Sustainable Management of the Urban Greenspace

Applying GIS technologies to urban ecological management at the Rose Kennedy Greenway in Boston, MA


The Urban Ecosystem

Parks confer a myriad of well documented benefits to cities, their local environments, and their residents. Beyond just their natural beauty, urban green spaces quantifiably better life in the city by decreasing air pollution, improving mental health, increasing biodiversity, and cooling space to name just a few (Annerstedt et al. 2012). As our global population rapidly urbanizes, effective maintenance of the urban ecosystem is vital in ensuring the continued health of humans and environment a like (Moore et al. 2003).

Challenges of the Urban Environment

While parks represent an invaluable resource, they face unique and considerable challenges growing in urban centers. These include intensified heat, cold, and wind, run off from pollutants, acidification from carbon combustion, and overall heavy use. As such, these spaces require regular, thoughtful, and ecologically mindful maintenance in order to create a thriving environment (Grimm et al., 2008).

Working with Ecology...

Conventional urban park management primarily focuses on an input biased approach; The nutrients needed by the park are brought in to the park ecosystem from an outside source. However, this standard of care is inherently unsustainable and irreverent of a parks complex and unique ecology. Alternatively, the urban ecological model seats sustainability at the center of it's mission (Breuste et al. 2011). Although requiring a more careful monitoring, this approach encourages greater, long-term benefits and improved park resilience.

The Rose Kennedy Greenway

The Rose Kennedy Greenway is one of only a handful of parks pioneering ecological horticulture in the urban setting. Since its founding in 2008, it has employed this approach and stands as model proving the robust success of sustainable practices. As the park matures, it's ecology grows more complex and encourages that maintenance evolve to confidently address new and future challenges. GIS technologies provide promising capabilities that can help horticulturalists identify patterns in the health of the park and optimize the environment, allowing Bostonians to enjoy the Greenway for years to come.

History of the Greenway

In the 1950's, Boston underwent a devastating urban transformation with the construction of the "Central Artery" - a massive multilane high way cutting through the historic downtown. By the 1990's planners had realized this mistake and proposed that these car lanes be moved underground in a construction project that Boston would soon learn to call the "Big Dig." This became the most expensive and complex infrastructure project of any US city. A replacement for the "Central Artery," the Greenway was proposed as an innovative and ambitious public park - a major improvement from the noisy, polluting, and neighborhood destroying highway that came before. It was opened to the public in 2008 and public support of the park has only grown as the park matures.

The Greenway looking toward Faneuil Hall, both before destruction of the "Central Artery" in 2004 and as enjoyed today

Full map of the Greenway

Native Pollinator Meadow, a recent addition to the Greenway

The Greenway Today

Today the park is maintained by the Rose Kennedy Conservancy and employs forward thinking practices in horticulture to achieve its goals of beautifying Boston. "[Their] approach reflects the latest science about sustainable and healthy ecosystems. It means no synthetic fertilizers, chemicals and pesticides are used to keep our park looking beautiful, no pollutants run-off into Boston Harbor to harm marine life, and children and pets can play safely on our lawns." (Horticulture Department, RKG)

As the park establishes itself into a more functioning ecosystem, the ecology and symbiotic, interspecies interactions will grow more complex. Ecological approaches in sustainable management are inherently knowledge biased and require deep consideration of place, biology, and historical trends. Paired with the master knowledge of the horticultural department at the Greenway, advances in Geographic Information Systems (GIS) and related technologies can help the park moving forward to identify complex spatiotemporal patterns and react accordingly.


GIS at the Greenway


This project is an initial summery of ongoing work to implement GIS at the Greenway. The preliminary work is separated into three parts focused on (a) the establishment and continued updating of the geodatabase, (b) using developing technologies in remote sensing to monitor the growth of the trees in the park, and (c) GIS to elucidate temporal and spatial trends at the Greenway. Finally, future potential projects are explored and summarized.

Establishing a Geodatabase

Before more involved geo-analysis can occur, a well organized database has to be create. The horticulture department has a wealth of data that they have recorded over their 10+ years of stewardship, and, while it is indexed by locations in the park, it was not formally georeferenced. The first challenge in establishing the geodatabase was conceptualizing how the current systems of tree or bed maintenance and identification could be updated and then uploaded to reflect best practices in unified modeling languages (ULM). Cleaning this data is a continuing process and there remains a trove of data to be explored once joined to geo-referenced objects.

Satellite orthoimagery was used in conjunction with older reference maps and the expertise of long time employees to identify and give shape to the existing beds, lawns, and geography at the park.

The existing naming convention for these features functioned well as they were and did not need to be modified. Additional attributes clarifying bed type, soil, and plant species can continue to be added.

Soil data from the last 10+ years was then joined to these references, providing a base map around which future analysis can be run.

Trees are an important asset of the park and monitoring their health, attrition, and growth is crucial. A separate layer for the parks trees were created using similar techniques as the beds and lawns.

The existing tree inventory at the park is robust, but the method of identifying trees is more complex than the relatively immutable landscape. Trees need to be replaced often, removed, added, switched out to reflect more resilient species, and change in size and shape. As such, the IDs for this layer are more involved and must reflect these realities.

Trees were then joined to existing identifying information such as past heights, genus, status, and health condition.


Monitoring Growth at the Greenway

Once establishing a geodatabase, the capabilities of geographic information systems can greatly aid horticulturalists in the maintenance of the park. While the horticultural department at the Greenway has diligently kept record on many aspects of the park from soil health, to water use, to tree growth, this data collection can be difficult and time consuming. GIS can simplify some of these processes.

One particularly relevant and rapidly expanding application of GIS is in forestry due in larger part to improvements in remote sensing technologies. Remote sensing - such as LiDAR - utilizes laser and light sensors mounted to low flying drones in order to produce a high quality 3D model of the environment. While commonly used to create detailed elevation maps, as the processing of this data improves, it additionally stands to revolutionize how we monitor the health of forest ecosystems both in wild and urban settings.

While the computational techniques needed to consistently identify singles trees from remote sensing are complex and still developing their accuracy and broader scale application, the future of tree monitoring lies in this technology (Eysn et al. 2015; Matasci et al. 2018; Wang et al. 2018; Liu et al. 2021). Below, we explore the current capabilities for individual tree height identification from current, leaf-off, quality level 1 LiDAR data -generously shared with us by the City of Boston - for application at the Greenway.

Monitoring Tree Growth

This satellite photograph conveys a wealth of information, however it is limited in its capabilities as it is bounded in a 2D image.

LiDAR

LiDAR data overcome this two dimensional challenge and details very fine point elevation.

Processing LiDAR

Individual trees can be pieced out of this image, giving researchers the ability to answer complex questions about the built environment and model our 3D world.

Calculating a Canopy Height Model (CHM)

The elevation from LiDAR can be converted into a a more workable form that interpolates the heights of individual trees.

Automating Trees Identification

The CHM was run through a multidimensional clustering-based algorithm using the R package "LiDAR Data Processing and Visualization" to derive tree points.

Improving Accuracy

A density-based clustering was run on the raw extracted tree points as a way of improving upon the automated algorithm.

Individual Tree Points

By deriving the central feature of these points, a more accurate tree map was created. The maximum value from the original clusters was used as the hieght of these central feature points.

Joining Heights to the Geodatabase

These derived heights can then be spatially joined to the existing tree inventory. These heights were joined in multiple, calculated stages in order to improve accuracy of the joins.

. . . Future Analysis

From these tree points, additional characteristics can be extracted. Using a reverse watershed method (Eysn, 2015), we hope to calculate an estimate of tree crowns and canopy within the park.


Ecology in Action

More than just monitoring growth and change at the Greenway, this technology helps park management to understand complex geographic relationships and optimize planning for the future of the park. We can answer questions such as: do areas of the park show higher rates of tree attrition and if so, what spatial elements may be contributing to that loss? Does the parks soil vary uniformly in water retention, or do some areas show perpetual drought or saturated conditions? Given current trends, what is the most optimal location for future plantings?

While our geodatabase remains in its infancy, we can begin to ask some of these questions. In the future, a more robust picture of soil health at the Greenway will be mapped, but to start, we can begin by exploring a particularly important quality of soil relevant for plant health: acidity. Healthy soil pH typically falls within the range of 6.0 to 7.0, and is regularly monitored to ensure the optimal growing conditions. Below, we look at just a few ways of visualizing change in soil pH at the Greenway since 2010.

Temporal Trends in Soil pH

The horticultural department tracks soil metrics annually and tweak their care of the park accordingly. An important question to ask is: how, and in what significant ways has this soil changed over time? The map to the left shows how the soil pH has changed in the park since 2010 as well as how significant these changes are. As a whole, the park has only seen significant decrease in pH (i.e. the soil has become more acidic) as time has moved forward. The Warf District particularly appears to have grown more acidic.

This data however, does not give a full picture of the soil pH change in the park. At the global scale we have seen a decrease, but at a more local scale, distinct groupings of park beds exist that further complicate the picture.

Temporal Clustering of Soil by Historic Trends in pH

Soil at the park has not change uniformly over time. Some areas of the park show more similarity in the change of their pH than others. However, using only the human eye, it can be difficult to identify these temporal similarities. A time series clustering can aid us in identifying similar location and group them accordingly. For the map to the left, a K-mean temporal clustering was used to achieve these groups.

These complex groupings can help horticulturalists identify larger trends in the health of the park or verify existing assumptions. This can be applied, for example, to create new and individualized care plans or influence future plantings.

From this map specifically, the temporal clustering seems to fall along different bed types (i.e. lawns vs. planting beds) and may have some type of underlying spatial relationship; the Warf District shows high incidence of the first and second cluster group, while more south it leans more to the third and fourth groupings.

A Deeper look at Soil pH at the Greenway

When we look at this graph showing the average pH by cluster, we gain even deeper view into the soil care at the Greenway over time.

A key feature distinguishing these clusters is their initial pH values. We begin with a large range of pH that approach a more stable and unified value as time goes on. This stabilization of the soil pH may reflect the diligent care on the part of the horticultural staff as they have cultivated a healthy, living soil that is more resilient to annual variation and can hold nutrients more securely.

As we continue to build the parks geodatabase and run these types of analysis, we gain a deeper understanding of how to cultivate a thriving ecology. Like a doctor assessing a patient, these spatiotemporal tools can aid in the prognoses of health challenges and identify interventions that are working.


Just the Beginning . . .

This represents the just earliest of analysis's that can be carried out using GIS. As the geodatabase grows, several exciting potential projects become possible. Three such projects are briefly outlined below, exploring pollinator plant integrations, conservation of water resources, and how to prepare the park for the threat of climate change.

Tracking Pollinators and Tailoring the Environment

The Greenway has worked hard to build an ecological refuge supportive of native pollinators. In return, these pollinators contribute to the diverse and robust ecosystem of the park and health of the plants.

By tracking these species, the Greenway can monitor its role in supporting native biodiversity. A citizen science approach can help to engage the community with activity at the Greenway while simultaneously building up this data base. Eventually, this can be used to identify areas of increased pollinator presence. In conjunction with this real time record of pollinator activity, a suitability analysis can be formulated that could optimize planting and guide establishment of pollinator friendly infrastructure such as Mason Bee houses.

Efficient Management of Water Resources

As the Greenway looks to mitigate the impacts of its management in the context of the local and global environment, monitoring of its water use will be important to ensuring sustainability. In addition to the natural features of the park, inputting the below- and above-ground irrigation into the geodatabase can help allocate watering and ensure only the necessary resources are being used.

To do this, a vulnerability score of the park can be constructed from existing humidity data, direction of run-off flow, shade cover, plant cover, and weather patterns, to provide a map to minimize water waste and provide moister to areas that need it the most.

Preparing for a Changing Climate

Our changing climate necessitates innovative and responsive monitoring of our natural resources. The New England in particular has experienced major climatic changes that increase stress on plant life. (State of the Plants, 2015). Many of our native plants are feeling this increased inhospitality which is strikingly exemplified by the decline of the native sugar maple in our region - an unofficial emblem of New England forest. As the climate changes, horticulturalist must not only identify which plants will be most suited to these new condition, but also question how their planting can be used to fight against an ecological decline or other externalities such as rising tides. Temperature and climatic data mapped using this software can help steer us in the right direction to combat and react to climate change.


Thank you to the staff in the Horticulture Department for providing the data and ecological expertise needed to making this project possible, the City of Boston for sharing LiDAR of the park, and Dr. Sumeeta Srinivasan for guidance in the GIS techniques.

Literature and Works Cited

Annerstedt, M., Nielsen, B., Maaruthaveeran, S., Konijnendijk, C. (2012) Benefits of urban parks: a systematic review. IFPRA World. Volume: 2012, Number: 6, pp 10-12.

Moore M, Gould P, Keary BS. Global urbanization and impact on health. Int J Hyg Environ Health. 2003 Aug;206(4-5):269-78. doi: 10.1078/1438-4639-00223. PMID: 12971682.

Grimm, Nancy & Foster, David & Groffman, Peter & Grove, Morgan & Hopkinson, Chas & Nadelhoffer, Knute & Pataki, Diane & Peters, Debra. (2008). The Changing Landscape: Ecosystem Responses to Urbanization and Pollution Across Climatic and Societal Gradients. Frontiers in Ecology and the Environment. 6. 10.1890/070147. 

Breuste, J., Qureshi, S. Urban sustainability, urban ecology and the Society for Urban Ecology (SURE). Urban Ecosyst 14, 313 (2011). https://doi.org/10.1007/s11252-011-0186-3

Matasci, G., Coops, N.C., Williams, D.A.R. et al. Mapping tree canopies in urban environments using airborne laser scanning (ALS): a Vancouver case study. For. Ecosyst. 5, 31 (2018). https://doi.org/10.1186/s40663-018-0146-y

Eysn, L.; Hollaus, M.; Lindberg, E.; Berger, F.; Monnet, J.-M.; Dalponte, M.; Kobal, M.; Pellegrini, M.; Lingua, E.; Mongus, D.; Pfeifer, N. A Benchmark of Lidar-Based Single Tree Detection Methods Using Heterogeneous Forest Data from the Alpine Space. Forests 20156, 1721-1747. https://doi.org/10.3390/f6051721

Liu, Haijian & Dong, Pinliang & Wu, Changshan & Wang, Pin & Fang, Meihong. (2021). Individual tree identification using a new cluster-based approach with discrete-return airborne LiDAR data. Remote Sensing of Environment. 258. 112382. 10.1016/j.rse.2021.112382. 

Wang, Zhihui & Wu, Jie & Wang, Yuanjian & Kong, Xiangbing & Bao, Hongzhe & Ni, Yongxin & Ma, Li & Jin, Chengran. (2018). CROWN-LEVEL TREE SPECIES CLASSIFICATION USING INTEGRATED AIRBORNE HYPERSPECTRAL AND LIDAR REMOTE SENSING DATA. Archives of Photogrammetry, Cartography and Remote Sensing. XLII-3. 2629-2634. 10.5194/isprs-archives-XLII-3-2629-2018. 

New England Wild Flower Society. 2015. State of the Plants: Challenges and Opportunities for Conserving New England’s Native Flora. Framingham, MA.

The Greenway looking toward Faneuil Hall, both before destruction of the "Central Artery" in 2004 and as enjoyed today

Full map of the Greenway

Native Pollinator Meadow, a recent addition to the Greenway