For the more information about natural sounds and night skies in the National Park Service, please visit http://www.nature.nps.gov/sound_night/.


Types of Data

Several types of data are needed to accurately quantify the soundscape of a park.

24 hour spectrogram

This spectrogram displays 24 hours of sounds from Olympic National Park. Each row displays two hours of acoustical data, beginning and ending at midnight. Brighter colors indicate louder sounds, which stand out from the quiet blue background. The brightest spikes are mostly high-altitude aircraft. NPS Photo.

Sound Amplitude and Frequency

Describing the soundscape.

At each site, monitoring equipment takes sound pressure level and frequency readings. The sound pressure level (loudness) is recorded in decibels (dB) and the frequency (pitch) of a sound is recorded in hertz (Hz). Currently, sound equipment used in the parks can record sounds from 12.5 to 40,000 Hz, which exceeds the human hearing range. High frequency sounds (a cricket chirping) and low frequency sounds (water flowing in a river) often occur simultaneously, so we split the frequency spectrum into 33 smaller ranges, each encompassing one-third of an octave. For each one-third octave band, dB level is recorded once per second for the duration of the monitoring period. Recording the sound intensity of each one-third octave band (combined with digital audio recordings described below) allows acoustic technicians to determine what types of sounds are contributing to the overall sound pressure level of a site. At sites where the sound pressure levels are very low, it is ideal to monitor with low-noise, high-sensitivity microphones. For some types of analysis, dB levels are A-weighted (dBA) to more closely represent the sensitivity of the human ear to different frequency ranges (see the Science of Sound for more information on A-weighting).

Sounds can be represented visually using a spectrogram, an image that incorporates frequency (y axis), time (x axis), and amplitude (brightness of color).

a.Spectrogram of bird calls b.Spectrogram of a high-altitude jet c.Spectrogram of rain and thunder d.Spectrogram of a vehicle

Different sounds occur in different frequency ranges. For example, bird song (a) is much higher in frequency than a jet (b), thunder (c), or a vehicle (d). In these spectrograms, brighter yellow indicates louder sounds. NPS Photos.

 

Digital Audio Recordings

An archive of sounds.

To provide an accurate characterization of the natural and non-natural acoustic conditions in a park, it is important that we know not only how loud the park is, but also the sources of sound. Digital recordings (.mp3 or .wav files) provide us with one way to identify specific sound sources. This can help park staff when determining noise intrusions and how to manage them. Park rangers also use sound recordings for educational programs (see Teaching Soundscapes for example programs).

Advancements are continually being made in acoustical data analysis.Therefore, it is crucial to obtain high-quality archival recordings that can be used to compute any conceivable metric for future analysis. It is almost certain that metrics specified today will be inadequate to meet all future needs. Digital recordings provide an archival record of the soundscape of the area.

An acoustic technician records the sounds of Mount Rainier National Park

An acoustic technician records the sounds of Mount Rainier National Park. NPS Photo.

On-site Listening

Do you hear what I hear?

Acoustic technicians conduct several hours of on-site listening at each monitoring location. This involves sitting quietly and noting each sound heard in a handheld digital device. From these data, we can discern how often and how long each sound is audible to a human observer at the site as well as how much noise-free time occurs in the area. On-site listening data create a sound inventory of each site and help ground-truth later data analysis.

 

Weather station

Weather station. NPS Photo.

Meteorological

Weather matters!

Sound waves traveling through the air are influenced by air temperature and humidity. Wind can sometimes mask other sounds, making them difficult to hear. Consequently, each acoustical monitoring system is paired with a miniature weather station that logs temperature, humidity, wind speed and direction for the entire monitoring period. Days with wind speeds in excess of 5 meters per second are excluded from certain types of analysis due to interference.

Last Updated: February 06, 2012