**Table 1.** Coding sheet.

Landscape setting refers to the type of landscape in which the greenway is situated and includes the following classification codes: *adaptive reuse* (greenways developed along spaces that served a previous use such as highways and railways); *waterfront* (greenways that run adjacent to water bodies); *complete street* (greenways that are part of multimodal transit corridors); *new-build* (greenways that are conserved or designed as part of new development); and *multiple settings* (greenways located in two or more of the above). A description of each category and associated classification codes is provided in Table 1.

Categories 1–5 were classified based on deductive (a priori) terms drawn from precedent [35]. However, greenway extent and landscape setting were classified inductively based on terms that emerged in the papers under review. In this case, 20 articles were reviewed and classification terms were established based on this sample. Our original review included both study city and the institutional location of the first author origin; however, there was much overlap between the two and we only report on study location.

In two categories—research theme and disciplinary orientation—we diverted slightly from precedent. Drawing upon Bentsen, Lindholst, and Konijnendijk (2008), we use the term "disciplinary orientation" instead of "type of science," as the former is, in our opinion, a clearer description of the intent

and associated coding terms (humanities, natural science, social science, and interdisciplinary/planning) for this review category. In this category, we also use the term "interdisciplinary" instead of "multiple;" and we added "planning" to this classification scheme, as many of the papers qualify as planning studies and many planning studies address both social and ecological concerns.

Based upon the same review precedent, we also adopted the classification scheme of James et al. [40] to depict the main research theme of the paper. Their original scheme included five classification codes: "physicality, experience, valuation, management, and governance." However, we combined "management" and "governance" into a single classification code ("management") because these terms encompass many overlapping ideas, and it was difficult to disambiguate the two. *Physicality* encompasses outcomes related to microclimate, soil, air, and water quality functions and is essentially synonymous with "environmental." *Experience* encompasses people's interaction or contact with green spaces and includes aesthetic, health, and sociocultural dimensions. *Valuation* encompasses links between green space and economic outcomes, and includes topics such as property value and business development. James et al. (2009) also include ecosystem services—human health and well-being benefits of ecosystem functions that are quantified and monetized—in this category; so we classified ecosystem services in both *physicality* and *valuation*. *Management* encompasses planning, design, and governance of urban greenways. In keeping with precedent, we also included *Other* for terms that did not directly classify into a priori categories; however, we removed this classification code from the total count as these terms did not address a research theme. To further minimize risk of misclassification in this category, we systematized the process by using the articles' keywords as the underlying source of data. This had the added benefit of providing quantitative data on the scope of terms associated with each research theme.

For categories requiring little subjective determination (categories 1 to 4), one co-author coded all papers. For categories requiring some subjective determination (categories 5 to 7), both co-authors reviewed all papers and arrived at a shared classification based on definitions and classification codes described above and listed in Table 1.

In addition to these review categories, we also reviewed articles for definitions of urban greenways and applied keywords from these definitions to a word cloud generator. This systematic review provides a foundation for developing a definition and a typology of urban greenways as a subset of greenways writ large. We drew upon Little (1990) [4], Ahern (1995) [1], Hellmund and Smith (2006) [16], and Rupprecht and Byrne (2014) [41] as precedents for the typology which includes descriptions, goals, and examples of five types of urban greenway. Descriptions and goals were based upon review of scholarship and practice. To acknowledge that greenways within each type can be naturalistic or highly constructed, two photographs are included exemplifying each greenway type across this continuum (see Table 7).
