Cowling Arboretum Study

The National Park Service (NPS) initiated soundscape monitoring and preservation practices in 2000.  The NPS established the National Sounds and Night Skies Division to determine how anthropogenic sounds impact the wild and what management practices need to be utilized to restore or preserve the natural soundscape.  Since the program’s foundation, acoustic monitoring has been performed in more than 70 National Parks.

I was fortunate to be connected with Davyd Betchkal, physical scientist and soundscape researcher in Denali National Park, where I spent time during the summer of 2012 learning the methodology to record and analyze soundscapes.  I used this knowledge to conduct a soundscape study in the Cowling Arboretum at Carleton College.  The study was conducted from September 17th until November 15th.   The primary emphasis of the study was to define the acoustic characteristics of Best Woods, a location within the Arb, during the fall season.  To accomplish this main objective, five goals were established: 1) ascertain the acoustic signatures of Best Woods by cataloging the present sounds to discover their numerousness and domination; 2) determine the daily and overall seasonal acoustic rhythms of Best Wood’s soundscape; 3) qualify Best Wood’s acoustic health and coherence by analyzing the diversity and density of sounds’ bandwidth; 4) establish an audio archive to preserve Best Woods’s soundscape; and 5) create a soundmap of the Arb as a whole that reveals how the sound changes throughout its soundscape.

The Arb consists of two sections: the Upper Arb, and the Lower Arb (Map).  Audio was collected only from the Lower Arb (the area depicted in the Map), which is larger in area and further away from the downtown-area of the city of Northfield.  To complete the goals, a weatherproof/waterproof audio recorder, SM2+ developed by Wildlife Acoustics, was stationed in a deciduous forest region of the Arb, Best Woods.  The SM2+ recorded audio 24hrs/day.  I visited the site every two to four days to change memory cards and batteries.  Because the study was focused on classifying the Arb’s soundscape, the recorder was used to maximize data collection while also capturing decent audio samples.  The recorder operated at 16bit 48000k resolution.  The pre-amplifiers on the SM2+ generate high frequency hiss; to minimize this noise while ensuring the audibility of recorded sounds, low gain levels of +4.5db were used (a maximum of +12 is possible on the SM2+).  As a result, to hear the nuances in the sound samples in the Audio Archive, headphones are recommended.

To analyze the data, I used Acoustic Monitoring Toolbox (AMT), software developed by the NPS.  The program samples audio in specified time intervals and provides spectrograms – visual analysis of the frequencies of sound.  Under the Air Tour Management Protocol developed by the NPS, to attain a statistically accurate representation of a soundscape, 2.5% of the total time recorded is required to be analyzed.   To obtain a statistically accurate representation of the soundscape, 18 days were sampled (generally two per week) and over 36 hours of audibility analysis (listening) was completed.  AMT generates audio samples that are in ten-second intervals from every two minutes.  To perform audibility analysis, I listened to all 721 samples/day in AMT and inputted sound source codes, indicating what is heard, into the program.  Subsequently, I analyzed statistics generated during the audibility analysis using Microsoft Excel to understand the soundscape of Best Woods.