Mapping the shade provision of street greenery using Google Street View
Thermal comfort in cities has become more and more important, especially in the context of global warming and rapid urbanization. Urban thermal comfort would influence human outdoor activities and the utilization of urban space. Severe heat has negative impacts on human well-being and efficiency. In the United States, extreme heat events are responsible for about one-fifth of natural hazard deaths. With the rise in global warming, many cities are expected to experience more severe heat waves (IPCC, 2007). The urban heat island effect would cause additional warming to the urban areas. This additional warming caused by urban heat island effect could bring 2.6 times more economic costs of climate change to cities.
As an integral part of urban ecosystems, street trees help to regulate urban microclimates and mitigate urban heat islands. In hot summers, street tree canopies can block the sunlight from radiating to the ground directly and provide shade for pedestrians. Using urban greening projects is an effective method to increase human thermal comfort in cities without changing the existing built environment.
It is therefore important to assess the shading service provided by street trees. Traditionally, remote sensing data have been widely used to evaluate the amount of urban greenery. However, the two-dimensional cover information cannot fully reflect the shading service of street trees, because the top-down view satellite imagery cannot reflect how the solar radiation reaching the street canyons. Recently, I have developed a method based on Google Street View panorama and building height model to quantify the exact shade provision of street trees in Boston.
Let first take a look at the study area and the datasets I used (Fig. 1). The first step is to prepare the spatial dataset in Boston. Fortunately, Boston has most of the datasets I need available. I get the LiDAR data and building height map from MassGIS (https://www.mass.gov/service-details/massgis-data-layers), the vegetation canopy cover map (Fig. 1 (b)) from Prof. Hutyra at Boston University (http://sites.bu.edu/hutyra/).
The second step is to collect Google Street View (GSV) panorama images. However, considering the fact that some GSV images were captured in different seasons(Fig. 2a). In order to accurately estimate the shade provision of street trees in summer, here, I only used those images captured in leaf-on seasons. Fig. 2b shows the spatial distribution of the green sites and non-green sites. In Boston, most sites have GSV images captured in green seasons. Therefore, we can use the GSV based method to estimate the city scale shade provision.
I used hemispherical images to analyze the shade along streets. The hemispherical images can be generated from GSV panoramas using geometrical transform and building height modeling using viewshed analysis. Fig. 3 shows the different hemispherical images generated from GSV and building height model, respectively. For those places with no street trees, these two methods can get a similar result (bottom two images). However, for those sites with tree canopies, the shade provision of street trees and building blocks may overlap (top two images). Therefore, the difference between these two methods should be considered as the actual shade provision of street trees.
I further used the sky view factor (SVF) to quantify the amount of shade in street canyons. As a dimensionless parameter of urban geometry, the sky view factor (SVF) indicates how much sky is obstructed by buildings and tree canopies. The SVF also represents the ratio between radiation received by a planar ground and that from the entire hemisphere’s input radiation. When the sky is totally obstructed, the SVF is zero while the SVF is one when there is no obstruction. Fig. 4 shows the SVF map generated by building height model. The downtown area has much lower SVF values than periphery parts, because of high-rise buildings there. Fig. 5. shows the SVF map generated based on GSV. The downtown area also has low SVF values, but the southwestern area also has low SVF values because of the abundant vegetation (Fig. 1b).
By comparing the Fig. 4 and Fig. 5, we can easily find that using the building height model method (Fig. 5) usually get larger SVF values than GSV based method (Fig. 5). This is not difficult to understand, because the building height model only considers the shading effect of buildings, and tree canopies are not considered.
Therefore, the difference of these two SVF maps can be considered as the shade provision of the street greenery in Boston. Based on my estimation, the existence of the street trees helps to decrease the SVF by 25%. That means the street trees in Boston help to provide 25% shade in hot summer. This is cool. Google Street View is publicly accessible (hope so after Google changed their data policy recently) and Globally available, the workflow can be used for many other cities to understand the value of street trees.
Reference
Xiaojiang Li and Carlo Ratti. (2018), Mapping the spatial distribution of shade provision of street trees in Boston using Google Street View panoramas, Urban Forestry and Urban Greening, 31, 109–119.
Xiaojiang Li, Carlo Ratti, Ian Seiferling. (2018). Quantifying the shade provision of street trees in urban landscape: A case study in Boston, USA, using Google Street View, Landscape and Urban Planning, 169, 81–91.
