Agriculture Reference
In-Depth Information
Broader Adoption of Acoustic Technology for Ecological
Monitoring
The Biophony Grid Portal
Soundscape research at KBS LTER resulted in a pilot grid computing initiative,
led by the National Center for Supercomputer Applications (NCSA) (Butler et al.
2006). The Biophony Grid Portal was developed to demonstrate the potential of
grid computing to enhance collaboration and sharing within and external to the
national LTER community, based on a large acoustic dataset and algorithms devel-
oped by KBS LTER researchers to identify entities in the soundscape. The grid
utilities were developed at NCSA, the access system was developed by the LTER
Network Office (LNO), and the digital data were located on a grid-enabled server
at MSU.
The Biophony Grid Portal was designed to allow an investigator to identify an
entity in a subset of a digital sound archive from a set of available locations. The
recordings on the MSU server were linked to metadata on the LNO server. An
investigator could log onto the grid and access the Biophony Grid Portal via the
Internet. The investigator could then select the entity sound signature from a list
of recognized entities (e.g., train whistle, chipping sparrow, etc.), together with a
location and a range of dates to search. Based on location and the date range, the
subset of sound recordings was retrieved from the MSU server and transmitted
to the NCSA High Performance Computer (HPC). Results were provided via the
Biophony Grid Portal where the investigator could listen to the entity signature,
examine the soundscape spectrograms, listen to the sounds, and retrieve a table of
signature match probabilities based on the recordings examined.
The grid computing infrastructure contributed to LTER Network-level syn-
thetic science. Scalability of solutions has emerged as an increasingly significant
issue, and grid technologies are an important approach to addressing and solving
large-scale data and analytical requirements (Butler et al. 2006).
Current Technology
The application of automated soundscape recording, and subsequent storage, anal-
ysis, and interpretation have advanced considerably over the past decade. Today,
recording technologies are available commercially (e.g., http://www.wildlifeacous-
tics.com) and new models and acoustic sensor innovations are under way. Digital
libraries to archive, analyze, and access acoustic observations have been developed
(Villanueva-Rivera and Pijanowski 2012, Kasten et  al. 2012, Aide et  al. 2013);
sound pattern recognition applications have evolved (Kasten et al. 2010, Acevedo
et al. 2009, Aide et al. 2013, Ospina et al. 2013); and acoustic indices have been
further developed (e.g., Sueur et al. 2008, Joo et al. 2011). The importance of the
soundscape as an ecological attribute has been acknowledged (Pijanowski et  al.
2011a), and a research plan has been devised to apply the principles of soundscape
ecology to monitoring ecological phenomena across landscapes (Pijanowski et al.
Search WWH ::




Custom Search