Results
Maps of Bird Collision Risk Criteria
For a detailed explanation of how these criteria were chosen, processed and visualized see my project methodology section
Multi-Criteria Evaluation (MCE) Results
Note that building shadows are not included in my MCE criteria due to uncertainty. Instead, I have overlaid shadows onto my MCE results just for visual purposes. From the MCE results, whole buildings were rated on a scale of low, medium, high, and highest priority for bird friendly retrofits as seen in the legend.
Regarding building facade specific retrofit recommendations: thick outlined facades denote where the building front would be, which I assumed would have higher glazing than the side or back. For this project, I also hypothesize that building facades shrouded by a shadow from another building will likely have less reflection of vegetation or open sky and potentially reduce bird collision risk.
All weighted
Sensitivity Analysis Map
All weighted + Building Shadows
Sensitivity Analysis Map + Building Shadows
The above sensitivity analysis map combines two weighing scenarios for bird collision risk criteria: all weighed and equal landscape weights. Buildings identified as high priority for just the equal landscape weight scheme are in yellow and buildings identified as high priority for just the all weighted scheme are in pink. Orange denotes buildings that, considering the different weight scenarios, would be high priority for retrofits.
Priority Buildings for Bird Friendly Retrofits
Weighted MCE - High Priority (Red buildings)
31 buildings in total
Allard Hall
AMS Nest
Beaty Biodiversity Centre
Beaty Biodiversity Centre Museum
Biomedical Research Centre
Brock Commons
Brock Hall
C. K. Choi Building for the Institute of Asian Research
Central
Centre for Interactive Research on Sustainability (CIRS)
Chemical & Biological Engineering Building
Continuing Studies Building
David Lam Management Research Centre
Earth Sciences Building
Engineering Student Centre
Gerald McGavin Building
Henry Angus Building
Hugh Dempster Pavilion
Indian Residential School History and Dialogue Centre (IRSC)
Irving K. Barber Learning Centre New
Liu Institute for Global Issues
Orchard Commons
Pharmaceutical Sciences Building
Ponderosa Commons East
Ponderosa Commons North
Ponderosa Commons West
Robert H. Lee Alumni Centre
The Brimacombe Building
UBC Bookstore
University Services Building
Walter C. Koerner Library
Sensitivity Analysis (Orange buildings)
32 buildings in total
Allard Hall
AMS Nest
Beaty Biodiversity Centre
Beaty Biodiversity Centre Museum
Biomedical Research Centre
Brock Commons
Brock Hall
C. K. Choi Building for the Institute of Asian Research
Central
Centre for Interactive Research on Sustainability (CIRS)
Chemical & Biological Engineering Building
Chemistry Building
Continuing Studies Building
David Lam Management Research Centre
Earth Sciences Building
Engineering Student Centre
Gerald McGavin Building
Henry Angus Building
Hugh Dempster Pavilion
Indian Residential School History and Dialogue Centre (IRSC)
Irving K. Barber Learning Centre New
Liu Institute for Global Issues
Orchard Commons
Pharmaceutical Sciences Building
Ponderosa Commons East
Ponderosa Commons North
Ponderosa Commons West
Robert H. Lee Alumni Centre
The Brimacombe Building
University Services Building
Walter C. Koerner Library
Wayne and William White Engineering Design Centre
Buildings in red are only identified as high priority for the all weighted MCE, buildings in orange are only identified as high priority using the combined all weighted and equal landscape sensitivity analysis.
Buildings in black text are identified as high priority by both maps.
Priority Buildings based on Financial Feasibility
While my model identifies up to 32 buildings as high priority for retrofits, it is relevant to consider the financial feasibility of undertaking such an enormous task.
I raised this point with my supervisor, Penny Martyn, Green Building Manager for Campus and Community Planning and she suggested that aside from deliberate, targeted retrofits to install bird friendly film, fritted, or acid etched patterns, aging windows (approx. 30 years old) that are in need of replacement to maintain building insulation can be used as opportunities for bird friendly design.
​
The following buildings were originally constructed in the 1980s and if not renovated yet, have windows that are about 30 years old and should be high priority for full window replacement bird friendly retrofits.​
-
Food Nutrition and Health Building (1982)
-
Douglas Kenny Building (1983) - Currently being retrofitted with bird friendly glazing
-
UBC Bookstore (1983)
-
Pulp and Paper Centre (1985)
-
Biomedical Research Centre (1987)
Uncertainties and Limitations
Limitations of GIS Representation
Building Dimensionality
Building facade glazing and surface area is in fact much more complex than what can be represented in 2D visualizations (ex. Pharmaceutical Sciences building). Additionally, my glazing model doesn’t consider the impact of glass roofs or horizontal glazing. The issue of building footprint print dimensionality matters to bird building collisions in particular due to configurations that can create “design traps” such as vegetated courtyards where windows face each other or glass hand railings.
​
Glazing Classification
As I could only sample and create classifications based on 9 buildings, I broadly categorized building glazing percentages of 31% and higher as “high glazing”, but I am unsure of what the true minimum and maximum glazing percentage for UBC buildings would be. There could also be outliers in glazing percentage which are unaccounted for in my classification as each building has its own aesthetic presentation outside of the general architectural style it falls under. In attempt to mitigate the effect of outlier buildings, I did look up images of some buildings to see if they look like they generally fit into my glazing classification schemes, but this was eye-balled and not accurately measured like my sampled buildings.
Vegetation Representation
While I chose to convert tree points to 5 meter diameter polygons, tree canopies can range up to 20 meters and more. Bird species interactions are also specific to vegetation types and this can possibly impact building collision risks.
Sun Shadow Analysis
I only used buildings shades to each other but shadows could also be produced by features not accounted for in my model.
​
High Glazing Facade Analysis
I identified high glazing facades through proximity to roads, however, it is possible that the impact of the glazed surface area could be minimized or cancelled out due to the magnitude and frequency of human activity as seen Kahle et al.'s (2016) study. Further research into the interactions of human activity to glazing surface area and vegetation would be necessary before assuming that building fronts have higher glazing and thus increases bird collision risk.
Data Gaps
Tree (2009), soft landscape (2015), and road (2016) shapefiles were not up to date and as construction is always ongoing on the UBC campus, it’s highly likely that many trees have been removed and planted.
Pharmaceutical Sciences building
Image source: HCMA Architects
Limitations of Project Scope
The north facade of the Museum of Anthropology.
Image source: UBC Undergraduate Programs and Admissions
Representing Glazing to Vegetation Interactions
Due to the coarse, building level approach of my project, I was unable to determine facade specific bird collision risks. The surface area of glazing itself is important to consider, however the biggest determinant for bird building collisions lies in how vegetation is reflected onto the glazed surfaces which can only be observed at a finer scale than my model (De Groot, 2018).
For example, I generalized glazing percentage to the whole building, and this could mask the possibility of one building side, such as the north face of the Museum of Anthropology, having a disproportionate amount of glass which is directly reflecting mature trees and a pond.
Temporal factors
Migration seasons, especially at UBC, are a critical aspect that my GIS model did not take into consideration. Changes in sunlight access and reflectivity, vegetation, and the weather could all impact bird collision risk as well.
Lack of Statistical Confirmation
This project was conducted over 5 weeks, and due to time and resource limitations, I did not use a statistically accurate random sample for building glazing percentage.
I also tried to test my model of glazing percentage to 2 sets of building collision points: the combined collision counts of a 2015 SEEDS study by Huang & Porter and 2017 SEEDS study by Gentile & Cheung which examine the same sets of buildings, and volunteer reported data from the FLAP Canada https://birdmapper.org/ site which were collected for about 9 buildings to perform a regression analysis. However, as the collision points are so limited for both sets, the results are inconclusive.
SEEDS 2015 and 2017 Bird Collision Points
FLAP Canada BirdMapper Bird Collision Points
Systematic Bias
As I was the sole researcher involved in data collection, data sampling, data transformation, methodology design and criteria choice and weighting in the MCE, my results were somewhat subjective and likely to be affected by systematic bias.
It is also important to consider my positionality as a staff member of UBC Campus and Community Planning. Over the course of working on this project, I had engaged in many discussions with coworkers of particular ‘problem buildings’ where bird collisions had been anecdotally mentioned and this may have swayed my imagination of the urgency and magnitude of bird collision risk for those certain buildings.
Recommendations
The golden crowned kinglet is one of the common victims of bird building collisions at UBC.
Image source: Wikimedia Commons
Directions for future studies
The mission to make UBC’s campus and landscapes more bird friendly is ongoing and never complete! The following are some ideas for future studies:
-
Include the broader landscape context with Pacific Spirit Park and the shoreline within my study. As these areas are more likely to attract birds and biodiversity in general, they can be used to understand the broader patterns of bird building collision risk.
-
Using updated or LiDAR tree and landscape data, map the hot spots of vegetation to inform new building design of the potential for bird building collisions. This encourages bird friendly design thinking early into the proposal stage if one intends to construct something over ecologically sensitive areas.
-
Compile building glazing through a representative sample of architectural styles.
-
As more bird building collision records are collected, test if my collision risk model is statistically significant.
-
Analyze bird building collision risk in residential neighbourhoods, especially the edge of Wesbrook neighbourhood which is surrounded by a forest.
Directions for UBC planning and operations units in using my tool:
-
Conduct building-specific audits. While my project provides a broad overview of potential buildings for bird friendly design retrofits, each facade and its surrounding context is extremely important to consider before installing bird friendly design retrofits. Here are some questions that only a building-specific assessment would be able to answer:
-
Does your building have any design traps?
-
What vegetation surrounds the building?
-
Can it be reflected in the windows?
-
What bird species rest and forage around your building? Are these species endangered or at risk?
-
What are the ecosystem interactions around your building?
-