Deep rumbling road noise is a true companion on my nature recording trips around in Denmark. Maybe it is due to modern headphones or to bass sensitive microphones – the rumbling is just ever present when I get home and hear the recordings. Noise from cross-cutting motorways can be particularly audible over long distances, however the noise level do vary surprisingly from time to time at the same location.
I have started to realize that weather conditions play a major role for this ambient background noise. The typically far distant nature of these sounds gives a high sensitivity to weather conditions. The atmosphere plays an active role in reflecting and guiding long distance sounds. It is actually a well-founded but highly complex research field. Sound differences of 10-20dB are often stated.
To my experience, road noise (closely followed by aircraft noise) is the most common and persistent type of noise encountered on my field recordings. Other man made noise types include trains, industrial activities, shooting ranges, lawn cutters, dogs and …wind turbines. I definitely don’t see this as all bad. The background ambience is very important for the identity of a place, often at a subconscious level. Imagine 100 years ago how completely different the ambience must have been, and again in 100 years time it will likely be completely different, e.g. with the emergence of electric cars and climate changes etc.
All in all, I have come to see that studying weather forecasts and road maps can be very instrumental when planning audio nature recordings. Considering weather factors is, in general, a great help to understand the often surprising outcomes of audio field recordings. In this blog I will try to dig more into this topic and compile various related information.
Noise from roads
The density of cars and their speed determine how much noise a road produces. The noise of every single car will add up. Here is an example of typical variation of traffic density through the day (figure, ref).
Saturday and Sundays have typically much less traffic, and especially lacking the morning and afternoon peaks. The speed is crucial for the amount of noise produced, which makes motorways particularly noisy. The below figure shows how the noise increases with the speed for different types of vehicles (figure, ref)
A wet asphalt surface adds an extra powerful hiss to the noise, while a snow covered surface can be very quiet, leaving the engine as the only sound.
A busy road can be seen as a line source (as can a sea shore) which means that the sound levels will decay over distance much slower compared to a point source. A typically stated figure is 3 dB per doubling of distance, which is highly theoretical. Furthermore, as the road curves and bends the image gets more complex, and it is important to note that the road also produces a lot of noise along the direction of the road.
As for most sound sources, a large part of the road noise is directed upwards into the air, and this underlines the importance of looking at how the sound is reflected and guided by the atmosphere.
The wind has the most direct and audible effect on sound propagation, especially for distant sounds. The wind direction is very important at the recording location. It is really possible to have much less noise from roads/motorways considerable when the road is downwind.
It is not the wind speed it self, but the wind speed gradient that bends and refracts the sound. The vertical wind speed gradient is the main factor. Figure after J. S. Lamancusa, Penn Stat.
Upwind, the sound is generally bend upwards into the sky and therefore rapidly decaying. Downwind, the sound is bend in the direction of the wind and downwards to the ground. The wind may not need to be very strong to have a clear effect. High winds may be more turbulent and unpredictable causing odd fluctuations for far distant sounds.
As a motorway road curves and bends strong noise tails can occur in certain regions especially when the wind goes along the direction of the motorway.
In windy conditions the majority of the ambient background noise will often appear to come from the direction of the wind, -like focussed into a certain direction. For other directions the noise is kind of cut away. Unfortunately, this also counts for the delicate nature sounds that you try to capture . In general nature recordings seldom benefit from windy conditions.
Another profound effect relates to how the temperature varies with the height (vertical temperate gradient). This effect may happen at calm evening or night when the sound transmission suddenly appears to be radically enhanced. The normal temperature gradient decreases with height, but if the temperature increases with height, ie. a ground level inversion, the atmosphere can act as a ceiling that reflects and guides the sound waves over long distances along the ground. (figures, Ref)
Above shows the normal condition with decreasing temperatures with the height. Here the sound tends to bend into the air, enhancing the decay over distance. This is especially pronounced for hot sunny days and afternoons with high incoming solar warming to the surface. This is also a typically situation with unstable air, maybe after a cold front with convection and cumulus clouds and risk of showers. I believe this is also a classic condition at a hot sunny beach.
A ground level inversion is apparently typical when the ground cools rapidly, such as for clear evenings and nights with high out going radiation. Typically at winter and spring with low solar heating. It could be calm high pressure situations at winter with foggy conditions or be in connection with cold water and sea breezes.
A good resource to evaluate the weather with respect to inversions is to use air gliders’ weather services and forecasts, such as rasp.skyltdirect.se for Scandinavia. This one also provides vertical atmospheric profiles, with wind, temperature and humidity.
Temperature and Humidity
The direct attenuation of sound (absorption of sound in air) due to humidity and temperature is said to be more negligible (ref), but still important over longer distances. The attenuation increases sharply to the frequency and approaches zero for low frequencies, which explains why especially very deep rumble can travel over long distances. Typical values can be see below (according to wikibook data).
The trend goes for a lower attenuation for higher humidity and for higher temperatures. Comparing a humid summer day, 20C and 90RH% to a dry winter day with 0C and 50RH% gives a difference (per 100m) of 1.5dB@2000Hz, 5bB@4000Hz and 8.4dB@8000Hz, which actually seems quite significant for far distance field recordings.
Precipitation, rain, snow, or fog, has an insignificant effect on sound levels (ref).
The surface condition of the road being dry, wet, snow covered, or somewhere in between has a high impact on the sound produced. A wet road surface is reported to produce 4-7 dB(A) more noise compared to a dry road. The main difference is an increase of the high frequency hiss for frequencies above 2 kHz, while the lower frequencies are less affected. A sudden snowfall with snow covered roads can change the soundscape radically. It puts a natural limitation on the speed, and reduces/changes the noise from tires to a minimum.
Another way weather may affect noise levels is through changes of the surface conditions over which the sound is propagating. A dry, wet, frozen or snow covered etc ground can have different absorption characteristics.I have not been able to find much information on directly how big a role this has. Ground absorption is again mostly affecting higher frequencies, and is normally said to be stronger than air absorption. A hard and flat surface attenuates less, while a soft, porous surface attenuates more. A newly snow covered ground can be highly sound absorbing, grass and vegetation is also quite absorbing, while hard asphalt on the other hand has very little absorption (Ref). I imagine that dew wet grass and vegetation surface is more absorbing than dry or frozen vegetation. Dry soil summer soil more absorbing compared to fully wet winter/spring soil.
Test recordings of motorway noise and weather
I have made some simple recordings of the nearby motorway to evaluate the variation due to weather. The Motorway begins about 300 meters away behind some building. All recordings are made around 8-9 pm in the evening, which should give relatively constant amount of traffic.
Downwind of the motorway, 5m/s, 5C, 85%RH, dry asphalt
Side wind to the motorway, light wind 2 m/s, 3C, 85%RH, dry asphalt
Upwind of the motorway, very light wind, 1m/s, 6C, 90%RH. Clear sky, dropping temperatures.
Cold & dry downwind of the motorway, light wind 3 m/s, -2C, 85%RH, dry asphalt
Rain, downwind of the motorway, light wind 2m/s, 2.8C
From the clips, I think the conclusion is that wind direction relative to the course of the road has a very strong effect and is the all dominating factor. The other weather factors only add smaller variations on top. Noise from aircraft is also present on many of the clips. Prevailing flight routes can also change according to the weather conditions.