Aviation and the Environment: A National Vision Statement,
Framework for Goals and Recommended Actions
Id like to start with a few quotes from a report FAA recently
released outlining a vision statement for aviation and the
environment.
I suspect that everyone in this room agrees that DNL 65 dB does
not adequately address the noise problem at airports. If it did,
after achieving 95% reduction in people exposed, wed all have
claimed victory and retired I remember being concerned in 1992 at
the beginning of the Stage 2 phase out all kinds of folks told us
wed be out of business by 1999. Well, our presence here more than
15 years later is probably proof enough that the phase out alone
despite its huge cost was not enough.
On the other hand, no one ever died from environmental noise
unlike air toxins or ground water pollution, aircraft noise does
not pose life-threatening effects, and its a real stretch to even
try to imagine it. We have a joke around the office that theres no
such thing as a noise emergency. This makes our job both easier and
more difficult at the same time: easier, because the price for
failing to adequately describe and address noise is potentially
less of a risk than, say, failing to learn that some air pollutant
really is toxic. It is more difficult though, for the same reasons:
perhaps because no one dies, many folks dont see noise as an
imperative. It also becomes a quality of life issue that is
difficult to quantify, even harder to monetize what should a 1.5 dB
increase in noise cost an airline, if the primary effects are a few
more people annoyed or occasionally waking someone up?
What Id like to do today is to propose a notion that contemplate
communicating about noise using something other than decibels, and
that the metrics we use to describe noise bear really do say
something about the way people live. We have identified several of
the main complaints about noise the concerns people share when they
show up at public meetings to protest runway expansion and tried to
use available noise metrics and modeling techniques to show how
where effects might be observed.
Why present information this way?It allows us to get away from
talking about decibels and start talking about experienceIt
validates peoples real exposure lets face it, no one has ever heard
DNL 65!It provides a more transparent methodology for policy-makers
to consider other metrics and criteria
I am not proposing any criteria here I know that many of you
would like someone preferably FICAN or FAA to say, the threshold
for sleep disturbance shall be x. I dont pretend to know what a
reasonable level of annoyance or sleep disruption is. My goal in
this presentation is to start a dialog about how we might
communicate aviations effects more meaningfully to the people who
live around airports, and provide more meaningful information to
those who are in a position of making policy decision.All of the
analyses I am presenting are based on samples of data from BWI. I
chose BWI for a number of reasons:It has a very long history of
noise abatement you many not know it, but Part 150 was based on the
Maryland Code that was enacted in 1974. In fact, one of the reports
prepared for Maryland in the 1970s recommended that DNL 60 be
adopted as a land use compatibility criteria when it became
economically feasible, or when noise levels dropped by 10 dB. COMAR
requires that MAA prepare an Airport Noise Zone every 5 years, that
is the outer limit of the current, 5-year, and 10-year forecasts.
This is a more comprehensive planning horizon than Part 150 one
could argue it is better suited and more compatible with master
planning concepts.One very interesting provision of COMAR is that
land uses within the ANZs DNL 65 contour are deemed incompatible
with residential land use development; developers are generally
prohibited from building residential, and need to apply to the
Board of Airport Zoning Appeals (BAZA) in order to do so. This is
occasionally granted, but requires developers to include sound
insulation and obtain easements. It is one of the few places I can
point to where a Balanced Approach has been taken seriously, and
local land use jurisdictions have lived up to their end of the
bargain.HMMH has a long history of analysis for this airport
(pre-dates HMMH, in fact some of the founders were involved all the
way back to the 1970s). We have access to lots of data, and have
recently started preparing daily noise contours we recently
submitted the first Noise Exposure Maps in the country with 365
daily contours providing the basis of noise exposure.
So, in a word, BWI should be a test case for everything going
right, and huge success. And in many ways, it is. This figure shows
the 1988 DNL Contours we prepared for BWIs first official Part 150
Submission.
Just for orientation, the primary mode of operations is landings
to runway 28 and 33L, and takeoffs from runway 28. Note though that
there were left-hand turns from occasional departures from runway
15R. Residential land use is shaded in yellow. Schools are
indicated by this little clue flag.
2003, by contrast, is much smaller. Primary modes of operations
are basically the same, though turns from Runway 15R are to the
right (not only for noise, but also for more efficient airspace).
Also, Runway 15R/33L was extended in the 1980s and now accommodates
general aviation jets this area to the north of Runway 33R is the
only area where the contour has grown in the last 15
years.Observations:MAA has been reasonably successful at preventing
development inside DNL 65 contours.Encroachment occurring now on
land that was formerly inside DNL 65 MAA cant do anything about
it.Lets face it. Weve been talking about the adequacy of DNL and/or
DNL 65 for a decade. Id rather not spend our time here today doing
that. Weve been talking about supplemental metrics now for several
years. Implicit in many of those discussions has been an assumption
that supplemental metrics are better understood by the community.
Why? Perhaps because they relate better to peoples experience. What
Im suggesting is that rather than be implicit about what those
effects are, we be very explicit in relating the noise metrics to
the potential effects of noise.
Instead, lets talk about effects of noise, and use the tools
already available to us to describe some of those effects. Perhaps
then, we can have a more informed discussion about noise effects of
aviation.
Again, weve identified the major effects of noise, and what we
believe is a reasonable way to describe them with tools at hand;
i.e., using a metric that fits the type of exposure.Ill describe
each of these effect contours, and provide an example. Youll notice
that several of the contours bear some relationship to traditional
DNL contours, while others are much larger. Land use compatibility
DNL: based on CNEL (based on EU work on transportation annoyance)%
Awakening (based on FICAN sleep curve, joint probability)N70 (based
on SEL contour)8-hour Leq (based on ANSI standard for interior
noise level of 45 dB)C-weighted Lmax contour (based on Indoor home
vibration)We have the ability at BWI to model INM using every
flight track that arrives or departs BWI. The figure shown here is
for one day. As you can see, the airport was in its normal
operation mode arrivals to Runway 33L, departures from Runway
28.This figure presents DNL contours for that days operations. As
compared with the 2003 annual contours, they are longer on the
predominant departure end, longer and skinnier on the predominant
arrival end. Also, there was a fair bit more use of Runway 33R than
normal for GA jets. But overall, a fairly typical day.Now lets look
at some other effects. First, annoyance most people agree that the
Schultz curve, shown here, is the basis for the decision that DNL
65 dB is the threshold of land use compatibility. In fact, as you
know, the EPA Levels document of 1974 recommended 60 dB as the
threshold of land use compatibility. The commonly used criteria of
65 was adopted primarily for financial reasons it barely possible
to imagine removing everyone from the DNL 65 contour, let alone DNL
60. In any case, here is the Schultz curve. You can see that at DNL
65 dB, the curve predicts that approximately 12-13.59% of people
will be Highly Annoyed Ive heard many elected officials say that
that level of annoyance becomes actionable. These curves are from a
recent study of annoyance conducted in Europe, which set out to
identify annoyance levels for different transportation modes. These
dose-response scales are quite similar to the Schultz curve, but
using Lden instead of Ldn (Lden is equivalent to CNEL)..Im afraid
that aircraft noise is the biggest loser here.Observations:Aircraft
noise more annoying than road, and rail probably not due to the
single event nature, since rail has single events that are
comparable to aviation.Curve for aviation higher than previous
dose-response curves which looked atSo, we can pick of levels of
annoyance, and present DNL or Lden contours in those termsGetting
back to my concept of showing effects, instead of presenting DNL or
Lden contours, what if we show annoyance? Where do we think 25% of
people will be highly annoyed? 30%Thus, 10% Highly annoyed = CNEL
5520% Highly annoyed = CNEL 6230% Highly annoyed = CNEL 67We dont
have to explain what DNL or CNEL means; dont even need to show
decibels!Turning to sleep this is a bit more complicated, but hang
in there its worth it!
The most current federal guidance with respect to sleep
disturbance is from FICAN in 1997. This dose-response curve shows
that maximum likely percent of people who will be awakened by a
single event for a given indoor Sound Exposure Level, or SEL
(recall, SEL takes into account both maximum level and
duration).WhereDose = Aircraft indoor SELResponse = Probability of
awakeningFICAN curve applies to Noise dose of a single aircraft
flyoverChance that the noise dose will awaken an average person
Use of FICAN curve is problematic.Past analyses have Selected a
low percent awakening (5-10%)Used the upper bound of the curve to
identify an indoor SEL that would produce the same percentage
awakeningConverted the indoor SEL to an outdoor SELUsed outdoor SEL
contours to tabulate the number of awakenings due to a single
eventAnd similar analysis for person-to-person differences, which I
wont belabor here, but if youre interested, I can send you a
technical paper on the concept. What we end up with then, is a set
of awakening contours that predicts the likely awakenings for a
given time period here a single night, but we could do average
annual, or seasonal, or whatever. This is also adjusted for
observed variability in behavior, not just an average person.I am
starting with the most engineering-oriented type of noise effect,
which is probably the easiest and least controversial to measure:
speech interference.
One of the primary effects of aircraft noise is its tendency to
drown out or "mask" speech, making it difficult or impossible to
carry on a normal conversation without interruption. The sound
level of speech decreases as distance between a talker and listener
increases. As the level of speech decreases in the presence of
background noise, it becomes harder and harder to hear. The figure
below presents typical distances between talker and listener for
satisfactory outdoor conversations in the presence of different
steady A-weighted background noise levels for three degrees of
vocal effort: raised, normal, and relaxed. As the background level
increases, the talker must raise his/her voice, or the individuals
must get closer together to continue their conversation. As
indicated in the figure, satisfactory conversation does not always
require hearing every word; 95% intelligibility is acceptable for
many conversations. This is because a few unheard words can be
inferred when they occur in a familiar context. However, in relaxed
conversation, we have higher expectations of hearing speech and
require complete 100% intelligibility. Any combination of
talker-listener distances and background noise that falls below the
bottom line in the figure represents an ideal environment for
outdoor speech communication and is considered necessary for
acceptable indoor conversation as well.One implication of the
relationships in the figure is that for typical communication
distances of three or four feet (one to one and one-half meters),
acceptable outdoor conversations where 95% intelligibility is
acceptable can be carried on in a normal voice as long as the
background noise outdoors is less than about 65 dB(A). If 100%
intelligibility is desired, the interior background level must be
less than about 45 dB(A). If the noise exceeds either of these
levels, as might occur when an aircraft passes overhead,
intelligibility is lost unless vocal effort is increased or
communication distance decreased. I am starting with the most
engineering-oriented type of noise effect, which is probably the
easiest and least controversial to measure: speech
interference.
One of the primary effects of aircraft noise is its tendency to
drown out or "mask" speech, making it difficult or impossible to
carry on a normal conversation without interruption. The sound
level of speech decreases as distance between a talker and listener
increases. As the level of speech decreases in the presence of
background noise, it becomes harder and harder to hear. The figure
below presents typical distances between talker and listener for
satisfactory outdoor conversations in the presence of different
steady A-weighted background noise levels for three degrees of
vocal effort: raised, normal, and relaxed. As the background level
increases, the talker must raise his/her voice, or the individuals
must get closer together to continue their conversation. As
indicated in the figure, satisfactory conversation does not always
require hearing every word; 95% intelligibility is acceptable for
many conversations. This is because a few unheard words can be
inferred when they occur in a familiar context. However, in relaxed
conversation, we have higher expectations of hearing speech and
require complete 100% intelligibility. Any combination of
talker-listener distances and background noise that falls below the
bottom line in the figure represents an ideal environment for
outdoor speech communication and is considered necessary for
acceptable indoor conversation as well.One implication of the
relationships in the figure is that for typical communication
distances of three or four feet (one to one and one-half meters),
acceptable outdoor conversations where 95% intelligibility is
acceptable can be carried on in a normal voice as long as the
background noise outdoors is less than about 65 dB(A). If 100%
intelligibility is desired, the interior background level must be
less than about 45 dB(A). If the noise exceeds either of these
levels, as might occur when an aircraft passes overhead,
intelligibility is lost unless vocal effort is increased or
communication distance decreased. Much research has been done on
the effects of noise on childrens learning. Two recent findings of
interest are the following:Researchers in the EU developed a
dose-response relationship between aircraft noise and learning
impacts, and showed a 5 dB increase in noise equated to a 2 month
delay in reading scoresWork we conducted for FICAN demonstrated
that low-performing students test scores improved with the addition
of sound insulation to school classroomsThe ANSI has developed a
standard for interior classroom noise of continuous noise of 40
dBA. Computation of this is pretty straightforward: compute an
8-hour Leq (8 am to 4 pm) we dont care about night noise in schools
and adjust for outdoor-indoor noise reduction (15 dB). Get outdoor
Leq contours of 55 dBA.Here, we could postulate that a learning
environment that meets ANSI standards is probably adequate for the
purposes of noise note that this bears no relationship to the
current use of DNL 65 dB for school funding.Finally, low frequency
has been an on-again off-again hot button issue around airports.
Its really only an issue near ground operations close to the
airport. Since the usual A-weighted decibels that are used for
other environmental analyses discount low frequency noise, we
recommend the use of a C-weighted contour.Our work for FAA
(actually, at BWI) in the late 90s showed that:C-Weighted Better
Correlated with Induced Vibrations and Resident Ratings than
A-WeightedPerceptible Wall Vibrations Likely to Occur for
C-Weighted Lmax Exceeding 75 - 80 dB.Propagation of C-Weighted Lmax
Close to Spherical SpreadingC-Weighted Lmax Possible Predictor of
Subjective Judgements of Takeoff Noise
Here, then, are C-weighted Lmax contours for the same days
activity at BWI. Again, operations that exceed this criteria would
be more likely to have perceptible vibration indoors.This figure
presents DNL contours for that days operations. As compared with
the 2003 annual contours, they are longer on the predominant
departure end, longer and skinnier on the predominant arrival end.
Also, there was a fair bit more use of Runway 33R than normal for
GA jets. But overall, a fairly typical day.Aviation and the
Environment: A National Vision Statement, Framework for Goals and
Recommended Actions
Id like to start with a few quotes from a report FAA recently
released outlining a vision statement for aviation and the
environment.
I suspect that everyone in this room agrees that DNL 65 dB does
not adequately address the noise problem at airports. If it did,
after achieving 95% reduction in people exposed, wed all have
claimed victory and retired I remember being concerned in 1992 at
the beginning of the Stage 2 phase out all kinds of folks told us
wed be out of business by 1999. Well, our presence here more than
15 years later is probably proof enough that the phase out alone
despite its huge cost was not enough.
![arXiv:1511.01631v1 [cs.CV] 5 Nov 2015 · Manjunath Narayana narayana@cs.umass.edu Allen Hanson hanson@cs.umass.edu University of Massachusetts, Amherst Erik Learned-Miller elm@cs.umass.edu](https://img.dokumen.tips/doc/110x75/5fbd15add9b35115b4645db6/arxiv151101631v1-cscv-5-nov-2015-manjunath-narayana-narayanacsumassedu-allen.jpg)
