Technical Document: Using EnviroAtlas Data and Remote-Sensing to Identify Locations for Urban Heat Island Abatement This document was written to accompany the EnviroAtlas use case: Using EnviroAtlas to Identify Locations for Urban Heat Island Abatement. Data Sources Weather Underground: https://www.wunderground.com/historv/ 1. Historical air temperatures for reference location at Portland International Airport USGS Earth Explorer: http://earthexplorer.uses.gov/ 2. Landsat 8 thermal infrared cloud-free images for summer dates in 2013, 2014 and 2015 were downloaded from USGS Earth Explorer. Seven were included in the study. Table 1. The imagery for seven dates shown in this table were included in the creation of the UHI Index. USGS File Name Date Year Time (UTC) Day/Night Min Temp (deg F) Max Temp (deg F) LC80460282013184LGN00 3-Ju 1-13 2013 18:58 Day 60 85 LC80460282014187LGN00 6-Ju 1-14 2014 18:55 Day 63 88 LC80460282015190LGN00 9-Ju 1-15 2015 18:55 Day 62 87 LC80460282013216LGN00 4-Aug-13 2013 18:57 Day 61 89 LC80460282014219LGN00 7-Aug-14 2014 18:55 Day 57 82 LC80460282013232LGN00 20-Aug-13 2013 18:57 Day 57 83 LC80460282015238LGN00 26-Aug-15 2015 18:55 Day 55 90 EnviroAtlas: https://www.epa.gov/enviroatlas 3. Portland, OR Meter-scale Urban Land Cover 4. Estimated percent of tree cover within 26m of a road edge 5. Boundaries for US Census 2010 Block Groups for Portland OR 6. Population over 70 years old 7. Population under 13 years old Portland's Metro Data Resource Center: http://www.civicapps.org/datasets 8. Streets (Region) 9. Neighborhood Organizations National Hydrography Dataset for Oregon: http://nhd.usgs.gov/data.html 10. Streams and Rivers from "NHDArea" shapefile 11. Lakes and Ponds from "NHDWaterbody" shapefile Landsat 8 Land Cover Companion document to Using EnviroAtlas to Identify Locations for Urban Heat Island Abatement. ------- Methods 1. From USGS EarthExplorer, use multiple criteria to select Landsat 8 scenes to include in the analysis: a. High maximum air temperature at reference location of Portland International airport (according to Weather Underground's historical record), b. Low percent cloud cover over image, c. Complete coverage over the study area, and d. Collection at the same time of day. Starch Criteria Summary isnowi Clear Criteria Q ' httftt -' • ' Ui9i.gOV P • fl C [ EE EaithEnplorer X | Result oeeons seaicn nom;|mmfflaryyy» fgio: ImmWym ~]1 Loom Reqttioi ryn FeedbatH HiiIb | 1. Enter Search Criteria To narrow your search area type in an address or place name enter coordinates or click the map to define your search area (for advanced map tools, view the help documentation), and/or choose a date range Download Landsat 8 imagery from USGS EarthExplorer. 2. In QGIS, use the Semi-Automatic Classification Plugin a. Apply DOS-1 Atmospheric Correction, yielding an image of brightness temperature for band 10, which contains the thermal infrared information. (B i^i Semi-Automatic Classification Plugin ^ ^ Download images Tools ~I Preprocessing Postprocessing m Band calc Band set ^ Batch S > $ Landsat Sentinel-2 ASTER Clip multiple rasters Split raster bands ^ PCA kti Vector to raster Directory containing Landsat bands a Select MTL file (if not in Landsat directory) c t Brightness temperature in Celsius ~ Apply D0S1 atmospheric correction Use NoData value (image has black border) | 0 0 J Perform pansharpening (Landsat 7 or 8) Create Band set and use Band set tools Satellite Date (YYYY-MM-DD) Sun elevation Earth sun distance Band RADIANCE MULT RADIANCE ADD REFLECTANCE MUU REFLECTANCE 1 0 1 In QGIS, unpackage the Landsat 8 imagery to yield the brightness temperature for band 10. Companion document to Using EnviroAtlas to Identify Locations for Urban Heat Island Abatement. ------- 3. In ArcGIS, create a raster of emissivity corresponding to land cover a. Reclassify Portland lm land cover TIF to lm emissivity, using the following emissivity values from Setturu, Rajan and Ramachandra (2013): Table 1. Surface emissivity values by land cover type, taken from Setturu et al. (2013). Land Cover Emissivity 1 Built-up / Impervious Surfaces 0.946 2 Non-Crop Vegetation 0.985 3 Water 0.990 4 Agriculture 0.974 5 Other 0.950 b. Aggregate cells, with resulting 30m cells having the mean emissivity of the original lm cells. c. The result is a 30m raster of emissivity, with grid snapped to Landsat raster imagery. 4. In ArcGIS, estimate land surface temperature a. Use raster calculator to adjust at-satellite brightness temperature (Kelvin) to land surface temperature (LST; Kelvin). From Congedo (2014): LST = b / ( 1 + ( 10.8 * b / 14380 ) * ln(a)) where a is the emissivity raster and b is the brightness temperature raster. b. Convert units from Kelvin to Fahrenheit. 5. In ArcGIS, calculate final UHI raster a. Remove areas in larger streams, rivers and lakes so that the relative scale is for land areas only. b. Selected Forest Park as the location for the reference temperature, due to its high amount of land cover that is natural (given the location). c. Rescale each day's LST map by setting the mean temperature of Forest Park to zero, and calculating the difference above and below this reference temperature in degrees F. d. Average across the seven LST maps to yield a single map of summer daily UHI index. It displays the average difference from the Forest Park reference LST, i.e. the warming or cooling response of a pixel's LST on a summer day relative to mean LST in Forest Park. i. Our daily maps had Forest Park reference temperatures ranging from 67.1 °F to 73.8 °F. ii. The mean of the reference temperatures is 70.6 °F. Companion document to Using EnviroAtlas to Identify Locations for Urban Heat Island Abatement. ------- Resources Butler, K. (2014). "Deriving temperature from Landsat 8 thermal bands (TIRS)." Retrieved from the ArcGIS Blog from Esri: https://blogs.esri.com/esri/arcgis/2014/01/Q6/deriving-temperature-from- landsat-8-thermal-bands-tirs/ Congedo, L. (2014). "Estimation of Land Surface Temperature with Landsat Thermal Infrared Band: a Tutorial Using the Semi-Automatic Classification Plugin for QGIS." Retrieved from the From GIS to Remote Sensing Blog: http://fromgistors.blogspot.com/2Q14/01/estimation-of-land-surface- temperature.html Setturu, B., Rajan, K. S., & Ramachandra, T. V. (2013). Land surface temperature responses to land use land cover dynamics. Geoinfor Geostat Overview, 1, 4. USGS. (2015). Landsat 8 (L8) Data Users Handbook. Retrieved from https://landsat.usgs.gov/l8handbook section5.php Zhou, W., Qian, Y., Li, X., Li, W., & Han, L. (2014). Relationships between land cover and the surface urban heat island: seasonal variability and effects of spatial and thematic resolution of land cover data on predicting land surface temperatures. Landscape Ecology, 29(1), 153-167. Companion document to Using EnviroAtlas to Identify Locations for Urban Heat Island Abatement. ------- |