Digital Classification and Mapping of Urban Tree Cover: City of Minneapolis

Transcription

1 Digital Classification and Mapping of Urban Tree Cover: City of Minneapolis FINAL REPORT April 12, 2011 Marvin Bauer, Donald Kilberg, Molly Martin and Zecharya Tagar Remote Sensing and Geospatial Analysis Laboratory Department of Forest Resources University of Minnesota St. Paul, MN Financial support was provided by the City of Minneapolis contract C 2742 to the University as part of a grant to the City from the Minnesota Environment and Natural Resources Trust Fund as recommended by the Legislative Citizen Commission on Minnesota Resources (LCCMR) and the Community Conservation Assistance Grant Program. 1

2 Digital Classification and Mapping of Urban Tree Cover: City of Minneapolis Introduction Tree cover is an important component of urban environments. In addition to the aesthetic values of trees, numerous studies have shown significant economic and environmental benefits and values of urban trees (Galvin et al.; including: Stormwater management: interception of rain, evapotranspiration, reducing runoff and erosion, and increasing the potential for improving water quality. Energy conservation: transpiration and shading reduce air temperatures and saves energy; reduces the urban heat island effect. Air quality: removes air pollutants, including carbon monoxide, sequesters carbon dioxide, and releases oxygen. Economic Value: enhancement of community vitality, stability and property values for residential and business areas. However, unless we can measure and quantify tree cover, we are not in a good position to manage it. Accurate maps and information on the amount of tree and forest cover are not routinely available and it would be expensive to acquire by field mapping methods. Interpretation of aerial photography is an alternative, but the most appropriate imagery, color infrared photography, is generally not available. Recent available high resolution color ortho imagery is early spring, leaf off imagery that is not suitable for mapping tree cover in Minneapolis where many trees are deciduous species. An alternative, and the approach of our project, is to digitally classify high resolution multispectral QuickBird satellite image data that was recently acquired of the Minneapolis area. QuickBird has four spectral bands, blue, green, red and near infrared, at 2.4 meter spatial resolution, with the infrared band being especially useful for mapping vegetation, plus a panchromatic band at 0.6 meter resolution. With pan sharpening of the multispectral bands, a 0.6 meter resolution image can be generated that is higher resolution than the 4 band, 1 meter USDA National Agriculture Imagery Program (NAIP) imagery acquired for Minnesota in the summer of An additional and significant advantage of the QuickBird data is that the entire City is included in a single image. The project objective was to generate a digital land cover classification of the City of Minneapolis in GIScompatible format, with emphasis on mapping the tree cover that can be used by the City to evaluate existing tree cover and potential for additional plantings. Tree cover is defined as the leaves, branches and stems covering the ground when viewed from above. Approach Imagery QuickBird satellite imagery acquired on June 25, 2009 was used for the image classification. The image was clear and cloud free. Natural color and false color (including the near infrared band) images are shown in Figure 1. 2

3 Figure 1. QuickBird imagery of Minneapolis acquired on June 25, 2009; natural color (left) and false color (right). 2

4 In addition, LiDAR imagery acquired in June 2007 was available from the U.S. Army Corps of Engineers. LiDAR (Light Detection And Ranging) is a remote sensing technology using pulses from a laser to measure the distance to the surface, and therefore can be used to generate elevation and height information. This imagery consisted of first return information as well as the last return or bare earth; using the two a normalized digital surface model (ndsm) which depicts height above bare earth (for example of buildings and trees). The horizontal accuracy of the data was roughly 0.5 meters and stated to be better than 1 meter. Its vertical accuracy compared to 33 control points was meters. The LiDAR data included full coverage for the entire City of Minneapolis. Figure 2. Normalized digital surface model (ndsm) imagery of Minneapolis acquired in June As shown in Figure 3 the LiDAR ndsm data corresponds very closely to the buildings and trees, with the height information providing excellent separation of buildings from streets and trees from grass. In the gray scale image in Figures 2 and 3, black is the bare earth surface elevation, and shades of gray to white are increasingly taller objects. 3

5 Figure 3. Enlargements of the LiDAR ndsm (top) and false color QuickBird imagery (bottom). Together these images were used to classify tree canopy and other land cover classes. 4

6 Land Cover Classes The land cover classes are described in Table 1. Table 1. Land cover classes and descriptions. Class Tree Canopy Grass & Shrubs Bare Soil Water Buildings Streets Other Impervious Description The layer of leaves, branches and stems of tree that covers the ground when viewed from above. Lawns and other grass covered areas and shrubs found in parks, golf courses and playgrounds. Areas free of vegetation, primarily in open industrial areas. Lakes and streams. Houses and commercial, industrial and public structures. Streets and highways. Includes driveways, sidewalks, parking lots and other impermeable surfaces that are not obscured by tree cover. In all cases the class is defined as the surface area viewed from above. It should be noted that tree canopies will cover and obscure from view some of the grass, bare soil, streets and parts of some buildings. To take one example, the amount of impervious will by definition typically be less than measured by other methods such as from leaf off high resolution ortho aerial photos in which all impervious surfaces can be seen. Therefore results from the two methods should not be compared. Of the two methods, impervious area measurements from the higher resolution photos should be more accurate. Classification Procedures The primary land classifications were produced using object based image analysis (OBIA) techniques available in ecognition Developer version 8.0. Ancillary software utilized included ArcGIS version and ERDAS Imagine version Additional customized routines were written in Python version 2.5 scripting language to support processing as required. Shapefile information was provided by the City of Minneapolis to help identify streets, buildings, roads and highways and water features. The following principle steps were followed to implement the project: The 2.4 meter resolution multispectral QuickBird imagery was pan sharpened using the 0.6 meter panchromatic band and subtractive resolution in ERDAS Imagine. QuickBird Imagery was georeferenced utilizing the available RPC files and a 30 meter DEM layer. LiDAR data were georeferenced to match the QuickBird imagery. A customized Python script was used to divide the georeferenced imagery into 750 x 1000 meter tiles with 10 percent overlap for further processing. This step created 262 individual tiles. 5

7 The street layer was buffered in ArcGIS by 3 meters to create a polygon shapefile for subsequent use in ecognition. The rule set was created using these process steps: ecognition workspace of all 262 tiles was created with a customized load procedure. Imagery was examined to locate a representative tile. Supportive image layers such as Normalized Difference Vegetation Index (NDVI) and Lee s Sigma Edge Extraction were created to aid classification efficacy. Image objects were generated representing buildings, roads and water features from shapefiles and classified as such. Since LiDAR data were available the images were first segmented into tall and short features. Remaining portions of the image were classified utilizing algorithms available in ecognition taking advantage of spectral information as well as other elements of image interpretation such as context, shape, size, site, association, pattern, shadows and texture. Classification was exported from ecognition into a TIF raster file. The rule set was fine tuned and tested on additional random tiles distributed throughout Minneapolis. The final rule set was used to classify all the tiles using ecognition Server. Individual classified tiles were joined into a single mosaic using geometric seam lines in ERDAS Imagine Mosaic Pro. The accuracy of the resulting classification was assessed in ERDAS Imagine using 1,413 stratified random points. The classification mosaic was then manually examined and edited to eliminate classification errors. Error corrections were re run in ecognition Server to incorporate the corrections. The final land cover mosaic was manipulated by ERDAS Imagine and ArcGIS into the output geodatabase utilizing both raster and vector forms of the data. A Python script was written to summarize classification information into various shapefiles such as parcels and neighborhoods. Key to the classification was use of an object based image analysis approach in which the imagery was first segmented into objects with similar pixels based on the spatial, as well as the spectral radiometric (color) attributes (Figure 4). Research has shown that it is the best approach for classification of high resolution imagery (Blaschke, 2010; Platt and Rapoza, 2008). Objects include more information than individual pixels, enabling the ability to take advantage of all the elements of image interpretation, particularly spatial information, including shape, size, pattern, texture, and context. Context is especially useful. Humans intuitively integrate pixels into objects and use contextual relationships to interpret images and draw intelligent inferences from them. Ancillary data such as GIS layers, for example, of streets and water bodies, can also be incorporated into the decision rules. 6

8 Figure 4. Major steps in the object based image analysis approach to image classification. The object based image analysis process in ecognition can broadly be split into two components, segmentation and classification. Segmentation primarily uses spectral information about individual pixels in the imagery to combine them into larger image objects or segments. As an example, individual pixels which comprise the roof of a building with similar brightness, normalized difference vegetation index (NDVI) and color values are combined to form an image object that represents the building. Other scaling information can be specified to regulate the size range of the desired objects. Once these image objects are created, they can be classified using a multitude of decision rules which utilize not only their spectral characteristics but also spatial information such as shape, size, proximity to other object types, texture, and context. The overall process is dependent on the quality of the initial segmentation into image objects. Accuracy Assessment Accuracy assessment was performed after the tiles were edited for misclassifications by generating stratified random points across the image and comparing the classified results to reference imagery of color ortho photos provided by the City and imagery from ArcGIS online. Stratified random point selection assures each class will be weighted proportionately to the total number of points in that class across the image. There were 1,413 points in the sample (Figure 5). 7

9 Figure 5. Locations of 1,413 sample points used for the final accuracy assessment. The assessment points are displayed large enough to be visible on the map, but in reality these points are geometric points that ERDAS Imagine randomly designates in the image. Figure 6 is a close up showing one of the assessment points randomly selected by the software. As is quite typical in an urban setting, it can be quite difficult to determine just what land cover is at a given point. In this example the shadowed areas exacerbate the task of determining if this point was tree canopy, grass or building roof. 8

10 Figure 6. Example of a randomly selected sample point for accuracy determination. Situations like these occurred quite commonly throughout the process. In an attempt to assure an effective process, mismatches were reviewed to confirm the interpretation of the reference image. 9

11 Results The final classification raster image is depicted in Figures 7 and 8. The results in Table 2 show that 31.5 percent of the area of the City is tree canopy. Figure 7. Land cover classification of Minneapolis. 10

12 Figure 8. Detailed view of the classification of the area near the intersection of Lake Street and Hennepin Avenue. Table 2. Tabulation of the percent area of each of the seven land cover classes. Land Cover Class Percent 1. Tree Canopy Grass and Shrubs Bare Soil Water Buildings Streets Other Impervious 17.3 A comparison of a high resolution aerial photograph to the image classification illustrating the high correlation between the classification and a reference image and providing a qualitative indication of the classification accuracy is shown in Figure 9. 11

13 Figure 9. Comparison of high resolution aerial photo (top) and image classification (bottom) of an example subset of the image near the intersection of W 36th Street and Calhoun Boulevard. 12

14 We also conducted quantitative assessments of the classification by comparing a stratified random sample of points from the classification to high resolution aerial photography. The results are presented in the form of a contingency table or error matrix; further details on interpretation of error matrices and the statistics derived from them are in Appendix 1. Our previous work had shown that automated object based classification, while effective, can still result in obvious misclassifications. To reduce this impact, the classifications were compared to reference imagery and were edited manually where necessary. As an example, grass near freeways can become quite dry and take on the appearance of impervious cover. Larger objects with height such as trucks and buses on roads are often interpreted as a building. We assessed the accuracy after these corrections were made and the results are shown in Table 3. The overall accuracy was 91.9 percent. Table 3. Classification accuracy following corrections. Classification Tree Canopy Grass/ Shrub Bare Soil Reference Data Water Buildings Streets Impervious Total User s Accuracy (%) Tree Canopy Grass/Shrub Bare Soil Water Buildings Streets Impervious Total Producer s Accuracy (%) Overall accuracy: 1299 / 1413 =91.9% 95 percent Confidence Interval: % Kappa Statistic: 0.90 Other Analyses Extraction of Tree Canopy The land cover classification raster data is the primary output of this project, but additional items that could be derived from it were requested by the City. The first of these was to extract only the tree canopy areas. This was accomplished with a raster extraction and is provided as a layer in the output file geodatabase. This layer can be utilized by the City to add tree canopy to other GIS applications. The raster was then used to create a polygon feature class of the same information. The polygons were simplified to smooth their boundaries and the resulting feature class was stored in the geodatabase. 13

15 The City would like to use this work to determine locations for potential tree planting sites. There are two requests which related to this effort. The first of these highlights parcels which have high amounts of tree canopy vs. those with low tree canopy. To create this analysis, the classified raster was used in conjunction with the parcel layer provided by the City to extract the area of the various types of land cover into each parcel. These fields were then used to calculate the percent of land cover along with other key information. For each type of land cover the following three fields were added to the parcel: An AREA field showing area in square meters of that class. A PCT field showing the percentage that area represented of the total parcel area. A CODE field which can be used as a legend code, for example 0 to 20 was assigned if the percentage field fell in the range of 0 to 20 percent inclusive. There are five ranges that could be assigned. Once the percentages for the parcel were calculated, a GIS analysis was done for all parcels in the neighborhood to determine the parcels ranking by tree canopy percent. The lowest 10 percent of parcels in the neighborhood were given a PercentileCode of 1, the next lowest 10 percent were given a PercentileCode of 2 and so on, until the top 10 percent of parcels were given a PercentileCode of 10. This analysis can be utilized in neighborhood meetings to look at the distribution of parcels and determine parcels to prioritize for further examination. Figures 10 and 11 below are some samples of these maps for Hale, Bottineau and Ventura Village neighborhoods. Maps can be created by the City from this layer for any neighborhoods or using any percentile groupings that are desired. Figure 10. Lowest and highest tree canopy parcels for Hale neighborhood. 14

16 Figure 11. Lowest and highest canopy percent parcels for Bottineau (top) and Ventura Village (bottom) neighborhoods. 15

17 Right of Way Analysis for Potential Tree Planting The second of the efforts to identify potential sites for additional planting uses a right of way analysis. The Planimetrics layer provided by the City was used to extract all SYMBOL_NAMES with a value of CWCBTOPO. These represented clockwise drawn (in most instances) curb segments. A right buffer of 3 meters was drawn to represent the right of way from the curb. The tree polygons from previously created layers were used to erase any right of ways which already had tree canopy cover. The remaining right of way segments were filtered to include only those segments which were 10 meters or more in length. These represent right of way which might be used for potential planting sites. The sample map below (Figure 12) depicts these segments along with shorter segments of right of way and tree canopy all superimposed over imagery of the City. Figure 12. Sample of right of way segments more than 10 meters in length. Using Tree Cover Classification for Energy Conservation through Strategic Tree Planting About half of the unwanted summer heat in residential homes comes from sun shining through windows, mainly on the west and east sides. Shade trees planted in adequate distance to the west (as first priority) and east (as second priority) of residential homes can significantly reduce air conditioning demand (Minnesota Department of Commerce), thereby reducing energy costs and greenhouse gas emissions. The digital classification of urban tree cover can be used to advance energy conservation in 16

18 the City of Minneapolis by informing existing city tree planting programs as well as residents and home owners about homes that currently lack shade trees in relevant distance and direction from buildings. A GIS analysis to derive this information is currently being completed. Figure 13 provides a brief illustrated description of the procedure. 1. Buildings in the City are overlaid on the QuickBird satellite image used for the classification. The buildings have the attributes of the city parcels, including address and owner name. Larger blocks are houses, smaller blocks are garages. 2. A shadow is determined for each home (in this case to the west). The shadow distance from the home is in the effective range for tree shading. The shadow is divided to the area within the parcel (light blue), where tree planting is the responsibility of home owners, and the area outside the parcel (darker blue), where tree planting (on boulevards) is within the city s responsibility. 17

19 3. Next, the tree cover classification is overlaid on the shadow areas determined previously. In this example, classified tree cover is colored green (other classes, e.g., grass/ shrubs or impervious surfaces, were omitted). 4. Tree cover is clipped from the shadow area. Homes near a shadow that remained largely intact after this operation are ones that have no trees in relevant distance to their west. Figure 13. Summary of GIS analysis procedures for identifying potential tree planting sites. The output of this operation will be an inventory of addresses in the city, where buildings have currently no tree shade in locations that are relevant for energy conservation. This inventory will include homes where trees planted by the city on boulevards would have energy conservation impact, as well as homes where trees planted on private property would have a similar impact. 18

20 Web based Maps The City also requested we create a web based mapping application that could be used to query and display tree canopy information generated by this work. We elected to utilize the FlexViewer (similar to Google Maps) application which can be downloaded from the ESRI website at no cost. This application works seamlessly with ArcGIS server and also allows use of ESRI owned data such as global imagery and street maps along with the layer information we generated. Examples of the levels of search and display available are shown in Figures 14 and 15. Figure 14. Web based maps and statistics from FlexViewer. The opening screen provides information on using the application, plus a summary report describing the project. In summary, the FlexViewer application provides easy to use navigational tool bars to zoom and pan across the images and query the database for amount of tree canopy by parcel and neighborhood. This application is currently hosted by the University of Minnesota for demonstration purposes. The web site URL is: 19

21 Figure 15. Using the Search dialog box allows the user to select neighborhoods graphically or with text (top) and then to display their classification statistics (top). Similarly, parcels can be selected graphically or by entering the parcel identification number (bottom). For neighborhoods the percents of all classes, plus possible tree canopy, are listed. For parcels the percent existing and possible tree cover are listed. 20

22 Discussion The overall accuracy of 91.9 percent and user s and producer s accuracies of 93.5 and 97.2 percent for tree canopy meets the expected accuracy goals of the project. The primary errors are some confusion between trees and grass/shrub, between buildings and impervious, and streets and impervious. The single largest area of confusion was the 23 points out of 284 grass and shrub points that were erroneously classified as tree canopy. Although the pixel size of the pan sharpened QuickBird imagery is approximately 0.6 meters, the lower limit for size detection of individual objects is between 2 and 3 meters square. More specifically, to improve the spatial resolution of the multispectral imagery we used a pan sharpening process which takes the spectral information from the 2.4 meter multispectral pixels and distributes it mathematically to the higher resolution 0.6 meter panchromatic pixels to create 0.6 meter multispectral pixels. While the pixel size is 0.6 meters, small or narrow objects (e.g., a sidewalk) may not be resolved in the imagery or classification. Another limitation to the study was the temporal mismatch between the QuickBird and LiDAR imagery. The QuickBird image was acquired approximately two years after the LiDAR data and this resulted in several inconsistencies in the classification. An example is trees present in the LiDAR but subsequently removed prior to the QuickBird image acquisition. Where the tree had overhung a street, the classifier interpreted the existing height as an impervious object and classified it as a building. If it overhung grass, it was interpreted as a tree. Where found, these errors were manually corrected. More difficult to correct was the reverse situation where a new tree planting did not have matching LiDAR height information. Many of these were classified as grass/shrub areas. As a final note, for this analysis, tree canopy was allowed to grow over any street and road layers and was not truncated by the latter. The exception to this is the parcel analyses. In the parcel analysis, only the percentage of the tree canopy that actually falls within the border of the parcel is included. Since parcels do not extend into the street, the canopy does not as well. Summary and Recommendations Multispectral QuickBird satellite imagery acquired on June 25, 2009 and LiDAR data acquired in June 2007 were classified into land cover classes of trees, grass and shrubs, buildings, roads, impervious, water, and bare soil using an object based image analysis approach. The overall classification accuracy was 91.9 percent. Citywide, the amount of tree canopy cover was found to be 31.5 percent. The most serious limitation of the classification was the temporal mismatch between the optical and LiDAR imagery which caused confusion between objects removed or added between the image acquisition dates. This limitation could be overcome by additional assessments with more recent, matched imagery. LiDAR data will be flown again in the spring 2011 and will be available for use sometime later in We suggest the City consider re analyzing tree canopy when the new LiDAR data is available and using satellite imagery matching the LiDAR acquisition dates as closely as possible. 21

23 Finally, we present examples in Figures 16 and 17 of the potential of what is possible in further GIS analyses by the City. The examples summarize the percent tree cover by parcel and neighborhood, but could be by zoning district or other areas of interest to the City. The first in Figure 16 is the existing or current tree canopy and possible tree cover summarized by parcel. Possible tree canopy is defined as areas with grass, bare soil or impervious surface (e.g., parking lots) where it is theoretically possible to plant trees. Figure 17 depicts the existing and possible percent tree canopy for all neighborhoods of the City. The next step would be to define criteria for identifying preferred area for adding trees. The capability exists in the data for the City to create any definition that is desired and do its own GIS analysis of where trees might be planted. Many factors will determine when and where trees are planted and maintained, but an urban tree canopy assessment is an essential first step in determining where trees can be planted if the requisite social political and financial capital exits. Figure 16. Example of parcel level GIS analyses summarizing existing and possible percent tree cover near Lake of the Isles. 22

24 Figure 17. Example of further GIS analyses with summaries of percent existing and possible tree cover by neighborhood. The classification maps and statistical data are available in a GIS database and a web based mapping application for further analysis by the City. 23

25 References Blaschke, T Object based image analysis for remote sensing. ISPRS Journal of Photogrammetry and Remote Sensing, 65(1): Congalton, R.G A Review of Assessing the Accuracy of Classifications of Remotely Sensed Data. Remote Sensing of Environment, 37: Foody, G Status of land cover classification accuracy assessment. Remote Sensing of Environment. 80: Galvin, Galvin, Morgan Grove, and Jarlath O Neil Dunne. Urban Tree Canopy Fact Sheet. FOS_files/UTC_FactSheet.pdf. Minnesota Department of Commerce Energy Information Center. Save Energy with Trees. dscaping.pdf. Platt, R.V. and L. Rapoza An evaluation of an object oriented paradigm for land use/land cover classification. Professional Geographer, 60(1): Acknowledgements We would like to thank Jarlath O Neil Dunne at the University of Vermont for his invaluable assistance during this project and for offering the use of his university s ecognition Server to process the 262 individual tiles, saving us considerable hours of personnel and processing time. 24

26 Appendix 1: Description of Accuracy Assessment Measures Accuracy assessment provides information on reliability and usefulness of classification needed to support decision making and management using maps and data derived from remote sensing data. Quantitative, objective assessment of the classification accuracy, defined as the agreement between a standard or reference map or data (assumed to be correct) is a critical part of any serious remote sensing project (Congalton, 1991; Foody, 2002). The standard method of communicating the results is in a contingency or error matrix comparing the classification results to reference data for a random sample of points. The error matrix is the starting point for a series of descriptive and statistical techniques to evaluate accuracy. Row totals equal the number of pixels in the reference data classes and the column totals equal the number of pixels assigned to each class. The diagonal show agreement between reference data and the classification (i.e., correct classification); points in off-diagonal cells are incorrect classifications. Overall accuracy is the number of correctly classified points divided by the total number of points in the sample. Columns include the errors of commission and rows include the errors of omission. Commission errors occur where a classified object (e.g., trees) is not actually that class. Omission errors are where the object was not classified as that class. Often the commission and omission errors are presented as user s accuracy and producer s accuracy. User s accuracy is based on the commission errors and is the probability that a pixel or object on the map actually represents that class on ground. Producer s accuracy is based on the omission errors and is the probability of a reference site being correctly classified. The Kappa statistic is a discrete multivariate technique to interpret the results of a contingency matrix. The Kappa statistic incorporates the off diagonal observations of the rows and columns as well as the diagonal to give a more robust assessment of accuracy than the overall accuracy. The Kappa statistic is computed as the sum of the diagonal multiplied by the sum of each row multiplied by the sum of each column divided by the sum of each row multiplied by the summation of each column. It is a more conservative estimate of accuracy that measures the proportional (or percentage) improvement by the classifier over a purely random assignment to classes; in other words it removes the contribution of chance agreement to the accuracy. Possible causes of classification errors include: spectral-radiometric similarity of classes leads to confusion between them, alignment or registration errors, and incorrect reference, including due differences in the time of acquisition of imagery and reference data. 25

27 Appendix 2: Summary of Deliverables The following items are included in the ESRI file geodatabase provided to the City: 1. A classified raster with the following land cover classes: Land Cover Class Raster Value Tree Canopy 1 Grass and Shrubs 2 Bare Soil 3 Water 4 Buildings 5 Streets 6 Impervious 7 2. A vector layer of polygons derived from the classified raster with the same land cover codes. 3. A raster layer which contains only the tree canopy. This was derived from the classified raster selecting only raster values of 1. The layer can be used by City GIS applications to depict tree canopy as a layer in other GIS analyses. 4. A polygon layer depicting potential right of way planting areas. This layer was derived from the planimetric data. Code values of CWCBTOPO were selected from the main file and extracted. A right hand buffer of 3 meters was generated from this data. Overhead tree canopy was utilized to erase curbs currently under trees. The remaining segments were filtered to include only those segments whose length was more than 10 meters. 5. Neighborhoods Neighborhoods were derived from the original shapefile provided by the City. Fields were added to show total percent of each land cover type as well as potential tree canopy percentages. The NeighborCode field was added, which is utilized in the Parcel data to join parcels to neighborhoods. 6. Parcels Parcels were derived from the original shapefile provided by the City. Fields were added to show total percent of each land cover type as well as potential tree canopy percentages. In addition, each parcel is coded with the appropriate Neighborhood code of which it is a part and a Percentile code which describes which decade grouping its tree canopy percent cover falls into within that neighborhood. The codes range from As an example, a code of 1 indicates this parcel s tree canopy percent places it within the lowest 10percent of all parcels in the neighborhood. A code of 10 indicates its canopy coverage places it in the highest 10percent of all parcels in that neighborhood. 7. A web application ( allows users to select neighborhoods or parcels and display tree canopy percentages for those units. This application was created utilizing the ESRI downloadable FlexViewer utility. The following data layers were created and are managed by an ArcGIS server instance: a. Neighborhood layer with total tree canopy summarized within that neighborhood b. Parcel layer with tree canopy percentage for each parcel calculated. 26