A Citizen Science Project Ethnography by Ann Jeffers & Victoria Kim
What project did you decide to join and why? We chose to join Sounds of New York City which is a project focusing on sound pollution in NYC. We chose this project because it fit into our interests of sound and urban living. The project studied something that we both experienced in our everyday lives and that we have a personal stake in. What are the goals of the project? The objectives of this project were to create technological solutions for: • Monitoring noise pollution systematically and constantly at a city scale • Accurately describing acoustic environments in terms of their composing sources, using machine learning • Supporting citizen participation in noise reporting and mitigation • Helping city agencies to take effective, information-driven action to mitigate noise pollution How did it demonstrate the characteristics of citizen science that you explored in this course? One of the purposes of this project was to broaden citizen participation in noise reporting and mitigation. This is in line with the characteristics of citizen science, in which the general public is able to raise awareness and stay informed on issues they might be interested in through their participation in citizen science projects like SONYC. By being a part of a project that explores noise pollution, citizen scientists were able to take agency in noise identification and reporting. The citizen scientist would be more well-informed and aware of noise pollution and the components of noise in New York City as a result of their participation in this project. Through this citizen science project, scientists are able to collect data on a much larger scale than they would have been able
to if they were responsible for doing all of the sound identification on their own. The data collected from this project contributes to the process of identifying, monitoring, reporting, and mitigating noise pollution in the New York City area. Thus, the citizen scientist is able to observe a real-world impact of their project participation. Citizen scientists are also able to participate in projects without sacrificing the time they need to spend on their ordinary jobs and lives, which is supported by the fact that this project takes up minimal time with its ten-second audio clips and subsequent quick click-based identification process. What was the science of this project? The system used in SONYC includes a network of sensors and citizens for large-scale noise reporting. The data collected from these sensors goes through a cyber-infrastructure that analyzes, retrieves and visualizes data to facilitate the identification of important patterns of noise pollution by citizen scientists. This noise “mission control” center of sorts is intended for decisionmakers at city agencies to strategically deploy the human resources at their disposal to act on the physical world. Identifying sounds is the way citizen scientists contribute to data collection, and this is done chiefly through auditory observation of sound clips as well as visual observation of sound wavelengths. In terms of knowledge and practices, what science did you learn from participating in this project? Please be specific! As most citizen science projects go, not a lot of background knowledge or skill is required in order to participate in this project. Most of the skills and scientific practices utilized in the data collection process was simple auditory observation of sound clips, though visual observation of sound wavelengths was also involved. A lot of conscious decision-making was required when identifying sounds that were “on the fence,” between categories, so to speak. As observation is one of the most important practices used in scientific data collection, this is to be expected. The SONYC team of scientists includes experts in acoustics, machine listening, distributed networking, citizen science, digital media,
machine learning, data analysis and visualization. The project thus synthesizes human observation and machine learning to monitor noise pollution through both man and machine. Describe what you were required to do in order to be a participant in this project. Participants were asked to listen to short audio files of recordings made around the city and then asked to tag the sound files with the sounds present in the file. The sound waves of the file are also visually represented to help associate sounds with different patterns of represented sound waves. This is supposed to help the program produce an algorithm that will identify noise without user input. What question were you exploring and what sort of evidence did you generate as a participant to be able to answer that question? We were exploring what sounds exactly are considered pollution and not just normal sound. We also explored what sounds are more frequent in NYC than others. As a participant of this project, we experienced evidence of both of these questions as we were made more aware of the different sounds of New York from participating in their identification. The more sounds we identified the more experience we had with New York Sounds that we could form our opinions on. However, we do realize that this project does not represent ALL of New York as the sounds are pre-recorded for us, that is why we collected some of our own sounds that we come by on a daily basis to bring in our own experiences. Provide examples over time as evidence of your involvement. How does your citizen science log capture this process? We continued to listen to different sound clips on Zooniverse and identified them according to the categories presented by the SONYC team. Because the sound clips we identified were not in any particular order, our evidence of involvement is mostly in our citizen science log, where we recorded our reflections as we progressed through this citizen science project, additional questions or criticisms we came up with as a result of our participation, and our experiences interacting with the project website.
What were your achievements? We created twenty logs containing our observations, reflections, and questions as we participated in this citizen science project. We also recorded a few noise clips of our own, containing sounds that we hear and deal with on a daily basis within our own lives. This was to see, on a more personal level, the noise that we deal with every day as a part of living in New York. Since SONYC gives us pre-recorded noises without any information about the area or time it was recorded in, this personal tangent helped us better understand the level of noise pollution we experience in our lives. Use Bonney (2009) posted to Resources in NYU Classes to identify the type of citizen science project into which you would classify the project you studied. Sounds of NYC is a citizen science project in which citizens take part in identifying and labeling the sources of the different sounds found in the examples of urban audio recordings taken from the data that scientists’ sensors have already collected. This fits most closely with Bonney et al’s definition of a Contributory Projects, which are generally designed by scientists and members of the public primarily participate by contributing data. Citizen scientists contribute data through the identification of noises in sound recordings, but not much else is required of them; we as participants to not analyze findings, refine or interpret data, or draw conclusions. What are some recommendations you have for the developers of this project? Our biggest complaint was that in the process of identifying sounds in the noise clips, the categories we were given were not very well-organized. For example, there were four or five different types of categories available for specific types of machine engines and varying sizes of construction equipment, but only two categories to choose from when identifying crowd noises. Some sounds did not have a category option at all, such as airplane noises. This was
frustrating at times because we knew the results could not be 100% accurate without the exact categories available to assign each of the noises too. For future iterations of the project, I would say this categorization of noises is the biggest challenge that developers must work on to improve. Other aspects of the project, such as the tutorial or identification process itself, were fairly straightforward and unproblematic. How did this experience of citizen science compare with what you are doing in class as part of this course? In class, we are learning about climate change through ocean acidification, ozone destruction, and global warming. These are all processes that occur through humans putting pollutants into the oceans and atmospheres that cause change to the already present processes of homeostasis in these ecosystems. In this project, we focused on sound pollution, pollution that is different from what we are studying in class but still has an effect on an ecosystem’s well being. Sound pollution is different because noise does not accumulate over time but different noise levels that are produced momentarily in the same location do have an effect on the human and animal populations that are present there.
Ann | Log 1 | 9.11.19 After looking over the wealth of Citizen Science projects that Zooniverse and the other links had to offer I was particularly attracted to SONYC. Sound pollution is an absolute reality of living in New York that can often be forgotten (thank god) but in critical cases can be extremely present day in and day out. I find myself to be a pretty observant person in my day to day life and I often get annoyed by the constant roar of the city. But past this feeling of annoyance, I am uneducated in the deeper consequences of noise pollution and am interested in finding out more and helping identify problems that hopefully can be mitigated one day. Ann | Log 2 | 9.21.19 I have registered an account with Zooniverse and SONYC and am ready to start identifying sounds to help identify different contributors to sound pollution in New York. I have visited the SONYC website and have found that most of the contributors on the project are from NYU, specifically the Music and Audio Research Laboratory. I am very interested in this space and would like to go visit the Lab. I have also toured the Zooniverse and actual website of SONYC, I am interested in who has participated as a volunteer in this project thus far and where they are getting their noise samples. Ann | Log 3 | 10.3.19 Today I watched the tutorial on how to correctly identify sounds for SONYC by using their sound visualization screen. Sounds are visualized on the bar according to the amount of energy they give off at a present time. High energy is represented by yellow and low energy is represented by purple. The higher the sound’s pitch the higher up it appears on the sound visualization bar. This reveals different sound patterns that can be identified. Ann | Log 4 | 10.14.19 I have helped identify about 30 sounds. I find the process to be sort
of frustrating because the sound identifiers are not specific and are lacking some very common new york sounds that should be monitored as part of sound pollution. For example, I got a clip with air traffic sounds and there was no way that I could identify the clip. Also for people talking they only have a small group (3 people) and the large crowd (demonstration size) and some of the clips I have gotten have been distant kids or medium-sized groups of people talking. Ann | Log 5 | 10.14.19 Today I held a party at my small apartment in the East Village. Weekends are usually very noisy times in my neighborhood and noise complaints are much rarer than they were in my suburban hometown. I attribute this to a younger and more lively population who keep later hours and are less apt to be disturbed by loud music late at night. I was however always aware of how loud we were being and what the neighbors could hear: shouting, music, etc. However, these are common noises in my building on the weekends. Victoria | Log 6 | 10.16.19 I identified about 20 different sound clips from the project website. The process itself is simple and straightforward, but there are not enough options for identifiable noises in the recordings. For example, there are four different option choices for vehicle engines, but none for noises that are more vague and difficult to specify, such as mechanical beeping sounds. While these are certainly loud enough in the recordings to contribute to noise pollution, the website did not provide any categories these sounds would fall under. Victoria | Log 7 | 10.20.19 I helped identify another 20 sounds. Some of the most common sounds seem to be construction noise, traffic noise, and chatter. I noticed again that the category options for sounds we could select were lacking - I heard what seemed like a bird from one of the noise recordings, but bird noises were not provided as a selectable option. Airplane noises are also not a viable option. This seems a bit odd to me considering how specific some of the other options are; some-
how there are no bird or airplane categories, but four selectable categories for different types of saws. Ann | Log 8 | 10.21.19 Ever since my party, I have been thinking about residential noise in New York. Aside from the car noises and construction, there are a whole plethora of noises that travel when living in small proximity to many people. I often hear parties with large groups of people talking, music, sports games, yelling, baby cries, dogs barking, and occasionally the vibration of my neighbor’s phone through our shared wall. This led me to start thinking about the psychological effects of always being exposed to noise and having everything I do able to be heard. I decided that I have built a large resistance to sound in New York and block most of it out. Sometimes I enjoy hearing sounds like a football game and people cheering because it makes me feel less alone and I have good memories associated with those sounds. But sometimes when the sounds are repetitive and annoying like a dog barking or a baby crying I can get really annoyed. Victoria | Log 9 | 10.22.19 I identified 10 more sounds. It seems that a lot of the recordings contain medium-sized crowd noise from a reasonably far distance, yet medium-sized crowds are not an option that we can select. Once again, the project itself is straightforward and easy to figure out, but its lack of provided options for noise category identification may prove as a limitation for its findings. Victoria | Log 10 | 10.25.19 I identified approximately an hour of sound recordings. I’ve been noticing that a lot of the sound recordings on this project contain more mild to moderate levels of noise. The website says the recordings were taken from various parts of the city for more accurate representation, but from my experience living and working in many different neighborhoods in New York, it feels like the average noise levels on any given day are much louder than these recordings. I think it would be interesting for Ann & I to record some sound
recordings of our own and see how they compare to this citizen science project. Ann | Log 11 | 10.28.19 Today, I worked on identifying more sounds for the Zooniverse project. I am still upset about the lack of identification tags for the sounds but I figure that the leaders of the project probably limit the possible responses on purpose. I now am thinking about what sounds are considered “pollution” and what is not. Sound is for most hearing people and uncontrollable external input that is a fact of life. Some sounds are good sounds. Construction I can understand to be a polluting force that needs to be monitored by the city. Noise pollution is also a good sign of spotting other types of pollution. For example, car noise is a good representation of traffic congestion and the carbon pollution that is happening with the influx of cars in the city. Victoria | Log 12 | 10.30.19 Because my internship is located in Midtown near Times Square, I deal with a lot of noise on a daily basis. I reflected on the time I spent identifying sounds for this project, and I wondered what the difference was between the levels of noise pollution in different neighborhoods and areas in New York, as well as the severity of noise pollution in more populated areas like Midtown. I hope the results of this citizen science project will give me more insight into the details of noise pollution in New York City. Victoria | Log 13 | 11.02.19 I spent another hour identifying sound recordings. It still frustrates me that the categories for the types of sounds are not as in-depth or broad as I would like them to be, but considering this citizen science project is over 90 percent completed, I do not think the website will be updated. Ann | Log 14 | 11.4.19 Today I tried to record some of my own “sound pollution” clip a la the Zooniverse project. Some sounds are easier to capture on my
phone than others. I was very successful in recording a siren. Listening to the varying types of sirens (cop vs ambulance) made me wonder who is in charge of deciding those noises. It is a very important job because it essentially impacts the whole city. I’ve been relating a lot of this project to stuff that we have been talking about in class. Sound pollution is not the same as the pollution we have been talking about in class because the sound doesn’t accumulate or remain present in the atmosphere it is, however, a good warning sign for processes that do pollute in these ways. Additionally, a lot of noise pollution is caused by a small group of people (developers in charge of construction and the people who decide siren tones) who create a problem that affects a large number of people. Car noises, however, are more at fault of the individual. Victoria | Log 15 | 11.07.19 As I am identifying more sound clips, I have been wondering if there are legal measures to prevent noise pollution from becoming too extreme. I know of a friend who lives right next to a construction site who finds the noise very difficult to deal with. She had sent a series of complaints to the city of New York, hoping for the noise level to be brought down to a more manageable level, but to no avail. It seems rather unfair as she is still paying the same amount of rent as she was before the construction project started, yet she is living in a much less desirable situation. I wonder if there is anything to be done about this. Victoria | Log 16 | 11.10.19 I have identified more sound clips for a little over an hour. I’ve also started recording my own sound clips of noises I encountered throughout my daily routine, as a sort of artifact and my own experience with the issues relating to this citizen science project. It is interesting to note how loud some of the sounds are in the environment around me, and how I’ve gotten used to ignoring some of these sounds that would easily bother someone who was not used to them.
Ann | Log 17 | 11.11.19 I recorded a few more sounds today and did my final round of noise identification on the Zooniverse project. The project is in its last stages and I will be interested in what the study finds. Going forward I would love to learn more about the people at NYU Tandon that orchestrated this project, what they set out to do and whether the data they collected supported or negated their hypothesis. I wonder if they will try to incite change in the city with their findings. Victoria | Log 18 | 11.12.19 I identified another hour of sound recordings. I am curious to see what an “acceptable” level of sound pollution would be - how quiet do we need our environment to be in order to consider it a “healthy,” environment, so to speak? Some of the recordings are at a volume that isn’t loud enough to be too bothersome, but those living near a hospital or a construction site would certainly have a hard time relaxing and resting with all that noise. Victoria | Log 19 | 11.27.19 For the Thanksgiving holidays, I went to Montreal, Canada. Being in this area has made me realize how loud New York City is. Even though Montreal is a fairly large city, it is much quieter than New York. The pollution is overall much less intense here - there is less garbage on the streets, public places are cleaner, and people tend not to yell or honk loudly. I felt much more relaxed while in Montreal, and wondered if that had any relation to the quieter atmosphere. I want to know more about the effects of noise pollution on humans. Ann | Log 20 | 11.28.19 I am on vacation in Deale, Maryland for the holiday. Deale is an extremely small town on the Chesapeake and there are little to no ambient sounds. I think about how strange the quiet is after loud New York City. My parents asked me if I sleep better with less noise and I’m not sure. Even though I am trying to stay aware of the impact of ambient noise around me at all times the human brain is so good
at adapting that the immense quiet doesn’t immediately change my body or mind’s regular state. This makes me question the importance of monitoring noise pollution as it relates to humans’ well being. I would like to know its effect on the broader ecosystem of the city (animals, etc).
Audio Visualization of a large group of people talking
the categories avzilable to classify sound clips on Sounds of New York City
Audio Visualization of a large group of people talking with what sounds to be a plate dropping
Audio visualization of construction noise (jack hammer) and a medium group of people talking
Audio visualization of more construction noise with traffic
The ongoing results of Sounds of New York City are illustrated on this graph. It was interesting to see the percentages of sounds identified by the citizen scientists contributing to this project. The most prominent noise identified in the sound clips was “people noise� featuring various sizes of groups talking, followed closely by two types of machine engine noises. It was surprising to see that siren noises did not make up a large percentage of the identified sounds, as in our personal experiences with noise around the city, sirens were more prominently featured.
Victoria Interacting with this project as citizen scientists has allowed us to get in touch with the reality of noise pollution in New York City. The loudness of the city is something that we have unfortunately grown accustomed to, but being forced to listen very carefully to sound clips taken from various parts of the city in order to identify and categorize them has made me much more aware of how severe the noise pollution is. I noticed that I’ve instinctively been “tuning out” a lot of noise that I encounter in my daily life, such as construction noises, sirens, and crowd noises, but being more aware of the sounds in my environment have made me realize how this level of noise must be impacting the animal populations around us as well as our own mental and physical health as humans. It also makes me question the severity and impact of other environmental disturbances in metropolitan areas, such as light pollution. In raising my own awareness of scientific issues and allowing me to explore further questions of my own, this citizen science project has clearly made its intended impact.
Ann
Being a citizen scientist has made me become more aware of my actions and their impact on the systems around me. I am much more aware of the interconnected nature of the world and my place in it. Simple realities of my daily life now prompt questions and exploration that I would never have occurred to me. Before this project, I never thought about the frequency of ambient city sounds or their type. It also never occurred to me that these sounds, if studied, could reveal deeper truths about my environment. I now am motivated to ask more questions about my surroundings and interrogate the method in which I perceive them. After identifying the problem as I have don in the Sounds of New York project I seek to find out the ripple effect each problem has on our interconnected environment.