1 / 11

Impervious Surface Area of the Conterminous United States

This research examines methods for estimating impervious surface area across the U.S. using data from Landsat TM, DMSP-OLS, road density, and population data, leading to improved modeling accuracy at 1km resolution. The study compares different models and data sources to find the most effective approach for estimating ISA, highlighting the importance of combining multiple data sets for accurate results.

lanierj
Download Presentation

Impervious Surface Area of the Conterminous United States

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Impervious Surface Area of the Conterminous United States Christopher D. Elvidge John B. Dietz Paul S. Sutton

  2. Research Approach / Methods 1 km Grids • Aggregated 30 meter land cover from Landsat TM (MRLC) from the early 1990’s. Classes found most useful include 21 , 22, and 23 (low density residential, high density residential, and commercial / transportation). • Radiance calibrated nighttime lights from the DMSP-OLS (2001) • Aggregated road density processed from the U.S. Census Bureau TIGER (2000) • DOE Landscan population density data (2000)

  3. Research Approach / Methods (continued) • For calibration, the percent cover of impervious surface was measured being measured with gridded point counts made on 1 km tiles of high resolution aerial photograpahy (1998-00) acquired along transects crossing major metropolitan areas in each region of the country. Each photo tile covers a 1Km Albers Equal Area grid cell.

  4. 80 Photos From 13 Transects

  5. Regression Results • Radiance only: %ISA = 21.17 + .348 Radiance (R2 = 0.53) • Radiance & roads: %ISA = 3.91 + 0.214 Radiance + .0021 Roads (R2 = 0.79) • Radiance, roads & MRLC: %ISA = 2.242 + .136*Rad + .0010*Roads + .00452*MRLC23 + .00286*MRLC22 + .00278*MRLC21 (R2 = 0.85 • MRLC only: %ISA =  7.6625 + .0085 MRLC23  + .0060 MRLC22 + .0047 MRLC21 (R2 = 0.78)

  6. % ISA Model From MRLC Classes 21,22,23

  7. % ISA Model From DMSP Radiances Only

  8. %ISA Model From DMSP Radiances and Road Density

  9. %ISA Model From DMSP Radiance, Road Density and MRLC Classes 21, 22, 23

  10. %ISA From Four Models DMSP Radiance MRLC 21,22,23 Radiance, Road Density, MRLC 21,22,23 Radiance & Road Density

  11. Conclusion To Date • ISA can be modeled regionally to globally at 1 km resolution. • DMSP radiance is insufficient as a sole predictor. • DMSP radiance plus road density works well. • DMSP radiance, road density plus Landsat land cover gave the best results. • Effort required to generate continental scale Landsat land cover results in long delays and high cost. Other sources of land cover should be considered (e.g. MODIS). • Level of effort for Landsat land cover for ISA estimation could be reduced by focus on areas with DMSP lights.

More Related