Jeffrey T. Walton

Subpixel Urban Land Cover Estimation: Comparing Cubist, Random Forests, and Support Vector Regression

Three machine learning subpixel estimation methods (Cubist, Random Forests, and support vector regression) were applied to estimate urban cover. Urban forest canopy cover and impervious surface cover were estimated from Landsat-7 ETMimagery using a higher resolution cover map resampled to 30 m as training and reference data. Three different band combinations (reflectance, tasseled cap, and both reflectance and tasseled cap plus thermal) were compared for their effectiveness with each of the methods. Thirty different training site number and size combinations were also tested. Support vector regression on the tasseled cap bands was found to be the best estimator for urban forest canopy cover, while Cubist performed best using the reflectance plus tasseled cap band combination when predicting impervious surface cover. More training data partitioned in many small training sites generally produces better estimation results.

Assessing urban forest effects and values, New York City's urban forest

An analysis of trees in New York City reveals that this city has about 5.2 million trees with canopies that cover 20.9 percent of the area. The most common tree species are tree of heaven, black cherry, and sweetgum. The urban forest currently stores about 1.35 million tons of carbon valued at

Assessment of 2001 NLCD Percent Tree and Impervious Cover Estimates

24.9 million. In addition, these trees remove about 42,300 tons of carbon per year (

Assessing urban forest effects and values, Washington, D.C.'s urban forest

779,000 per year) and about 2,202 tons of air pollution per year (

Difficulties with estimating city-wide urban forest cover change from national, remotely-sensed tree canopy maps

10.6 million per year). The structural, or compensatory, value is estimated at

Assessing urban forest effects and values, Casper's urban forest

5.2 billion. Information on the structure and functions of the urban forest can be used to improve and augment support for urban forest management programs and to integrate urban forests within plans to improve environmental quality in the New York City area.

A Ground-Based Method of Assessing Urban Forest Structure and Ecosystem Services

The 2001 National Land Cover Database (NLCD) tree and impervious cover maps provide an opportunity to extract basic land-cover information helpful for natural resource assessments. To determine the potential utility and limitations of the 2001 NLCD data, this exploratory study compared 2001 NLCD-derived values of overall percent tree and impervious cover within geopolitical boundaries with aerial photo interpretation-derived values for the same areas. Results of the comparison reveal that NLCD underestimates tree cover and to a lesser extent, underestimates impervious cover. The underestimate appears to be consistent across the conterminous United States with no statistical differences among regions. However, there were statistical differences in the degree of underestimation of tree cover among mapping zones and of impervious cover by population density class.

Projected Urban Growth (2000 - 2050) and Its Estimated Impact on the US Forest Resource

An analysis of trees in Washington, D.C. reveals that this city has about 1,928,000 trees with canopies that cover 28.6 percent of the area. The most common tree species are American beech, red maple, and boxelder. The urban forest currently store about 526,000 tons of carbon valued at

Assessing Urban Forest Canopy Cover Using Airborne or Satellite Imagery

9.7 million. In addition, these trees remove about 16,200 tons of carbon per year (

Effect of Plot and Sample Size on Timing and Precision of Urban Forest Assessments

299,000 per year) and about 540 tons of air pollution per year (