HEMA Theme 3: Soundscape Ecology Research
Our most recent research thrust in the HEMA lab has been the exploration of sound dynamics in landscapes. We refer to this area of discovery as “soundscape ecology”. We are interested in characterizing the patterns of sounds that occur across different spatial extents and time periods. The connection to land use is obvious, we are interested in the types of sounds that occur in various major land use types and how planning and management might impact the quality and quantity of natural sounds in these environments.
The Tippecanoe Soundscape Study is one of the first that embarks on continuous recording of sound in several different habitats and land uses. We collect data for 15 mins starting at the top of the hour and use information in these sound files to quantify biophony (sounds from biological organisms), geophony (sounds from the geophysical environment, the most common is rain and wind), and anthrophony (sounds from humans, which are generally machines).
The work also includes the application of acoustic sensors in sensory arrays. Sensor networks are complex groupings of instruments that require synchronization technologies, wireless communication (data transfer and sensor control), independent power support (e.g., solar power), among other technical challenges.
Currently (as of February 2011), three graduate students are working on different aspects of our soundscape ecology research theme. Work is being conducted on the processing of massive sound files for the use of quantifying spatial-temporal patterns of sound across landscapes, the use of acoustic devices to monitor the behavior of anurans in tropical rainforests, the correlation of Biophony metrics with on the ground census data, and issues surrounding the conservation of soundscapes in natural areas.
The work that we do requires a diverse set of skills and approaches. Our conservation work for example is based on the theories of managing the commons. In particular we are using the Ostrom SES framework to construct a framework for managing sound as a common pool resource.
We are also employing sophisticated tools to extract sound information from sound files. This includes a host of machine learning tools and multivariate statistical methods. We are also attempting to interface the soundscape metrics that we develop to remote sensing imagery. For example, we are interested questions such as "how does biophony vary across vegetation structure?"
A very successful part of our program has been the development of web-based query and analysis tools that allows users to view acoustic trends. The database currently is over 30,000 records (15 min recordings) and is approximately 30 TB in size.
We are also hosting data from other researchers around the world who are collecting soundscape data. We plan to make the data that we collect available to the broader research community.
To learn more about our work, visit the Soundscape Ecology Projects website.
Last updated March 6, 2011