Above-ground Carbon Storage of Urban Trees on UBC Campus

Introduction

Climate change has become a critical issue worldwide, particularly in urban areas. The UN predicts that by 2050, 68% of the world's population will reside in urban regions. However, urbanization also leads to an increase in greenhouse gas emissions, with 70% of carbon emissions originating from urban regions. In Canada, despite urban areas constituting only around 0.25 percent of the country's landmass, more than 80% of the total population lives in cities. Urban trees, particularly in the form of an urban forest, can store substantial amounts of carbon dioxide and mitigate air pollution by absorbing carbon dioxide from the atmosphere through photosynthesis.

Assessing urban tree carbon stocks at the single-tree level is essential because of the diverse forms and configurations of trees found in urban areas. Various approaches, such as the canopy height model obtained from light detection and ranging, have been used to identify individual trees. The University of British Columbia (UBC) Vancouver campus, with over 50,000 students, faculty, and residents, represents a sizable urban area within Vancouver. The main objectives of this project are:

1. Estimating the above-ground carbon storage of UBC campus trees.
2. Mapping the tree carbon storage over the UBC campus.
3. Providing recommendations for policy in alignment with the UBC Climate Action Plan.

By analyzing the above-ground carbon storage in the campus ecosystem, this study will enhance the understanding of the ecosystem services provided by UBC campus natural assets.

Study Area

The University of British Columbia (UBC) Vancouver campus is situated in Vancouver, British Columbia, Canada, at the western end of the Point Grey Peninsula, covering an area of 4.02 km2. The campus is divided by the Pacific Spirit Regional Park and is directly east of the University Endowment Lands, which are governed by the province. The area was originally a coniferous woodland clearing, but since 1925, UBC has planted over 8,000 trees of various species, with more than 10,000 native trees in their natural habitats. The dominant species on campus include western redcedar, pin oak, and red maple. The campus is located in the Coastal Douglas-Fir biogeoclimatic zone of the Moist Maritime Subzone, ideal for temperate mixed woods, with an average elevation of 87 meters above sea level. The climate in this area is moderate, with dry and warm summers and rainy winters, with an average annual temperature of 11°C and 146cm of precipitation.

Data Summary

The Lidar point cloud data used in this study was obtained from the Open Data Portal of the City of Vancouver ( opendata.Vancouver.ca ), which was acquired on August 27th and 28th, 2018, with an average density of 30 points/m2. The data consisted of 11 tiles with a horizontal accuracy of 0.36 m and a vertical accuracy of 0.18 m under a 95% confidence interval. The map projection used was NAD1983 UTM 10N. The data was classified by the City of Vancouver into different categories such as unclassified, bare-earth and low grass, low vegetation (height 2m), water, buildings, other, and noise.

The legal boundary and land-use planning map of the UBC Vancouver Campus were obtained from UBC open geospatial data on GitHub ( github.com/UBCGeodata ), which were projected in the WGS84. The legal boundary of the UBC Vancouver Campus was used to obtain the polygon of the study area, while the land-use map was used to calculate the tree carbon storage for each land-use type.

Methods

The analysis involved five stages, which included data pre-processing, generation of digital terrain model (DTM) and height normalization, generation and smoothing of a canopy height model (CHM), individual tree detection (ITD) and individual tree segmentation (ITS), and estimation of diameter at breast height (DBH) and above-ground carbon storage. The “lidR” package in the R programming language was used to perform the ITD and ITS algorithms. The overall workflow is presented in Figure 1.

During the data pre-processing stage, the LiDAR data were filtered to exclude non-forested areas and buildings. A noise filter function was developed to identify and remove high-level outliers or noise caused by flying objects such as birds or aircraft. The Inverse Distance Weighting (IDW) method was used to create the DTM, which facilitated terrain normalization to eliminate the effects of terrain on above-ground measurements. A hybrid method of DTM normalization and point cloud normalization was applied, and a 1 m spatial resolution was chosen for the CHM generation. A 3 x 3 median filter was applied to the CHM to remove nearby local maxima caused by tree branches and other non-canopy features. The resulting CHM was used for individual tree detection and biomass estimation.

The Local Maximum Filter (LMF) was used to detect individual tree tops from the LiDAR point cloud data. Alternatively, the Canopy Height Model (CHM) was used to speed up the process, but the output was more complex. A 5x5 window size was applied to the smoothed CHM with a resolution of 1 m, and the Dalponte algorithm was used for tree segmentation due to its highest accuracy. The algorithm identifies individual tree tops and crowns from point cloud data using the growing region technique. After segmentation, the “crown_metrics” function was applied to the LiDAR data to delineate the crown shapes and calculate the area of the concave hulls.

DBH was estimated using the crown area (CA) and maximum tree height (TH) obtained from crown delineation, and the DBH estimation model presented in Equation (1) was used.

DBH = b1 × ((TH-1.3) ^b2 ) × (CA ^b3 ) (1)

To improve accuracy, trees with a height of less than 4 m and a crown area less than 12 m ²  were filtered out. Using the estimated DBH, the above-ground carbon (Cag) was estimated using an existing biomass model (Equation 2) with the assumption that carbon is 50% of biomass.

Cag = (exp (a + b × ln⁡DBH)) × 0.50 (2)

The total carbon storage for the study area was calculated by summing the estimated above-ground carbon of all individual trees.

Results

Above-ground Carbon Storage of Campus Trees

Using the TH and CA for each tree calculated using “lidR” package, the DBH values were estimated using the existing model. For each tree the DBH was then used to estimate above-ground carbon for each tree using Equation (2). Figure 5 shows the estimated above-ground carbon of each tree on the UBC Vancouver Campus. According to our results, the total urban forest carbon storage of UBC Vancouver Campus is estimated at 24.63 Gg. It is worth noting that there is a higher density of trees with high carbon storage on the southern part of the campus, where the UBC Botanical Garden and UBC Farm are located. Additionally, we observed relatively dense carbon storage at the northwestern corner of the campus, where the Nitobe Memorial Garden is located, as well as smaller areas at the middle part of the campus where the Rhododendron Wood and Totem Park Residence are situated.

Carbon Density Map

With the Cag of individual trees, we generated a kernel density map in ArcGIS, which presents above-ground carbon storage of urban forests on campus, the units is kg/m2 (Figure 6). Similar to the carbon storage map, areas with the highest carbon storage density are strongly agree with areas with dense high carbon storage trees. Based on our results, the average carbon storage of UBC Vancouver Campus is 6.13 kg/m2. In addition, we combined our carbon storage results with the land use map of the campus. Our results indicated that academic land use had the highest amount of carbon storage of 18.57 Gg (72.4%), the neighborhood (residential) land use had 6.12 Gg of carbon storage (23.9%), and the future planned neighborhood areas had the least amount of carbon storage of 0.95 Gg (3.7%). The amounts of carbon in each land use type are in consistent with the area of land use types.

Discussion

Our study aimed to estimate the above-ground carbon storage of urban forests on the UBC Vancouver Campus using LiDAR point cloud data. We found that the total above-ground carbon storage for urban trees on campus is estimated at 24.63 Gg, with an average carbon storage of 6.13 kg/m2. Our results are consistent with other cities in North America, Europe, and Asia, but at the higher end of the range.

We generated a spatially explicit carbon storage map for the campus, which can be used to inform future campus planning and assess above-ground tree carbon storage at various scales, from campus-wide to city-wide. We recommend incorporating tree species data in carbon storage assessments to more accurately quantify the carbon storage capacity of urban forests.

Urban forests offer a range of socio-economic and ecological benefits, including carbon storage and sequestration. They play a crucial role in mitigating carbon emissions in urban areas by functioning as carbon sinks. Our study highlights the importance of considering carbon storage in urban forests when making land use decisions and the need for continued research to better understand the carbon storage potential of different tree species in urban settings.

In conclusion, our study demonstrates the utility of LiDAR data for quantifying urban forest carbon storage, and our findings can contribute to broader ecological assessments and inform future campus planning.