A briefing paper prepared by the Urban and Community Forestry Committee of the National Association of State Foresters
The NASF Urban and Community Forestry Committee is grateful to Dr. David J. Nowak, Project Leader USDA Forest Service, Northern Research Station, and Dr. Monica Lear, Deputy Associate Director, Urban Forestry Administration, Washington DC Department of Transportation, for their contributions to this paper.
Urban forests are comprised of trees in residential areas, parks, public areas and along streets and provide important forest cover benefits to cities large and small (Raciti, et al). The urban forest provides economic, environmental, public health, and social benefits that include cleaner air and water, decreased heating and cooling costs, reduction of the urban heat island effect, enhancement of storm water treatment practices, increased livability, aesthetic value and reduced ultraviolet radiation (http://nrs.fs.fed.us/urban/utc/data) to communities. Urban tree canopy is defined as the layer of leaves, branches and stems that cover the ground when viewed from above.
Urban tree canopy (UTC) assessment is a tool to help urban and community forestry programs and communities increase urban tree canopy cover to maximize the benefits provided by the urban forest. It can also provide an important opportunity to bring community members together to develop a vision of the future urban canopy. UTC is an easily understood measure of the community's overall success in meeting agreed upon canopy goals
The Value of Urban Tree Canopy Assessment
Urban and community forests can be considered part of the "green infrastructure" that complements our grey infrastructure and should be managed with equal importance. Urban tree canopy assessments can help a community determine how much of their land area is covered by trees, location of those trees and where there are new opportunities to plant trees. UTC assessments also determine the amount and location of impervious cover in a community. The power of an UTC assessment is the GIS framework where it resides. Canopy cover can be assessed by watershed, zoning or land use category, political boundary, neighborhood, business district, census tract or individual parcel. Results of analyses focus on specific community needs and provide a blueprint of available and unavailable planting opportunities.
A community needs to determine available resources prior to conducting a UTC assessment. Remotely sensed sub-meter resolution digital imagery is necessary as well as technical expertise to process and analyze the data. Imagery and analysis can be expensive so smaller communities should work with state and federal entities to leverage resources or acquire available grants. However, analysis may be equally expensive depending on the level of land cover information sought.
There are two primary ways of assessing the structure or composition of the urban forest: 1) a top-down approach, which involves aerial assessments of canopy cover, and 2) a bottom-up approach, which involves collecting field data on tree species composition and physical attributes of the vegetation. The top-down approach can be relatively easy and quite accurate, but is limited to measuring the quantity and distribution of canopy layers. The bottom-up approach involves collecting field vegetation data in the field on the vegetation and may provide more detailed information needed to assess urban forest structure and ecosystem services, and to aid in urban forest management e.g., species composition, number of trees tree locations, tree sizes and tree health). Specialized software tools developed by the U.S. Forest Service such as the i-Tree suite (Eco, Streets and Hydro) are available at no cost. This paper focuses on an overview of top-down approaches to assessing urban tree canopy cover.
Methods of Urban Tree Canopy Cover Assessment
There are three common methods of assessing urban tree canopy cover: 1) NLCD analyses; 2) high-resolution image analyses; and 3) aerial photo interpretation, each with distinct advantages and disadvantages.
NLCD analyses - The National Land Cover Database (NLCD) has recently released tree and impervious cover maps (30-meter resolution) for the entire lower 48 states with percent tree and percent impervious cover estimated for each pixel. These maps and data are available for free at http://www.mrlc.gov/ and summaries of the cover data for each state down to the community level are being produced.
Advantages - 1) free, 2) wall-to-wall coverage of lower 48 states, 3) maps of canopy cover distribution, 4) can integrate with GIS
Disadvantages - 1) relatively course resolution (cannot see individual trees), 2) tends to underestimate tree cover by approximately 10% on average (see Greenfield and Nowak, 2009), 3) not designed for local scale (e.g., city) analyses (better for regional analyses).
High-resolution imagery - This approach uses high resolution in-leaf aerial digital imagery (typically less than one meter pixel resolution) to classify each pixel as to its appropriate cover class. This type of analysis produces detailed maps of canopy cover where tree canopies are visible. Accuracy of these maps typical 75-85% for tree cover depending upon the processor. Additional hand editing can increase accuracy. Using Light Detection and Ranging (LIDAR) data, which provide height information, can increase accuracy to over 90%.
Advantages - 1) high resolution cover map, 2) good estimates of cover amounts and locations, 3) integrates well with GIS, 4) can process data for sub-area analyses (e.g., tree cover by neighborhood), 5) can locate potentially available spaces to plant trees.
Disadvantages - 1) may be costly if digital images and process data to produce maps must be purchased and analysis must be subcontracted 2) cloud cover can be an issue, requiring multi-date images, 3) significant effort and time needed to produce quality maps..
Photo-Interpretation - This approach uses existing digital aerial images and lays of series of random points that are interpreted to determine the cover type at each point center. This process produces statistical estimates of cover with known error. A sample of 100 points will produce an estimate with a standard error of about 4.6% (assuming 30% canopy cover) and can be interpreted in about an hour. A sample of 1,000 points will produce an estimate with a standard error of about 1.4% (assuming 30% canopy cover) and can be interpreted in about a day. Photo interpreters can zoom in or out on screen to help with interpretation. Photo interpretation has been used for accuracy assessments of the other top-down methods.
Advantages - 1) low cost - most images can be acquired freely (e.g., Google Earth or other public sources); cost involves set up and interpretation time, 2) cover assessment can be done quickly, 3) accuracy can be increased by adding more points and calculated quickly. 4) can produce sub-area analysis and maps (e.g., tree cover by neighborhood) if sample size is large enough), 5) in-leaf photography is recommended, but high resolution leaf-off imagery may be with the assistance of a trained interpreter.
Disadvantages - 1) does not produce detailed cover map, 2) photo-interpreters can create errors through misclassifications (training and quality control is recommended).
Applications of Urban Tree Canopy Assessment
Cities of all sizes are looking for a way to optimize and increase ecosystem services to support a livable community for residents. Planting trees and maintaining tree cover is a way to enhance environmental benefits and provide a platform to build community support to increase the environmental health and economic vitality of places where we live, work and play.
The assessment of geographic data can help inform the municipal and state decision making process at a range of scales, from prioritizing communities within a state to targeting individual properties for tree plantings by determining the amount of existing and potential tree canopy (Troy et al). Existing and potential tree canopy data used in conjunction with ownership data can help a community develop an overall urban forestry management strategy that stakeholders can use to plan budgets and allocate resources; it may also be used to support specific decisions such as issuance of a stormwater permit based on the amount of impervious area in a neighborhood or subdivision.
UTC assessments may also serve as a valuable tool as a required component of a city comprehensive plan to aid in developing forest conservation ordinances in light of forest loss associated with growth and in incorporating "strategic tree planting" in State Implementation Plans as a means to help meet air quality standards set by the U.S. Environmental Protection Agency. UTC assessments in city planning can also enhance and direct local and regional landscape watershed restoration efforts and track tree planting to address Clean Water Act regulations and permit requirements.
- Cappiella, K., T. Schueler and T. Wright. 2005. Urban Watershed Forestry Manual, Part 1: Methods for Increasing Forest Cover in a Watershed. United States Department of Agriculture, Forest Service Northeastern State and Private Forestry and the Center for Watershed Protection, Newtown Square, PA.
- Greenfield, E.J., D.J. Nowak, and J.T. Walton. 2009. Assessment of NLCD tree and impervious surface cover estimates. Photogrammetric Engineering and Remote Sensing. Nov. issue
- Nowak, D.J., R.A. Rowntree, E.G. McPherson, S.M. Sisinni, E. Kerkmann and J.C. Stevens. 1996. Measuring and analyzing urban tree cover. Lands. Urban Plann. 36:49-57. http://www.treesearch.fs.fed.us/pubs/18804
- Raciti, S., M. F. Galvin, J. M. Grove, J. P. M. O'Neil-Dunne, A. Todd and S. Claggett. 2006. Urban Tree Canopy Goal Setting: A Guide for Chesapeake Bay Communities, United States Department of Agriculture, Forest Service, Northeastern State & Private Forestry, Chesapeake Bay Program Office, Annapolis, MD.
- Troy, A.R., J.P.M. O'Neil-Dunne and J.M. Grove. 2006. Geospatial Assessment in Support of Urban & Community Forestry Programs. USDA Forest Service State and Private Forestry and the University of Vermont. 8 pp.
- Walton, J.T., D.J. Nowak, and E.J. Greenfield. 2008. Assessing urban forest canopy cover using airborne or satellite imagery. Arboric. Urb. For. 34(6):334-340. http://www.treesearch.fs.fed.us/pubs/19515