This project maps the soundscape of Olympiapark in Munich, providing valuable information
for sound-sensitive individuals to help them choose where to go. As urban environments grow
increasingly noisy, finding quiet spaces—even in parks—has become more challenging. This map
aims to simplify that search by highlighting areas suitable for contemplation and relaxation
while offering insight into the distribution of sound quality throughout Olympiapark.
Soundscapes of Olympiapark
“This project maps the soundscape of Olympiapark in Munich, helping soundsensitive individuals find quiet spaces for relaxation and contemplation while providing insight into the park's sound distribution.”
What it is about
How we built it
Our first step was to select the locations where we would record audio and measure noise
levels. We ended up with 58 locations, which we tried to place evenly throughout the park. We
paid special attention to locations where points of interest are located, such as the observation
deck, bridges, and stadium.
After we recorded about 10 audios, we felt it was important to categorize them into three groups: ‘pleasant’, ‘unpleasant’ and ‘mixed’. This classification is based on research about what sounds are “harmful” (cars, construction works, etc.), irritating the psyche, and what sounds are favorable (birds singing, the sound of water, etc.). There are quite a few points that had both beneficial and unhelpful sounds, so we categorized them as ‘mixed’.
We recorded all the data about sounds in Notion application, but soon we faced impossibility to get a connection to such a database. It was decided to move everything to Google Drive. Attempts were made to display our data as a web map. Points were displayed according to their properties: quality, geolocation, dB level. At the same stage, a gradient scale was created to show what color corresponds to what level and quality of the recorded sound. We were not able to find free ways to display the data correctly and beautifully, so we decided to switch to ArcGis Dashboards.
We processed the data in Java Script to automatically categorize the points into classes, and then loaded the resulting layers into ArcGis Dashboards. We ended up with two layers: the points with their properties and the interpolated grid. To the resulting map we added a table with information about each point. This included the ID of the point, the date the point was recorded, the average noise level, the sound quality and what kind of sounds are heard. Also, when you click on a point, you can get a link to the audio to see exactly what that location sounds like. In addition, a map legend, decibel frequency, perception of sound and general information about map have been added.
After we recorded about 10 audios, we felt it was important to categorize them into three groups: ‘pleasant’, ‘unpleasant’ and ‘mixed’. This classification is based on research about what sounds are “harmful” (cars, construction works, etc.), irritating the psyche, and what sounds are favorable (birds singing, the sound of water, etc.). There are quite a few points that had both beneficial and unhelpful sounds, so we categorized them as ‘mixed’.
We recorded all the data about sounds in Notion application, but soon we faced impossibility to get a connection to such a database. It was decided to move everything to Google Drive. Attempts were made to display our data as a web map. Points were displayed according to their properties: quality, geolocation, dB level. At the same stage, a gradient scale was created to show what color corresponds to what level and quality of the recorded sound. We were not able to find free ways to display the data correctly and beautifully, so we decided to switch to ArcGis Dashboards.
We processed the data in Java Script to automatically categorize the points into classes, and then loaded the resulting layers into ArcGis Dashboards. We ended up with two layers: the points with their properties and the interpolated grid. To the resulting map we added a table with information about each point. This included the ID of the point, the date the point was recorded, the average noise level, the sound quality and what kind of sounds are heard. Also, when you click on a point, you can get a link to the audio to see exactly what that location sounds like. In addition, a map legend, decibel frequency, perception of sound and general information about map have been added.
Challenges we ran into
Initially, we chose the Notion app to record and store all the sound data. It seemed
convenient for structuring information about each point: location, noise level, sound quality and
additional descriptions. However, we soon encountered a technical problem - the inability to
connect to the Notion database at a moment's notice made it difficult to work with the data.
This led to considerable inconvenience as the data became inaccessible and we were forced to
look for a more reliable solution. Eventually, the decision was made to move all the information
to Google Drive, which took additional time to migrate the data and set up the storage
structure. Google Drive provided stable access to the data, allowing us to continue our work
without interruption.
We aimed to create a user-friendly and understandable web map that would display all collected data, including sound quality, geolocation and noise levels in dB. Initially, we tried to use free visualization tools. However, none of them were able to correctly display the data in the right format, especially given our requirements for visual aesthetics and functionality. This became a major hurdle, as the map was to be a key deliverable of our work.
After exploring various options, it was decided to switch to ArcGIS Dashboards, whose platform allowed us to create a map that met our requirements. To do this, we first processed the data using JavaScript to automatically categorize points based on their properties.
We aimed to create a user-friendly and understandable web map that would display all collected data, including sound quality, geolocation and noise levels in dB. Initially, we tried to use free visualization tools. However, none of them were able to correctly display the data in the right format, especially given our requirements for visual aesthetics and functionality. This became a major hurdle, as the map was to be a key deliverable of our work.
After exploring various options, it was decided to switch to ArcGIS Dashboards, whose platform allowed us to create a map that met our requirements. To do this, we first processed the data using JavaScript to automatically categorize points based on their properties.
What we're proud of
Aesthetically, the project looks really good, there is a coordination between the base map,
the representation of data and the color palette in the dashboard. We also managed to
represent the data successfully, but not without trouble, so it was more gratifying.
What we learned
Through this project, we improved our programming skills, particularly in JavaScript, gained
experience with ArcGIS Dashboards, and learned about the challenges of collecting and
visualizing complex environmental data effectively.
What's next
In the future, our database and the structure that we have created can be used as the basis
for a large project with an interactive map and frequently updated data. For example, you can
install small sound sensors in the park and upload sound data in real time.
Sources
Cui, P.; Li, T.; Xia, Z.; Dai, C. Research on the Effects of Soundscapes on Human Psychological
Health in an Old Community of a Cold Region. Int. J. Environ. Res. Public Health 2022, 19, 7212. https://doi.org/10.3390/ijerph19127212
Students
Agata Kiseleva and Pedro Bonilla Artigas
14th intake
Supervisor
Juliane Cron, M.Sc.
Keywords
Sound mapping, Olympiapark, Noice pollution