Shadow analysis and
Sun shadow frequency analysis
In this exercise, I conducted a visibility analysis – specifically shadow analysis – in the urban locality of Umeå, Sweden, focusing on restaurants with outdoor seating. Outdoor seating constitutes an important feature for many restaurants during the short summer season. The provision of outdoor seating represents an opportunity to attract customers who want to enjoy the long and bright summer evenings in the Swedish North. However, due to the structure of the built environment in Umeå, outdoor seating locations with sufficient amount of sunlight are a scarce resource. Restaurants with a favorable “sun location” thus have a substantial competitive advantage during this important season for the industry.
The analysis in this exercise was carried out for central Umeå and was primarily based on a 3D model of the built environment provided by Umeå municipality. The exercise was comprised of two separate, but related, tasks. The first task was to conduct a shadow analysis in relation to all restaurants with outdoor seating facilities, while the second task was to assess and compare the effects on outdoor seating conditions of a hypothetically planned 35-meter-high office building development in Renmarkstorget, central Umeå.
↑Assigning shading status to each outdoor seating area.
↑17:00 part of the result from
the first shadow analysis.
First, I had to represent the physical layout of all outdoor seating areas in central Umeå by downloading a high-resolution RGB aerial photo of central Umeå. In addition, I needed to download a digital elevation model (DEM) that could make up the base heights for my ArcScene layers in ArcGIS Pro. Using the downloaded aerial photo of central Umeå, together with Google Street View, I could then draw the precise layout of all the outdoor seating areas.
Having done this, I could then calculate the number of sun hours every outdoor seating receives during a selected time by running the “Sun Shadow Volume” tool. Given the specific activity, eating out and consuming drinks, I needed to specify a time interval that was most relevant for this activity. I therefore chose the 21st of June. With the results from running the tool, I checked for intersections between shadow multipatch features and outdoor seating areas at 17:00 and then assigned a "shade status" (see upper-left screenshot). I then repeated this process for 18:00 and 19:00 to see how the shadows developed for each outdoor seating area throughout the early evening.
With that done, I summarized the findings by logging the number of sun hours in a table (see below) and therefore compiling a ranking of each outdoor seating in Umeå.
↑A snippet of the final ranking table
from the first shadow analysis.
For the second task of the shadow analysis, I used the same data as before to assess the effects of the hypothetically planned 35m-high office building development on Renmarkstorget in central Umeå. This time, however, I used the ‘Slice Multipatch’ and ‘Scale’ editing tools to raise the correct segment of the building complex up to the desired planned height of 35 meters (see screenshot below to the left). Moving on, I used the same layer for the outdoor seating areas, as the point was to see how this development affects the same outdoor seating areas using the same ranking scale.
After raising the planned office building development to a height of 35 meters, and then running the shadow analysis again, I could start ranking the shade status of the outdoor seating areas on Renmarkstorget and compare the results.
↑The final ranking table from the second analysis.
This ranking comparison shows, perhaps not surprisingly, that the outdoor seating areas of the restaurants on Renmarkstorget get a worse overall ranking after the development or remain unchanged, at least for the 17th hour. All in all, most outdoor seating areas on the square would generally suffer from the planned development.
↑Planned 35m-high office building
development in Renmarkstorget, Umeå.
↑17:00 part of the result of the second shadow analysis.
Sun shadow frequency analysis
Shadow analysis, like the one covered above, is a form of visibility analysis where the aim is to answer the question of where a shadow is and to calculate the amount of shadow cast by a given object at different times of the day. Both the inputs and outputs of the 'Sun Shadow Volume' tool in ArcGIS Pro are multipatch features, and this has the benefit of storing data such as the name of the feature class casting the shadow volume, a unique ID of the feature casting the shadow volume, local date and time used to calculate sun position, the horizontal azimuth angle and vertical angle of the shadow volumes.
However, there is a different, but more rudimentary approach to sun shadow analysis, and that is sun shadow frequency. Sun shadow frequency analysis is a semi-3D technique used for analyzing the frequency and duration of shadows created by objects on the Earth’s surface. The tool calculates the number of times a fixed position on a surface has its direct sight line to the sun obstructed by multipatch features. In other words, this simply means “how often is a given area shaded within a chosen time frame” which doesn't sound too different from what sun shadow volume is all about (the previous method), except the main difference is the analysis of volume amount vs. the frequency of shadow.
This method works by taking 3D multipatch objects as an input along with a digital elevation model (DEM) or digital surface model (DSM) and spits out a 2D raster that represents the number of time intervals each pixel experiences during the chosen time frame.
By analyzing the frequency of shadows over a given time period, we can determine how often and for how long certain areas are affected by shadows. Sun shadow frequency can provide valuable insights for many fields and a wide range of applications, such as in architecture, urban planning, solar energy potential, and more. For example, in urban planning, knowing the frequency of shadows can help determine which areas of a city receive the most sunlight throughout the day and which areas are more likely to be shaded. This information can be used to optimize the placement and orientation of buildings and public spaces, to ensure that they receive sufficient sunlight throughout the day.
To try this out, I chose to look at the frequency of shadows on two town squares in Visby, Gotland in Sweden during the the three summer months of June, July, and August.
For the analysis, I chose a time frame starting solar noon until sunset on the 21st of June of 2023, or summer solstice. Then again at solar noon until sunset on the 21 of July, and again the same thing for August.
With that, I had two research questions for this project:
Summarizing in a table how frequently any given pixel is shaded enabled me to compile the results in a chart.
↑For Stora Torget, we can see that ‘Shaded in the last 2 hours’ and ‘Shaded in the last hour’ account for the majority, or 56.9%, which is not bad, but only 2.7% is ‘Never shaded during time frame'.
↑Conversely, for Södertorg, we can see that the majority enjoys no direct shadow during the entire time frame, or 55.3%, as well as a total of 28% being only shaded in the last hour.
When comparing the frequency of shadows between months, we can see that Södertorg is less shaded during the entire time frame compared to Stora Torget. However, likely explained by the downward slope of its terrain, as well as the lower building density in its vicinity. There is minimal change in shade frequency between June and July, but something seems to change in August and become less frequently shaded. This seemed to contradict the transition we see between June and July. Perhaps this may be because in August, the days are beginning to get shorter, and the sun sets more quickly, resulting in shorter twilight and shorter shadow durations.