Using Remote Sensing to Calculate Impervious Surface in Los Angeles

Impervious Surface in Los Angeles

Posted on December 15, 2012 

Assessing Impervious Surface in Los Angeles

Anshu Kishore
Geog 169
Prof. TW Gillespie, Department of Geography, UCLA
December 14, Fall 2012

View Map in ArcGIS Online of Impervious Area of Sample Parcels in Los Angeles

Abstract

The urban development and expansion has triggered the spread of Impervious surfaces across the nation. As cities grow and more development occurs, the natural landscape is replaced by roads, buildings, housing developments, and parking lots. The Los Angeles region in Southern California has experienced explosive growth over the last 50 years, and, along with it, large amounts of impervious surfaces have replaced the natural landscape. The goal of this research paper is to assess the expansion of impervious surfaces in Los Angeles through satellite images, compare the satellite imagery by Landsat and Quickbird, overlaying of GIS data on Satellite imagery to calculate impervious surface of UCLA.

Introduction

“Impervious surface” refers to all the hard surfaces like paved roads, parking lots, roofs, and even highly compacted soils like sports fields. The problem with impervious surfaces is that they prevent the natural soaking of rainwater into the ground and slowly seeping into streams. Instead, the rain water accumulates and flows rapidly into storm drains.
The population of Los Angeles has skyrocketed in the past 100 years, and it’s putting a lot of pressure on our environment. To accommodate the needs of growing population requires building of roads, parking lots and paved sidewalks. Although the urban development and expansion facilitated the system, its impact on the environment can’t be justified. Unlike the soil and vegetation in natural watersheds, most urban areas like LA are covered in impervious surfaces, such as asphalt or concrete. As a result the rainwater does not percolates through the ground slowly rather the water is rushed along the street picking up trash, oils and other kinds of waste into storm drain followed by the ocean. No doubt urban runoff in Los Angeles has become major source of water pollution . Not only this , the spread of impervious area in Los Angeles has intensified the speed of water flow causing soil erosion and urban flooding. Impervious surfaces is also playing major role in urban heat island effect too ( Kent). According to EPA, the term “heat island” describes built up areas that are hotter than nearby rural areas.The Heat Island Group says that the urban heat island around Los Angeles, California, costs the city $100 million a year in energy. According to the study conducted by USDA forest service to assess Tree and impervious cover change in US cities, Los Angeles has topped the chart as a city with greatest annual increase in impervious cover change average 550 ha ( High availability) / year followed by Houston ( 400 ha/ year) and Albuquerque (480 ha / year)(Nowak, and Eric).

Study Area

The study area covers Los Angeles Metropolitan Area specifically which comprises of Los Angeles County (9,862,049), and Orange County (3,010,759) and is home to 15.4 million people, making it the most populous metropolitan area in the western United States and the largest in area in the United States. The UCLA campus is being used to verify the satellite data on the ground.

Method

Part I : Download and manipulate Satellite images

To access the spread of Impervious surface in Los Angeles, I navigated to the USGS site
http://glovis.usgs.gov/, to download the Landsat Imagery. The details of imagery is mentioned below;
Landsat data Collection-> Landsat Archive-> Landsat 4 5 TM -> Lat/Long 34.6, -118.3 -> 0 percent cloud
ID: LT50410362004285PAC01
CC: 0% Date: 2004/10/11
Qlty: 9 Product: TM L1 ( Oct 2004)
240m resolution
Once the download was complete, unzipped file was opened on ENVI. On layer stacking window , the band 3 2 1 loaded as Red Blue Green for Natural color display of the imagery.

Results

This section presents the results from study, and we also discuss the findings.

Fig 1: A satellite image of Los Angeles with the focus on UCLA campus

Fig 1 shows a satellite image of Los Angeles with the focus on UCLA campus . The natural color imagery provides a general view of entire Los Angeles area. The 240 meter resolution does not provide clear distinction between different types of impervious surfaces but it clearly specify different land cover such as vegetation, water bodies and build up area. The ratio of natural landscape ( except Santa Monica mountains on the upper side of the image) and build up area as visible in the image suggests expansion of impervious surfaces in Los Angeles.

Fig 2: A Landsat imagery of Los Angeles.

Fig 2 displays the landsat imagery being layer stacked on ENVI with the bands 4 4 3 loaded as RGB. It shows the general overview of vegetation in Los Angeles area. The bright yellow suggests strong presence of vegetation, and the dark grey is mostly impervious surfaces.

Fig 3: Classified Landsat Image of Los Angeles

The significance of this image is to show the classification of landsat image data on the basis of spectral signature and extraction of land-cover information from it. The the data has been classified into seven categories. The blue color corresponds to water bodies, yellow vegetation . Certain surfaces with similar pixel values has been combined automatically, which makes it harder to classify every surface types on the basis of spectral signature. I have used this tool to examine extent of impervious surface. The pink color ( refer to the Fig 5 below) corresponds to the impervious surface, as I have verified using natural color images

Fig 4. Impervious Surface in Los Angeles

After classification of Landsat images using ENVI, it is important to compare images from other satellites too. In this case, I am using  Quickbird images with classification done  by McPherson in another study of the Los Angeles area with resolution of 2 m by 2 m. This is the baseline imagery available for the analysis and comparison of impervious surface in Los Angeles. In this study, we will use the the Quickbird imagery to compare the accuracy of impervious area calculated from Landsat imagery. The following imagery sourced from Microsoft Bing shows the UCLA campus and the its neighborhood.
.
Fig 5: McPherson Land Cover data on ArcMap

Fig 5 is a land cover Classification data of Los Angeles overlayed on Bing maps. The imagery used for the land cover classification consists of 64 scenes collected by QuickBird satellite in different seasons from 2002 to 2005 (Wu, Qingfu, and McPherson).

Fig 6: Classified Landsat Impervious surface area overlaid on Ariel Photography of UCLA

Fig 6 shows classified Landsat imagery using unsupervised ISOdata classification technique. This method to classify impervious surfaces on Landsat imagery has drawbacks. It classified approximately half of the impervious on the given parcel of land. The reason could be low resolution. On the other hand, Classified land cover data based on Quickbird provides detail view of impervious surfaces in and around UCLA campus. The imagery has captured impervious surfaces almost precisely.

Fig 7: Quickbird Imagery showing UCLA campus (Gray color shows impervious area)

Bing map showing area in around UCLA campus. The accuracy of Quickbird imagery can be verified using high resolution bing map of that same area of interest.
Fig 8: Aerial Imagery of UCLA

Fig 9: Zoomed in Imagery of UCLA

Fig 10: Impervious Areas in Yellow Polygons Ignored by Landsat Imagery

Aerial Imagery showing a part of UCLA campus ( impervious surfaces is classified with pink color). The comparison between Fig 10 and 11 shows serious discrepancies on Landsat classified images. Fig 11 shows accuracy of Quickbird imagery on impervious surfaces on UCLA campus.

The measurement of impervious area on a parcel of land can be calculated using GIS shapefile (containing parcel data) and classified data on impervious surface provided by Quickbird.

Fig 11: Impervious Area Identified by Quickbird Imagery

Fig 12: Reference Aerial Imagery of UCLA Showing Actual Impervious Area

Fig 13. UCLA Land Parcel From Los Angeles County Office of Assessor

The following table summarizes the impervious area derived from Quickbird imagery for each of the three parcels in UCLA. The total area of UCLA campus is 419 acres. In this parcel data provided by the Los Angeles Tax Assessor Office, the three parcels that form the main campus of UCLA account for 373 acres (16,272,221 sq ft)  implying that UCLA owns parcels outside these 3 parcel areas. Overall, of the total 373 acres of main UCLA campus shown in the map above, about 221 acres are impervious or built up. The rest is probably vegetation or other type of land cover.

Table 1: Impervious Area Computed from Quickbird Imagery for UCLA Parcels

Fig 14: Attribute for UCLA Parcel 1 showing Total and Impervious Area

Practical Application

Impervious areas can be used by Counties to assess maintenance fee mandated by EPA to meet certain water quality standards. For this to happen, it is important to identify impervious areas for each parcel in the county. The final result of imagery classification was published on Esri ArcGIS Online, and can be viewed here at http://bit.ly/128BD47. You can also view this on a smartphone. Esri ArcGIS Online is automatically optimized for viewing on mobile devices.

The following is a live map published on ArcGIS Online. Click on parcels to see impervious area.

View Larger Map

The following image is a screenshot of the ArcGIS Online map. 

Discussion

I used different images from different sources to assess the access of impervious surface in Los Angeles. I used UCLA campus as my ground data verification site. All the images inserted in my paper above acted as a tool to examine impervious surface. The comparison between Landsat , Quickbird and Aerial imageries has been vital to my research. The ground data verification at the UCLA campus surprised me with accuracy of Quickbird images. Even Though the resolution of Landsat image lacks clarity, it is used by educational institutions and organization across the world because this is free data provided by USGS. On the other hand , high resolution images from different private satellite companies such as Quickbird is inaccessible to many as they are very expensive.

Conclusion

An in depth examination of classified satellite images and Aerial photography of the Los Angeles area, I came to the conclusion that remote sensing combined with GIS is a powerful tool to calculate and assess impervious surface. Urbanization of Los Angeles has brought some environmental issues which could be examined using remote sensing and GIS.

Works Cited

Barnes, Kent B., John M. Morgan III, and Martin . C. Roberge. Impervious surfaces and the quality of natural and built environments. Department of Geography and Environmental Planning, 2002. Web. 5 Dec. 2012.
Nowak, David J., and Eric. J. Greenfield. “Tree and Impervious Cover Change in U.S. Cities.” Urban Forestry & Urban Greening 11 (2011): 21-31. Www.elsevier.de/ufug. Elsevier GmbH. Web. 18 Nov. 2012.
Wu, Chunxia, Qingfu Xiao, and E. Gregory McPherson. “A Method for Locating Potential Tree-planting Sites in Urban Areas: A Case Study of Los Angeles, USA.” Urban Forestry & Urban Greening 7 (2008): 65-76. ScienceDirect. Www. Sciencedirect.com. Web. 15 Nov. 2012.