In the spring of 2019, researchers from the NYU Center for the Study of Asian American Health CSAAH) collaborated with community leaders in Chinatown to assess the impact of long-term construction on the health of older adults in the neighborhood. CSAAH’s work around this topic found that the urban soundscape unique to Chinatown may be a factor affecting the health and quality of life of its residents.
In February 2020, three noise-sensing monitors were installed on the 13th floor rooftop of Chung Pak LDC, a low-income senior residence. These sensors consist of a low-cost, custom digital microphone and a single board computer, and captured the longitudinal measurements of noise through sound pressure levels. The noise sensors were strategically placed to track average noise levels of the area around Chung Pak.
Chung Pak Noise Sensor Data
Sound waves are created by objects as they vibrate. These vibrating waves are transmitted through the air, creating pressure. The patterns of pressure of the sound wave corresponds to how loudly one hears a sound, with quieter sounds producing sound waves with relatively small sound pressure levels and loud sounds producing sound waves with higher sound pressure levels. These sound pressure levels (SPL) are measured in units of decibels (dB(A)). Because the range of sound is so wide, sound pressure levels are represented by a logarithmic scale rather than a linear scale. For example, 60 dB(A) is twice as loud as 50 dB(A), and 70 dB(A) is twice as loud as 60 dB(A).
These heatmap visualizations of SONYC sensor data recorded from the rooftop of Chung Pak show average loudness for each hour of every day for a selected month, recorded as sound pressure level (SPL) and measured in decibels (dB(A)). Each square represents an hour and is shaded according to its loudness. Each square represents an hour of the day. The color scheme shows a range of SPL values between 55 dB(A) in dark blue, and 85 db(A) in yellow. The darker the color of the square, the lower the sound pressure level – this can be interpreted as lower noise level. The lighter (more yellow) the color of the square, the higher the sound pressure level – this can be interpreted as higher noise level.
1. Let's start by looking at two specific days, April 19th and September 14th,
It is relatively quite at 2 AM on April 19th. It has an average sound pressure level of 59dB, roughly as loud as normal talking. When we color a block based on an average measurement of around 59dB, it looks like
It is relatively noisy at 9 AM on September 14th. It has an average sound pressure level of 80dB, as loud as an alarm clock. When we color a block based on an average measurement of around 80dB, it looks like
2. When we plot the data for each hour through out the day, it looks like this
3. When we plot the data for each day through out a month, it looks like this
4. When you hover your mouse over a block, a pop-up will show you details about the data for that hour and that day
Great! Now you are ready to play with the Heatmap!
I want to check the sound levels for ...
and compare it with
As you compare different heatmaps, here are a few things to think about:
Which hours comprise the quietest and noisiest parts of the day?
Were there seasonal changes in noise levels?
Which month had the lowest sound levels? Why do you think that’s the case?
Were there days or weeks with unusual noise levels? What do you think was happening during those days or weeks to cause these differences?