Environmental acoustic recordings have been ongoing at OKEON field network sites since 2017. Passive acoustic recordings enable valuable research pathways such as species identification, seasonal behavior dynamics, ecological disturbance and stability, diurnal soundscapes, and more. Our acoustic data covers a large variety of habitats and contains vocalizations from birds, insects, and frogs in urban, forested, and mixed environments. ESI also hosts several recorders and tools to capture recordings both passively and actively in the field either using long term deployment recorders or hand held equipment and mixers. Contact ESI if you are interested in doing research using our acoustic recordings or equipment.
Acoustic
Acoustic Data
ESI stores the OKEON field site recordings as lossless .flac files. We have over 13 thousand hours of recordings which have been processed for some soundscape metrics like biophony (sounds made by nature) and anthrophony (sounds associated with humans). Each recording includes the temperature at the device. Our data is rich in a diverse range of acoustic features, from biophony to anthropogenic noises. ESI maintains a collection of 10 minute of recordings every 30 minutes (10 on/20 off) from 2017 to 2021 and switched from 2021 to current with 1 minute every 30 (1 on/29 off).
Soundscape Ecology
Soundscape Ecology is the study of acoustic relationships between living organisms and their environment. Our recordings include day and night acoustic soundscapes and the more extreme highly disturbed soundscapes wrought by seasonal typhoons, torrential rains, and high winds. Our data has been used to measure how disturbed soundscapes respond to and recover to tumultuous events, identify bird species and their call timing, and map the preferred habitat range of owls in Okinawa.
Birds
Avifauna are a rich source of data in acoustic recordings. ESI's acoustic collection is resplendent with high quality passive recordings of birds in their natural environment. From these recordings we've been able to create species occupancy models, determine call statistics, study the acoustic environment that the birds vocalize in, machine learning models to find calls of specific birds and bats, and more. Currently there is ongoing work to create a broad scale machine learning model based on BirdNET models to identify bird vocalizations at the species level for the entire catalogue of recordings at the OKEON field sites.