A TYPOLOGY OF POTENTIAL SITES FOR GREEN INFRASTRUCTURE DEVELOPMENT IN THE GOWANUS CANAL WATERSHED FOR THE GOWANUS CANAL CONSERVANCY SPATIAL THINKING WITH GIS Leigh Brown | Lena Greenberg | Taylor Drake
Image from the Gowanus Canal Conservancy
GOWANUS CANAL CONSERVANCY x THE NEW SCHOOL: A COLLABORATION The following report is a suitability analysis performed for the Gowanus Canal Conservancy, a community-based non-profit organization that serves as the environmental steward for the Gowanus Canal Watershed. The analysis takes into account seven variables and three criteria for the proposal of green infrastructure interventions for three land use types. Maps for each of these variables and criteria were created in ESRI ArcGIS. Most and least suitable sites for public, private, and privately-owned public ecologically productive development. In the end, a few specific lots were identified for further research. The hope is that this tool will be useful for both the Conservancy and their current and future partners who are similarly invested in maintaining an open, clean, and alive watershed.
PURPOSE
LITERATURE Precedents
This project consisted of a suitability analysis of potential sites for ecologically productive development in the Gowanus Canal Watershed. Layering geographic features, existing and potential land uses, and demographic data resulted in analysis comprehensive enough to apply to developers and designers alike. Such analysis can assist them to achieve and maintain economic sustainability and ecological resilience in the watershed. Our overarching research question was: which public, private, and privately owned public sites in the Gowanus Canal Watershed are most suitable for ecologically productive development? Though future developers and designers will address individual lots depending on their projects, our final suitability analysis compiles relevant factors that can inform decisions.
Miller et al. (1998) discuss the complexity of doing a suitability analysis. The authors’ process included gathering and identifying information about the study area, classifying variables and criteria, and using GIS to analyze this information (Miller et al., 1998). Their study addresses a rural part of the American Southwest, where the primary concern is the interaction between development and wildlife that inhabit the area. While the factors that contribute to this analysis are quite different from ours, the ultimate goal is similar—to integrate green space and biodiversity with a changing built environment. Because not all criteria are independent of one another, the authors of the study take this opportunity to rank variables based on their relationships as well as the data that comprise each variable. The authors also cite the “inherent impurities” of making decisions about weighting variables and quantifying relationships that are often qualitative (Miller et al., 1998). Part of the reason the overarching task of this study is so complex is that aims to integrate biophysical characteristics with social geographic data. However, the authors conclude that use of GIS facilitates this multifaceted approach to suitability analysis.
RATIONALE A lot is changing in the Gowanus Canal Watershed and there are many disparate visions for how it should look, each in different stages of ideation. Stakeholders need a centralized space for information or series of GIS maps to reference to inform development and feature green infrastructure. Such centralization of information and creation of visuals could reveal overlap and invite collaboration. Planning forums must take into account the diversity of possibilities and decide which plans maximize and stack benefits. The Gowanus Canal Conservancy, an organization that sits at the nexus of these projects and plans, has a vision of a neighborhood that is cleansed of its past industrial pollution, home to a vibrant wildlife community, a destination for public recreation, and is bustling with industrial and commercial activity. Performing a suitability analysis of existing and potential sites for the Conservancy enabled the categorization of potential sites for green infrastructure development.
In a similar suitability study for multi-purpose greenways in Concord, North Carolina, Conine et al. (2003) suggest greenways can balance needs for environmental protection, recreation, and growth in areas facing increased development interest. Like the Miller et al. (1998) study, the aim is to use GIS to maximize benefits and identify potential greenway corridors for local conditions through a weighted assessment. GIS layers representing demand centers and attributes pertaining to connectivity were added to each data table, allowing for the identification of possible corridor alignments based on demand areas. The authors then used capability scores between 1 and 0. The sum of the products for all factors yielded a final suitability
3
score, to which a weighting system was applied to calculate overall suitability. The data layers chosen for their suitability analysis (among them schools, parks, floodplains, sewer lines, zoning, development, and future land use) are much like ours and the capability scores are much like our rankings. An overlay of final coverages for suitability, approachability, and proximity resulted in a raster layer for their corridor delineation. The authors suggest that the same methodology could be applied to other localized studies (Conine et al., 2003). In addition to looking to these suitability analyses, we looked to Gowanus-based suitability analyses, including Green Spaces, Blue Gowanus (2013), by students of Hunter College, and Reconsidering Gowanus (2010), by students of Baruch College.
low rents attracting artists and young professionals, development spreading along a transit corridor— is one familiar to all New Yorkers. But, the report cites the enormous obstacle of remediation as one that shapes development in Gowanus most significantly. All of the speculation that follows is of secondary consideration; the toxicity of the land transcends even big players in the real estate market. Research Gaps The research the Hunter students conducted is quite similar to our research, except that their study was completed before the community rezoning meetings began. The eventual zoning amendment for Gowanus will have an impact any work that is done. Acknowledging that the neighborhood is in flux is incredibly important. The authors also limited the extent of their study to the Gowanus Canal Watershed boundary. Consideration of lots beyond the watershed could provide new insight.
Completed in 2013, Green Spaces, Blue Gowanus seeks to understand opportunities for stormwater management on private property to improve the watershed and addresses the ways existing residents in Gowanus can be engaged to invest in such activity. The students found that not only was there considerable room for improvement such as bioswales and other green infrastructure, but the community members were interested in investing in these changes. This research served us well by demonstrating which reasonably small-scale green infrastructure strategies could be implemented in the area and citing the benefits and limitations of their implementation.
The Baruch College report was written before the construction—and approval—of the Whole Foods; this is quite impactful because although the opening of Whole Foods was not a singular landmark change for Gowanus, it is a huge asset to developers of new residential properties. The esplanade is also one of the only examples of a privately-owned public space, so it has to some extent set an example for others. If more ecologically productive privately-owned public spaces are to come about, there must be discussion on standards. Once again, the extent is limited; much of the analysis is based solely on Gowanus within the designated boundaries, not the watershed or surrounding neighborhoods.
Reconsidering Gowanus (2010), a study produced by Baruch College, details the potential for sustainable development along the canal in this time of dramatic change for both the canal and the surrounding area. It begins with a brief history of the neighborhood and outlines the ebbs and resurgences of industrial activity and population. The study outlines demographics of the neighborhood, noting a decline in residents since the 1970s. In addition to easy access to transit, rent is low compared to that of surrounding neighborhoods and most of the housing units are rented. The story told here—
4
METHODOLOGY
Who’s There: Analysis for the criteria measuring proximity to population of interest
At the start, we identified three overarching criteria to do with who’s there, what’s there, and how potential infrastructure will be used, specifically: Proximity to Population of Interest, Potential/Need for Green Infrastructure, and Public Accessibility. These criteria are qualitative in nature, but informed by variables that are represented quantitatively. The map for each criterion had extent in common (all data was clipped to the watershed boundary), and unique variables. For each variable, we reprojected data using the North American Albers Equal Area projection. Our study relies heavily on accuracy with respect to distance, which makes this coordinate system that preserves distance especially suitable.
For both Income and Population Density, datasets that were originally housed in the attribute table of our watershed layer, we had to convert from feature class to raster. We then followed the same steps as with the other raster datasets to complete analysis. First we used the Euclidean Distance tool (not specifying a distance) and then reclassified. Low income and high population density were assigned higher/more suitable rankings, with the interest of serving populations most in need of public space. What’s There: Analysis for the criteria measuring potential for green infrastructure Brownfields We identified contaminated sites, particularly brownfield sites, based on the Brownfield Opportunity Area (BOA) Study and data from the Toxic Release Inventory. Not only are these the sites of the most contamination in the watershed, these are also the sites that directly affect the canal’s water quality. There were sites in the BOA study around the canal identified as Manufactured Gas Plant (MGP), EPA “Dirty 7” Site, Asphalt Plant, Chemical/ Plastic Manufacturing, Coal Yard, Oil/Petroleum Business, Multiple Contamination Sources, Spill Sites, Other Manufacturing Use, and Superfund Area.
For those variables which required we examine distance, we used the Euclidean Distance tool. This tool allowed us to specify a distance (which we did in the case of Public Transit: 400 meters) or not (in the case of most others). Before using this tool, we set the watershed boundary as the mask, so that this boundary matched that of our analysis. Because the Euclidean Distance tool converts features to rasters, we only converted to raster those for which we did not use the tool (e.g. Income and Density). Each rasterized variable was reclassified using the Equal Interval classification to have a range of ranked values from 1 to 5 (5 being most suitable) to establish a standardized metric. Once all the variables were reclassified and symbology was altered, we created final suitability maps for the criteria. Since the three criteria do not apply to all of the land uses (Public, Private, and Privately Owned Public), we used the raster calculator to combine criteria maps according to different land uses. This allowed us to identify hotspots derived from the data in each variable for each criteria.
In addition, we overlayed Toxic Release Inventory data that had a generic listing of Brownfield sites in a shapefile format. We calculated lots based on the addresses listed in the attribute table of the shapefile, and merged the result with the BOA study map file. We used all of the ten categorized BOA sites, and ranked them all the same, with on-site being 5 and distance from site extending to one based on the Euclidean distance tool. Further analysis might indicate that implementation of development or green infrastructure may need to wait until post-dredge, depending on how toxic the site is.
5
How Will It Be Used: Analysis for the criteria measuring public accessibility
The translation of the BOA map to GIS took place on the clipped base layer from the MapPLUTO clipped map layer of lots. The MapPLUTO layer was clipped from a Brooklyn layer to the watershed boundary based on a blockgroup-level clip we made from the Proteus Gowanus watershed map.
Proximity to Subway Entrance or Bus Stop The variable pertaining to public transit refers to the distance (meters) from a site to a public transit hub. The data was retrieved from NYC OpenData and clipped to the watershed boundary. Preprocessing steps includes using the Euclidean Distance tool to convert to raster and setting a distance of 400 meters then reclassifying using the Reclassify tool. There were five rankings from 1 to 5, with 5 being the most suitable. The smaller the distance, the closer to 5.
Once this layer was clipped, we used the Clip tool to perform the new clippings from a copied layer that includes each lot, identified from the BOA study, for each Brownfield site type. Once identified, these sites were merged with the layer of TRI brownfield sites based on the addresses listed in the attribute table of the shapefile from TRI. We then created a single shapefile layer of the isolated brownfield lots in the watershed to perform the Euclidean distance tool on. Next, we used the Reclassify tool to reclassify into five categories using Equal Interval classification.
Proximity to School The variable pertaining to schools refers to the distance (meters) from a site to a school. The data will be retrieved from MapPLUTO and clipped to the watershed boundary. Preprocessing steps include: using the Euclidean Distance tool to convert to raster (without specifying a distance because we did not think it important to set a threshold) and reclassifying using the Reclassify tool. There were five rankings from 1 to 5, with 5 being the most suitable. Similar to the last variable, the smaller the distance, the closer the ranking to 5. The idea is that consideration of the locations of schools is critical; students should have access to green infrastructure as part of their education, whether formal or informal.
Proximity to Flood Zone The Flood Zone layer came from the FEMA floodplain map. The proximity to flood zones was created using the Euclidean Distance tool without specifying a distance, but using the watershed as a mask. We reclassified into 5 categories using Equal Interval classification, with 5 as the most suitable (closest to the flood zone). Proximity to CSO Outflow Point The CSO points were manually input to match Oasis. This layer, like Brownfields, has bearing on contamination. More often, CSO remediation projects might take the shape of on-water installations, but even though our focus was more on in-ground interventions, we included the CSO sites for an understanding of toxic release on the watershed as a whole. Future analysis could break the Potential/Need for Green Infrastructure down according to in- or near-water and in-ground. We used the Euclidean Distance tool on CSO points without specifying a distance, then reclassified. The closest sites to a CSO recieved a ranking of 5, for most suitable, and those furthest from a CSO a ranking of 1, for least suitable.
6
VARIABLES MEAN HOUSEHOLD INCOME BY CENSUS TRACT (2013) Source: U.S. Census
POPULATION DENSITY PER SQ. MILE BY BLOCK GROUP (2013) Source: U.S. Census
BROWNFIELDS Source: Toxic Release Inventory
PROXIMITY TO FLOOD ZONE Source: FEMA
PROXIMITY TO CSO OUTFLOW POINT
PROXIMITY TO SCHOOL Source: Map PLUTO
PROXIMITY TO SUBWAY ENTRANCE OR BUS STOP Source: NYC Open Data
Source: Oasis
CRITERIA
PROXIMITY TO POPULATION OF INTEREST
POTENTIAL/NEED FOR GREEN INFRASTRUCTURE
PUBLIC ACCESSIBILITY
PRIVATE LAND USE SUITABILITY
PRIVATELY-OWNED PUBLIC LAND USE SUITABILITY
PROXIMITY TO POPULATION OF INTEREST
PROXIMITY TO POPULATION OF INTEREST
PROXIMITY TO POPULATION OF INTEREST
POTENTIAL/NEED FOR GREEN INFRASTRUCTURE
POTENTIAL/NEED FOR GREEN INFRASTRUCTURE
POTENTIAL/NEED FOR GREEN INFRASTRUCTURE
PUBLIC ACCESSIBILITY
PUBLIC ACCESSIBILITY
PUBLIC ACCESSIBILITY
PUBLIC LAND USE SUITABILITY
PUBLIC LAND USE INTERVENTIONS
PRIVATE LAND USE INTERVENTIONS
PRIVATELY-OWNED PUBLIC LAND USE SUITABILITY
07
TRANSPORTATION/ UTILITY
05
COMMERCIAL
05
COMMERCIAL
08
PUBLIC FACILITIES & INSTITUTIONS
06
INDUSTRIAL/ MANUFACTURING
06
INDUSTRIAL/ MANUFACTURING
09
OPEN SPACE & RECREATION
11
VACANT
11
VACANT
11
VACANT
RESULTS
west. The highest values almost directly relate to distance from the canal. We chose the 530 President Street lot for our analysis, with proximity to the eastern branch of the canal where a school, several bus stations, and a subway entrance are located.
There were very clear areas in which the highest value of public land use suitability were located. Due to the fact that the public land use suitability map was comprised of all three of the criteria maps, its values were the highest, the darkest green having a value of 34. Upon examining the Proximity to Population of Interest map and the Public Land Use Suitability map side-byside, it is clear that the Proximity to Population of Interest map had a strong influence on the final Public Land Use Suitability map. This is because the high income/low density sites are focused primarily at the north end of the canal. The same high-suitability lots on the north end of the canal that were highlighted with high suitability in the Proximity to Population of Interest map were also the highest suitable lots in the land use map for public suitability. The 160 Wyckoff lot analysis came as a direct result of this overlap.
DISCUSSION Results Summary & Implications Of the lots with high suitability, we conducted research on three that stuck out to us upon performing a Google search. One of the three came from the Public Land Use Suitability map, one from the Private Land Use Suitability map, and one from the Privately-Owned Public Land Use map. Lots furthest from the canal were the lowest in suitability in the Public Land Use Suitability map, most likely as a result of the Potential/ Need for Green Infrastructure criteria, as its hottest spots were almost directly related to distance from the canal. Overlap between high-density and low income heightened the suitability value at 160 Wyckoff.
The Private Land Use Suitability map was no different from the green infrastructure criteria map, because the only criteria we thought applicable to the Private Land Use Suitability was Potential/Need for Green Infrastructure. This is the criteria map almost directly related to distance from the canal. The highest value is 15, meaning that because three variable maps with a suitability ranking value maximum of five went into this map, that the most suitable lots on this map are indeed highly suitable, as their variable lot value ranking was 5 in all three variable maps. The 268 Bergen street lot was the private lot chosen for analysis with the highest suitability, as it achieved a 5 in all three variable map suitabilities.
268 Bergen was identified because it is close enough to brownfield sites and to the CSO outflow points that green infrastructure would provide remediation. Additionly, its proximity to the canal put it within the flood zone, where it is prone to flooding and inundation. As stated in our client meeting with GCC, proximity to schools is a very important part of the accessibility criteria map, as it gives the potential for school programming around the site, with an opportunity for GCC’s role to foster a stronger connection between the nearby schools and the sites for potential green infrastructure. 530 President is a very large lot, and therefore might have higher green infrastructure efficacy potential, as its size would allow for more space for green infrastructure intervention.
Finally, the Privately-Owned Public Land Use Suitability map included the Potential for Green Infrastructure map and the Public Accessibility criteria map. On the POPS map, there is a school and several public transit hubs around the eastern branch of the canal and two sites on the north and south ends of the canal to the
9
Recommendations
appears unused. There is a community garden slightly east of the building; since there is already some greenery, and there is a bridge on Carroll across the canal, this might be a fruitful site for a green corridor.
The zoning district that the 160 Wyckoff building is within is often used as a ‘buffer zone’ between heavier industrial areas and residential areas. The designation allows for 1-2 story buildings used for light manufacturing. The building is owned by the Ulano Corporation, which manufactures screen printing products. It is difficult to gauge whether or not the roof of this building (which is quite large) might be suitable for a ground-cover green roof, for instance, because of the ventilation structures that cover much of the area. Green infrastructure on the sidewalk may also prove difficult because of loading and unloading on the block. Despite these challenges, it is a huge building, and perhaps with careful planning the building’s potential can be utilized. At 160 Wyckoff, there is a park in the middle of a set of housing projects. Public parks are not generally subject to zoning regulations. The space is bounded by a NYCHA development (Gowanus Houses) and has an assortment of ball courts and a playground. Much of the area is paved. Potential interventions include bioswales on the sidewalk in front of the park, and replacing hard surfaces within the park with absorbent, green alternatives. Perhaps by working with NYCHA and the residents of the Gowanus Houses, GCC could facilitate green roofs on the apartment buildings, further relieving storm water runoff. Also in a light manufacturing district, the 530 President building extends from President Street through to Carroll. The Textile Arts Center (TAC), a space that holds workshops and studio space for textile-based work and learning, uses the first floor of the Carroll street frontage. TAC has a program committed to growing plants for natural dyes and partnering with community gardens; perhaps there is room for a partnership with GCC using the skills of their program participants? It is not clear what the rest of the building is used for, but the roof is large and
10
Suitability
MEAN HOUSEHOLD INCOME BY CENSUS TRACT (2013) Source: U.S. Census
Suitability
POPULATION DENSITY PER SQ. MILE BY BLOCK GROUP (2013) Source: U.S. Census
PROXIMITY TO POPULATION OF INTEREST
Suitability
BROWNFIELDS Source: Toxic Release Inventory Suitability
PROXIMITY TO FLOOD ZONE Suitability
Source: FEMA
PROXIMITY TO CSO OUTFLOW POINT Suitability
Source: Oasis
POTENTIAL/NEED FOR GREEN INFRASTRUCTURE Suitability
PROXIMITY TO SCHOOL Suitability
Suitability
Source: Map PLUTO
PROXIMITY TO SUBWAY ENTRANCE OR BUS STOP Source: NYC Open Data
PUBLIC ACCESSIBILITY
Suitability
Suitability Gowanus Canal Privately Owned Lots
Suitability Gowanus Canal Privately Owned Lots
Suitability Gowanus Canal Privately Owned Lots
PUBLIC LAND USE SUITABILITY
PRIVATE LAND USE SUITABILITY
PRIVATELY OWNED PUBLIC LAND USE SUITABILITY
Address: 160 Wyckoff Street Owner: NYC DPR Zoning: Park
Address: 530 President Street Owner: PDS Carroll Street LL Zoning: M1-2
Address: 268 Bergen Street Owner: Ulano Corporation Zoning: M1-2