Isobel Map

Cowling Arboretum Isobel Map

The concept of an isobel map first appeared in Barry Truax’s book Acoustic Communication.  Truax defines it as a map that “joins equal points of sound level (e.g. in units of dBA)” (Truax, 72) through isobel lines.  Isobel maps can be used to quantify baseline conditions when SPL measurements are taken during times when the soundscape is reduced, such as early morning or late evening.  Alternatively, they can classify points of high activity, if recorded when the soundscape has increased sound events.

The first iteration of an isobel map (Fig. 1) appeared in a 1974 publication – The Vancouver Soundscape, No. 2, Music of the Environment– as part of the World Soundscape Project and mapped Stanley Park in Vancouver.  SPL measurements (dBA) were taken on footpaths at “intervals of about 100 yards, between the hours of 10am and 4pm, on several successive Wednesdays during May, June and July, 1973” under constant weather conditions (Truax, Isobel).  Three successive measurements were taken at each location and averaged to attain a representation of Leq (equivalent continuous noise level), which is the average of SPL readings.  The map parallels a topographic map, by connecting points of equal SPL in unique formations.  It is valuable to see a broad scope of SPL measurements throughout Stanley Park that reveals where differences and commonalities may exist.  These values visually structure the Park’s soundscape, making the general sonic composition of the area readily apparent.

The fixed nature of SPL measurements on Truax’s map is problematic.  It is impossible that SPL would remain constant, as topography does, because sound continually fluctuates according to the dynamics of sound events.  Hence, concretely defining an entire area – along a singular isobel line – as having an SPL reading of “50,” for example, misrepresents the dynamics of that location and falsifies the true physical structure of the soundscape.  In the Truax map, the isobel lines are precisely drawn with complex curvatures and unique shapes; permanence of SPL in these areas is implied.  These fixed isobels not only distort the physical characteristics but also undercut a central tenant of sound: that it is ephemeral and uncontrollable, which is at odds with the controlled and fixed construction of the lines.

In addition to the misrepresented depiction of Stanley Park’s isobels, the methodology is problematic.  Truax claims the SPL measurements were recorded on “footpaths” (Truax, Isobel), indicated on the map as the wide areas bordered by solidly defined lines.  Yet, there is no description of how the SPL measurements and isobel lines within the interior spaces of the Park were recorded and created.  Did he walk off the path yet not mention this in the figure caption?  Or, are the interior lines extrapolations and assumed values?  If the latter is true, by speculating about the aspects of the Park’s SPL, the map does not truly depict the soundscape because it falsifies measurements.

The second iteration of an isobel map (Fig. 2), constructed in 1977, also appears in Acoustic Communication.  The map charts SPL readings from Dollar, Scotland.  It is more simplified than the map of Stanley Park, and does not resemble a typical topographic map because although the isobels are generally incomplete, and they do not fully connect.  This map is also problematic because, like the Stanley Park map, it also fixes SPL readings along the continuum of isobels, thereby undercutting the dynamic and ephemeral nature of sound and misrepresenting the potential SPL variability of Dollar’s soundscape.

The Cowling Arboretum Isobel Map is the result of applying methods that capture the dynamic and variable characteristics of the Arb’s soundscape.  Truax’s map represents SPL measurements recorded within Stanley Park over three months.  By comparison, SPL measurements were taken from the Arb in one day (Sunday, May 26) under cloudy skies, light drizzle, and temperature in the upper 50’s.  Recording within a singular time is advantageous because weather conditions are constant at all locations, thereby maintaining consistency of variables.  For example, weather may influence animal activity and hence biophonic sounds.  While Truax tried to control for confounding variables, by recording on the same day of the week multiple times under similar weather conditions, atmospheric and biophonic variations most definitely occur across days.  Hence, Stanley Park’s isobel lines may not accurately reflect the correct relationship between these variables.  In contrast, while variability always exists, recording the entire lower Arb in one day minimizes confounding factors.

Another problem with Truax’s methodology derives from the practice of taking three SPL readings and averaging them to attain a representation of Leq.  Representation is highlighted, because Leq should be the continuous average of SPL – recording continuously over a time period.  However, taking three discrete measurements is not true Leq, but an approximation of it.  To attain accurate Leq, SPL readings from the Arb were measured using the Ipad + I-test mic to record continual SPL for a minute at 35 locations, one at each path juncture.  For clarity and ease of viewing, not all measured locations are shown on the map.  Files were stored in the Ipad with a numerical value and the locations were geo-tagged using GPS Toolbox on the Iphone.  Thus, locations could be precisely matched in Google Earth and Photoshop.

Much like Truax’s approach, a limitation of this method is that only the paths were used to catalog SPL.  Hence, parts of the Arb’s interior are not reflected on the isobel map.  For this reason, and to reduce misrepresenting the ephemeral quality of sound, the Arb’s isobel lines were not constructed in a topographic manner.  Rather, the Arb was divided into three regions containing a dynamic range of 6db.  This method has two benefits.  First, creating broader zones reflects the soundscape’s structure while leaving open the potential for change within them.  Consequently, zones better communicate organization while maintaining that sound is dynamic and ephemeral.  Second, 6db represents a doubling of SPL.  Because energy that exerts two times the pressure is a benchmark value within environmental acoustics, it is logical to divide the Arb’s isobel map into regions separated by 6db.

Solid lines join points of similar SPL on the Arb’s isobel map.  These isobels differ from the two predecessor maps because they do not qualify a value concretely tied to them.  SPL values are shown at each recorded location and marked by vertical pillars.  This method effectively outlines regions by assigning concrete values to isolated spots while not extrapolating these along a continuous “topographic” isobel line, as shown in the previous isobel maps.  As a result, similar SPL readings are connected not by direct association but by their dynamic range.  Dashed lines mark potential convergence of zones and are not associated with actual recorded SPL values.  Rather, they link regions that could overlap another region or depict areas where the SPL is possibly similar.

Data from the isobel map reveal the Arb is quietest (37-43dBA blue zone) in the interior where heavy forestation exists.  Areas closer to the road are louder (44-49 dBA green zone).  These locations are still protected by trees and brush that diffuse the anthropogenic sound, causing the SPL to remain lower than the maximum observed SPL (50-54dBA red zone) nearest the road.  Regions with the highest SPL values are logical because vehicles exert a high amount of sound power and hence have increased SPL.  Thus, areas of the Arb in direct proximity to the road are vulnerable to elevated SPL.

There are several SPL outliers.  The first outlier is the 51dBA at the north end of the Arb.  The roaring Canon River and the proximity to the road on the opposite bank explain this measurement.  Clearly it seems out of place because it is surrounded by dense forestation and near the blue zone.  However, this SPL highlights that anthropogenic sounds are not the sole sources of high SPL, but biophonic sounds also can have elevated SPL.  Second, it may be expected that 43.8dBA and 40.4dBA near the southern end of the Arb would be in the green or red zone because of their proximity to the road.  However, 40.4dBA can be justified because it is sheltered behind a thicket of sound-diffusing trees and also located down a steep hill.  Thus, sound propagation from the road may be greatly diminished by these landscape factors.  Although 43.8dBA appears peculiar because it is unsheltered in the prairie and exposed to the road, this value is two tenths away from 44dBA, the minimum for the green zone.  Thus, it is very likely that had the Leq been recorded longer, this value would have reached 44dBA.

These outliers reveal how cataloging SPL is dependent upon the sound sources, landscape factors, and length of recording.  Furthermore, they highlight how the first two isobel map iterations inaccurately portray SPL measurements from across their soundscapes by fixing them to locals that are extrapolated into definite, single-measurement, and “topographic” isobels.  Rather, as the Arb’s isobel map portrays, zoning enhances the map’s flexibility by maintaining structure that is generalized, yet contains specific examples of SPL while denoting potentially overlapping regions.

These results integrate with the data presented on Arbsonics.  The study revealed that anthropogenic sounds (91%), predominately from vehicle noise (82.3%), outnumber bio-geophonic sounds (50.9%) (these are independent variables.  Hence, the average percentage is not cumulative in that all groups add to 100%, but represents total percent audible per anthropogenic, biophonic, and vehicle group.  Thus, each group’s percentage is independently out of 100%).  When compiling data for Arbsonics, vehicle sounds were audible even though the recorder was stationed in the middle of the Arb (within the quietest dB zone), thereby indicating that the amount of sound-power (the fixed amount of sound energy exerted per unit time) vehicles exerted was so intense that their SPL (the size of the sound energy field, i.e., how far/with what intensity the sound travels) could still propagate into the interior of the Arb.  Attaining SPL measurements reveals that the SPL near the road is four times as loud compared to the recorder location in Best Woods.   This finding validates that over a distance of one mile from highway 19 or one-half mile from highway 3, 4xSPL infiltrates the quieter regions of the Arb such that the anthropogenic sources are audible 91% of the time.  Overall, this highlights how powerful human-generated sounds are and alludes to their potential for influencing biological sounds.

Arbsonics also concluded that Best Woods was acoustically healthy because variety, complexity, and balance were present.  These factors were influenced by the SPL in the region.  The isobel map confirms that there is four times more “openness” in the sound field in Best Woods, because it’s Leq was 40.9dBA compared to the loudest value of 53.8dBA near the southern end of the Arb.  This increased availability of dynamic acoustic space allows for increased transmission and audibility of changes in sound source event characteristics.  Hence, the Arb has an increased potential for variety, complexity, and balance that translates into enhanced acoustic health.

Overall, the SPL readings confirm the results from Arbsonics that Best Woods has the potential for increased sound transmission and reception compared to areas closer to the highways where the sound field is more overpowering.  Furthermore, when subjectively analyzing Best Wood’s soundscape to determine the appropriate location to station the SM2+ recorder, Best Woods seemed to be quieter than peripheral or less remote interior regions of the Arb.  The isobel map reveals four times more SPL in the exterior, thereby verifying that loudness, a perceptual sense, can subjectively be used to qualify the nature of soundscapes and derive information about their physical structure.

The methodological advantage employed to construct the Arb’s isobel map is that it is technologically accessible and easily understood.  However, there are two primary drawbacks to the map.  First, although recording all SPL measurements on the same day reduces confounding variables, physical conditions constantly change.  To mitigate this problem, using an array of time-synched sensors distributed throughout the Arb that catalog Leq exactly in-sync would provide a more accurate measurement of the soundscape’s acoustics per unit of time.  Soundscape ecologists claim this model provides the best way to obtain information about a location’s soundscape.  However, they also acknowledge that the use of multiple sensors is impractical and problematic, because of the cost and technical limitations such as battery life and memory storage (Pijanowski, 1225).  The second drawback to the isobel map is that, even though it is not “topographic,” it is still static and therefore does not fully depict the ephemeral quality of sound and SPL.  A perfect isobel map would be one that animates live and/or modeled SPL measurements, similar to a meteorological barometric pressure map.   However, such a map would be very expensive and, apart from of its interest to soundscape ecologists or nature-enthusiasts, may confer few benefits to the general public.

To make this study longitudinal using the current paradigm, SPL readings should be attained weekly over a season.  The Arb would have to be walked in its entirety one day each week to reduce confounding variables.  Alternatively, the necessity for an automated sensor array system could be reduced by having multiple people with SPL meters positioned at different quadrants of the Arb with time-synched SPL recordings.  If performed at different daily or seasonal intervals, multiple isobel maps could be created that compare the soundscape based upon season, time of day, weather conditions, number/type of sound events, etc.

Overall, the Arb’s isobel map provides a baseline to understand where future soundscape analysis should be conducted.  Most importantly, the map reveals a generalized view of the physicalenergeticcomposition of the Arb’s soundscape.  It provides another concrete measure in conjunction with sound event quantification data from Arbsonics that demonstrates the fragility of the Arb’s acoustic characteristics.  For example, because vehicles are audible 91% of the time and their energy is four times louder near the road, yet are still audible in the quietest region of the Arb, it is clear that soundscape preservation efforts are needed.  Ironically, the most direct way to preserve the soundscape is by preserving the surrounding landscape.  The isobel map makes apparent that external SPL are much more intense than internal SPL.  Thus, not only does the interior landscape need to be preserved, but measures need to be taken to ensure anthropogenic sound expansion from the Arb’s surrounding areas, due to increased industrial development, does not occur.  In sum, inhibiting the external SPL from increasing beyond the loudest measured 53.8dBA will protect the interior of the Arb.  Such action will preserve the internal dynamic range of the Arb and maintain the openness of the sound for effective communication and acoustic health.