) Water 2006 Wetlands 9. 44% 5, 78% Urban 21, 66% Ag 44% Grass Shrub 2, 78% 1, 63% Barren 0, 06% Forest 15, 28% Ag 36. 66% Grass 29, 04% Barren 0, 10% Forest 16, 04% Dallas - 2006 Urban 22, 71% Barren 0, 11% Forest 11, 63% Shrub 0, 49% 6, 68% Ag 21. 96% Wetlands 2, 10% Shrub 0, 06% Grass Shrub 2, 55% 1, 32% Ag 18. 26% Water 4, 25% Urban 35, 22% Grass 25, 17% Urban 30, 74% Ag, 36% Grass Shrub 2. 96% 1. 39% 3, 73% Ag 30. 38% 7, 74% Urban 27, 64% Dallas - 2001 Wetlands Water 1, 91% Minneapolis (West) Water Wetlands 2011 Forest Barren 10, 91% 0, 33% Barren Forest 0, 14% 14, 99% Dallas 2011 Wetlands -Water 1, 43% 5, 14% Urban 41, 56% Grass 23, 00% Shrub Forest 0, 03% 10, 21% Barren 0, 38%
Conclusions • Generally good link between temperature and percent impervious surface • Land cover type plays key role in daytime surface temperature patterns – Lower temperatures around water, forests – Highest temperatures in urban, agriculture, grassland
Next Steps • Compare 2011 and upcoming 2016 land cover data to newer ASTER imagery • See if trends continue to hold up • Compare to nighttime imagery if possible to see how UHI patterns differ. • Reverse Migration and Green Initiatives
Acknowledgements • • • Dr. Jay Parrish – Advisor Beth King and Dr. Doug Miller – Penn State MGIS Program Jon Dewitz, Joyce Fry, Dr. Jim Vogelmann – USGS EROS Center Questions? maw 323@psu. edu
- Slides: 21