3:40 - 4:10 pm
Winter Use Monitoring - Soundscapes
Shan Burson
Grand Teton National Park

Transcript

Female Voice:  The next presentation, I was going to say that so far it’s been all critters all the time. And now we’re going to actually hear some physical science presentations for the latter part of this afternoon. Shan Burson is with the National Park Service. He’s duty stationed at Grand Teton but he works part time for Yellowstone. And he’s going to be giving you a presentation on soundscapes, winter use monitoring, and I think there are wolf howls involved.  (Yes)  Yes!

Shan Burson:  Ok, I think we’re in business now. (Sorry)  Yup, that’s alright. I’m going to talk about soundwaves, which is quite different from what’s been talked about so far. And I’m also going to talk mainly about winter use monitoring, which is why I took this picture of the road in the winter. Those animals just somehow got in my photo.

Now the first thing talking about soundscapes is what is a soundscape?  Soundscape is just defined as all sounds, both natural and non- natural that occur in National Parks.  Soundscapes also occur outside national parks, but I’m concerned with them inside national parks.

I’m going to play a little medley of sounds right now that were all recorded either in Yellowstone or Grand Teton. I’m going to take a couple seconds beforehand, if everyone could be quiet and actually try to identify what you can hear in here before I start the medley, and then identify the sources of sound.  It’s a minute and fifteen seconds.

(1 minute, 15 seconds of recorded sounds)

The amazing thing about humans, and human ears and brain is that we can listen to sounds like this and identify many of them, if not all of them, and machines can’t do that yet. So that has implications in the research I’ve been doing. You’ll also no doubt recognize that it started with natural sounds, went into non-natural sounds, and human caused sounds and then back into natural sounds. Natural sounds are really important in natural ecosystems, because many animals use vocal communication for breeding and survival. And also natural soundscapes are important to visitors that come to national parks. Surveys nationwide have shown that many visitors come to national parks to experience solitude, to experience tranquility of natural sounds, and to escape from urban environments.

But the appropriateness of non- natural sounds on those natural soundscapes depends on the enabling legislation of a park, the park purpose, as well as within an individual park the management area. National Park Service in their management policies requires parks to protect natural soundscapes, to restore them if possible and to mitigate non- natural impacts on natural soundscapes. Natural soundscapes are considered a resource similar to wildlife and plants, air and water. And more locally, and what I’m going to talk the rest of the time about, is the acoustic standards that were set in the winter use plans for Yellowstone and Grand Teton National Park.

I have 3 slides that go over the last three plans. I’m not going into detail; I just want to indicate here that there are a variety of different acoustic measures you could measure in terms of the impact of non-natural sounds on natural soundscape.

The 2001 winter use plan used audibility. Audibility is just the ability of someone with normal hearing to hear a sound.  And here is percent time audible of motorized vehicles, over-snow vehicles.  Four different zones, those are just management areas, recognizing that different management areas would have different desired conditions for natural and non- natural soundscapes.

Two years later we added a couple acoustic standards, retained the audibility.  Sorry about the columns, every computer’s a little different I see.  We added two more. One is, not only it renders it now can you hear non-natural sounds or motorized vehicles, but how loud are they. We set maximum sound levels which is Lmax.  Leq is an energy average, it’s also a sound level, and I’ll get into that a little later.

Then for the current plan, the interim plan, we used similar categories.  Here we have three management areas. Instead of zones they’re defined, developed areas, travel corridors, and backcountry areas. And then have audibility % time over-snow vehicles are audible, or can be heard, and then the maximum sound levels in those different areas. This was taken from the environmental assessment, and these numbers indicate if the maximum levels were the thresholds that would kick these, the audibility, I mean... anything over these levels would be adverse major impact.

So what does that all mean, and why am I here?  I’m trying to measure the sounds in these different management areas to compare it to what these management actions have set.  So the thing is I need to find out is what are the sounds so I can identify the over-snow vehicles, where they occur, how much, how often, and how loud are they. 

To do that we developed automated sound monitoring equipment that would record, sit out in the environment unattended for days, weeks or months at a time... here you can see the tripod with microphones on it.  The other parts of the system are underneath the snow with the solar panels that are running it. This system does two different things: it records digitally digital recordings, and that gets that % time audible. I’ll illustrate some of these things with examples. It also collects decibel levels, and that we can measure and calculate the maximum sound levels.

More specifically it collects digital recordings in two different ways. One is a sampling scheme... every 4 minutes of the day the system records 10 second samples of recording.  So over the course of the day that’s 360 samples, and if you combine those altogether that’s one hour of digital recording per day.

Also for loud events, things that exceed a threshold that I set  in duration or sound level, it will record about 20 seconds of that, so that way at any site where I’m monitoring I can go back and listen to those and identify what are the loudest sounds in that particular site.  So that’s irrespective of what time of the day that occurs.

The third thing that these instruments collect are sound levels. Using a sound level meter collects decibel levels every second of the day. And I was thinking that 120 or115,000 browsed twigs was an impressive number. I have a little over a million seconds of sound, and also within every second there’s a pitch of sound goes from low frequency to very high frequencies goes within our human hearing. Every second I collect 33 of those 1/3 octave bands, from very low pitch to high pitch.

Here’s some of the equipment that’s inside... use a laptop computer with a program to do the digital recordings, and also to collect the sound level data. Here’s our sound level meter. And the tricky part is to get these things to run through the winter in environments like this. It hasn’t been done anywhere else in these climates except for Denali, when I was using it there before coming down here.

So the management zones I wanted to collect data in were developed areas, travel corridors, and backcountry areas.  The yellow dots are where I collected data over the last three winters. Three winters ago, the first year, we collected data at a few sites just for very short periods of time, but it gave us some data we could compare 2-stoke motorized use, snowmobiles, to subsequent 4-stroke use.

Anyway, this past winter had four sites that were out for most of the winter: one the travel corridor in West Yellowstone, a slight 8,000 feet of the Mary Mountain Trail as a backcountry site, a monitor at Lone Star Geyser, and then one near the ranger station at Old Faithful. Also collected data near Madison Junction over President’s Day weekend for the third winter, and collected about a weeks worth of data at West Thumb and in the Upper Geyser Basin at Old Faithful.

And here’s an illustration of a slide you’re not supposed to use in a presentation (Laughter), but the only reason I have this is because when we do these digital recordings we take them back to the office and then we listen to them. We just created a nationwide, and a park service wide code for the different kinds of sounds that we can hear. And it’s illustrated here, 1 - 20 are non-natural sounds, and 21 – 40 are natural sounds.

I have a series of slides that you’re not supposed to show. This is a spread sheet with these codes that we’ve listened to.  Now each one of these rows is a 10 second sample. This is at Madison Junction over a two hour period Saturday of Presidents Day weekend. So you can look down and any 10 second period may have several different sound sources. 4.1 is a snowmobile.  22.2 is a river of sounds. 25.5 is a magpie. 8.1 are people talking. 1.1 are jets. So you can imagine if you do this for a number of days you can start having some kind of idea of what’s happening there.

Here’s another one. This is just summarizing that one day, Saturday.  And these are just a number of samples that have these particular sound sources.  So whatever you’re interested in, and here we’re interested in snowmobiles and snowcoaches, for the whole day we heard snowmobiles on 77 of the 360 samples. And then the middle columns are the twelve hour period 7 am to 7 pm. And then 8 am to 4 pm, which is the time that we’re actually measuring for the winter use plan.

Another way of looking at that is you can see again, to summarize, different types of sound sources, and then when during the day it occurs, and you can look at snowmobiles. They started, one snowmobile went by between 6 and 7, and then during the day, and then a few people coming back from grocery shopping toward the end of the day. You can take this data, summarize it more, and get just one number of the percent time audible of the snowmobiles and snowcoaches per day, and then plot it over the winter, and this starts in mid December to early March. This is at Old Faithful. These are percent time audible between 8 am and 4 pm. And you can see, and these are just days that were sampled, the average was 61% of the time you could hear snowmobiles.  I also should say I’m using the percent of the samples as an analog to the percent time during that period even though I’m not listening to all eight hours of 8 am to 4 pm.  But with tests it shows that it’s very closely related. Anything that occurs in a fairly regular interval I’m picking up enough samples to make it reasonable to use that as an analog.

Looking at this same data at Old Faithful, 8 am – 4 pm, I’ve just broken this in times of day so you can see patterns of the day. The dark green on the bottom are snowmobiles, the light green are snowcoaches, and the top are those that we know are either a snowmobile or snowcoach, but couldn’t identify from the sounds.  And you can see the pattern over the course of the day, again 61% of the time for the whole winter.

One interesting thing is, after looking at this data from two winters ago, we saw the snowcoaches in the morning between 8 and 9 were the majority of oversnow vehicles.  At the time we thought, well, what’s going on here? It just turns out that the behavior of the snowcoach drivers, is they start up in the maintenance yard and drive over to the ranger station parking lot and then they idle there until they take their trips somewhere. So this past winter we asked them not to do that. I guess they’re warming up now in the maintenance yard, and actually from the preliminary look this year the % time audible of snowcoaches went way down in early morning.

This has data from this past winter. Again, the whole winter season, the percent time audible of these different days of snowmobiles and snowcoaches. Remember 61% a year ago, it went up to 69% of the days. And they showed a little bit more of a pattern here as opposed to just scattered around from 2 winters ago. Clearly at the beginning of the season there were fewer snowmobiles and snowcoaches, and I think it’s quite obvious that snowmobiles weren’t driving to Old Faithful before January 1 from the west entrance.

So data like this we can look at and try to figure out what’s happening.  One of the things is wind. You can imagine if it’s a windy day you might hear snowmobiles more or less. I just plotted here the percent time for each of those samples you could hear wind also. And I haven’t looked at wind speed but I have that data. It’d be interesting to look, but at this point it doesn’t look like the presence of wind is associated closely with our measures of percent time audible.

Now this is data from 2 winters ago, again.  Mary Mountain Trail, 1,000 feet from the road. And just again the same pattern, with average of 25 % for oversnow vehicles.  If you get farther off the road, a mile off the road, now at Lone Star Geyser, many days you couldn’t hear oversnow vehicles at all, but over the course of the winter the average was just about 4 % of the day, 8 am to 4 pm you could hear oversnow vehicles.

Now, there are two things, one is when you’re in the office listening to headphones you wonder how well that, how realistic that is, what does it represent. And so we’ve done quite a bit of testing to see what we’re hearing on our digital recordings, is that comparable to what you’d hear if you’re outside. To do that we do attended logging. We sit there and actually record for several hour periods, what we hear, and then we compare that to our analysis of digital recordings. And also  I’ve shown you lots of statistics for oversnow vehicles, but one of the things is we don’t know who’s driving these things, and that’s one thing the park is very interested in.  Are these park service snowmobiles, are they visitor snowmobiles, concessionaires, are they contactors. And so this winter we did a fairly big effort, and went out for hours at a time and just recorded.  We sat near a road and watched who was driving these things, and so we can come up with an index of the percent of the snowmobiles and snowcoaches that are park or administrative use, or visitor use.

Now I’m going to shift gears, and talk about the sound levels, the other part of this monitoring.  This is just a little primer on decibels if you’re not familiar with decibels. That’s how sound levels are measured.  Also I was talking about all these different 1/3 octave frequencies. I’m not going to talk about those much anymore. A measure of a overall sound level is just a one decibel number. And “A” is an A weighting, it’s more comparable to how we hear than the actual energy in these sound waves.

So anyway, I’ve a variety of things...I’m talking now probably at around 60 decibels. For those in the front of the room and the back of the room it’s quite a bit quieter than that. And that’s a really important thing when you’re looking at decibel levels is to know how far you are from the source, the measurement site to the source, because without that information the actual numbers, the decibel levels, are basically meaningless because of that attenuation, and I’ll show examples of that a little bit later.

Decibels are logarithmic, and that does some interesting things when you’re looking at them also. The difference of 10 decibel from 60 decibels to 70 we perceive as a doubling of sound level, and so 60 - 80, 80 decibels is 4 times as loud as 60. So it doesn’t take a lot of difference in decibels to make an appreciable difference in what we perceive as loudness. Now the scale is basically set because of human hearing. 0 decibels is about the threshold of human hearing. There can be negative decibels in a scale, if other animals are more sensitive then we are, and equipment is much more sensitive than our ears. At about 130 decibels it stops becoming sound and it becomes pain.

So I’ll just show you a couple of examples of some of the data that I collect. This was over Presidents Day Weekend at Madison Junction again. This is a 24 hour period, midnight to midnight, and then decibels from 0 – 80 decibels.  This is two days, Saturday and Sunday combined.

These red dots are just the maximum sound level at any one hour.  That’s a pretty straightforward concept. If you know anything about Yellowstone Park in winter you could predict that the sound levels, the maximum sound levels would be quietest early in the day until 7 or so and then they would jump up. In fact that’s what they do, and they stay throughout the day until about midnight that evening. But the nice thing about having digital recordings along with this is that we can actually go and listen to every one of these things, and determine what in fact this was...snowcoaches and snowmobiles, except for that one, which was a jet, and these two were groomers. And so it’s really quite a powerful way of describing soundscapes by a combination of sound levels and digital recordings.

This middle line indicates the Leq, that energy average. It’s just like many averages are affected by a few loud sounds, so it’s quite a bit higher than if you took a median.  If you compare that to a median, which is the next line down, you can see that the majority of sounds during this weekend, were quite, well, half the sounds were below this level, and half were above.

So you can look at these patterns and the bottom is not the minimum sound level, but the level exceeded 90% of the time, so the 10th percentile of sound levels if you stack them all up in any one hour from quiet to loud.  And when you don’t measure sound levels, if you don’t know what the natural sound level is in an area, the park service uses L90 as an analog to the natural sound level, and that’s just with the assumption that the loudest sounds would be human caused, and so the quietest sounds would be natural.

So you can look at these kinds of data over days and between or over different sites. Here I just have an illustration of comparing over Presidents Day Weekend sound levels at Old Faithful, the ranger station at that site, and Madison Junction. And here you have the minimum sound levels at these two sites, and the sound levels minimum at Old Faithful is louder and that’s just explained by there’s always utility sounds always occurring at Old Faithful, 24 hours a day you can hear a utility. The next is Lmax, and this you might think I wonder why it’s louder at Madison Junction. But the difference there is the monitor, the microphone is 100 feet from the road at Madison Junction, and 220 feet from the nearest road at Old Faithful. And so sound levels decrease by about 6 decibels every doubling of distance. So if you add 6 decibels to this the maximum sound levels at Old Faithful are fairly comparable, or maybe a little louder than at West. And so you can do that for a variety of other metrics, and start getting a pretty good idea of what’s happening soundwise.

Now here are some lame conclusions. I put this in kind of at the very end because I thought if I ended there people would say well, ya, so. After one year of analysis, 2 winters ago, it was very clear that the best available technology, using 4-stroke snowmobiles and using guides, substantially decreased the impact of non-natural sounds on the natural soundscape.  But in some areas the desired soundscape, natural soundscape hasn’t been obtained. That’s just saying that there’s some room for improvement. If you’re completely unsatisfied with these conclusions I have a report that is in detail that I’d be happy to share with you.  Thanks very much.  (Applause)  Do we have some...?

Female Voice:  Questions?

Shan Burson:  Yes...

Audience Member:  Did I recognize a Harley-Davidson in the...?

Shan Burson:  You did, yes.

Audience member:  What is its decibel rating of it compared to a snowmobile, say a 2 stroke snowmobile?

Shan Burson:  Well, that depends. I think a stock...I mean there’s so many differences between (yeah, you’re right) snowmobiles.  That Harley-Davidson measured at 100 feet was about 84 decibels.  That would indicate somewhere around 90 decibels at 50 feet, and a 2-stroke is somewhere in the 70s at 50 feet. But these are not... often Harley-Davison riders will put on after-market mufflers so they’re much louder.

Female Voice:  One more question right over here, and then we’ll move on to the next presentation.

Audience Member: You showed a slide of attended logging, and what were the results? What did you observe? What was causing (covered by cough)?

Shan Burson:  Well, that was just a month ago, basically.  I’m not sure exactly what you’re asking.  If you’re asking about what percentage of park snowmobiles are park staff I haven’t looked at it. We’re just entering that data, working with that now.