The basis for the urban tree canopy assessment is a high-resolution land cover analysis, provided by consultants at the University of Vermont Spatial Analysis Laboratory. The analysis, which is documented in a separate report, involves the integration of several input data sources:

  • National Agricultural Imagery Program (NAIP) leaf-on imagery (2017)
  • Cuyahoga County orthophotography leaf-off imagery (2017)
  • LIDAR-derived surface models (Northeast Ohio Regional Sewer District, 2017; Cleveland Metroparks, 2018)
  • Other local data, including building footprints and parcels.

The resulting base land cover data includes the following ten classes:

  1. Building
  2. Road
  3. Other Pavement
  4. Bare Soil
  5. Water
  6. Vegetative Cover
  7. Tree Canopy Over Vegetation
  8. Tree Canopy Over Building
  9. Tree Canopy Over Road
  10. Tree Canopy Over Pavement

These ten classes were further condensed to a set of seven classes, where all Tree Canopy classes (TC) are combined to a single “Tree Canopy” category. Using elevations above-ground, derived from a LIDAR generated Digital Surface Model (DSM), trees are distinguished from smaller “shrubs” by using a minimum height measurement of eight feet. Although efforts are made to manually identify smaller trees. LIDAR does generally require a minimum area of 30 square feet to be captured. This could result in some smaller diameter trees not being captured in the data set.

“Possible Canopy” (P) is a land cover combination consisting of areas where trees have not yet been planted but could be, including grass/shrub areas, bare soil, and even parking lots.

Another land cover class, which can be looked at from ground level, can be a combination of various land covers (building, road, pavement) and classified as Imperviousness (I). Impervious cover is a key contributor to increased stormwater runoff and urban heat island effect.

Tree Canopy Metrics Methodology

Using the land cover data, various summary metrics were tabulated for the County as a whole, its 59 individual communities, City of Cleveland neighborhoods, census tracts, census blocks, parcels, and watersheds. Those results form the bulk of this report. Additional details are available in spreadsheet and GIS formats on the downloads page.

Users will note that some of the data from 2011 has been revised slightly due to advances in algorithm development, better data, and data from multiple time periods. Please refer to the Land Cover report from the University of Vermont for further details.

Another item important to note is that all of the land cover estimates are subject to a degree of error. As of publication of this report, the margin of error for the 2017 measurements has not yet been calculated, although a countywide margin of error of +/- 1.5% for the tree canopy class was calculated for the 2013 study data. Due to improved measurement methods and higher quality input data, the accuracy of the 2017 data is expected to improve. However, for data analysis purposes a 1.5% margin of error for 2017 data will be