Distribution of noise exposure levels and annoyance scores
Low transportation noise exposure vs MHE2023 annoyance data
Aim Of this document
To evaluate the relevance of the noise map from the report Ljudmiljö i naturområden (https://www.naturvardsverket.se/om-oss/aktuellt/nyheter-och-pressmeddelanden/2024/september/75-procent-av-sveriges-naturomraden-har-bra-ljudmiljo/) (LiN method) and check its applicability vs Environmental Health Survey data 2023 (MHE 2023). Also at glance it checked howe inline is LiN method with SCAPIS noise by comparing MHE23 distribution of noise exposure levels in Stockholm County.
Exploring
Now I apply the same categorization to SCAPIS noise data i.e. all level above 47 dB Leq are coded as 47 
Generally distributions in Stockholm County look very similar with slight higher levels LiN method which is expected as it is based on more noise sources.
Survey data
In MHE 2023 survey we have a battery of questions about noise annoyacne by different sources (road, train, aircraft, wind turbine, industry, any noise). The annoyance is measured on a 5-point scale from “Very much” to “Not at all”. On example of annoyance data from road traffic noise we can see the distribution of annoyance answers both Figure 2. NB! in contrast with charts above all figures below are based on the whole MHE23 material (i.e. inlucding those outside Stockholm County).
Notably, not so many MHE23 participants are exposed to levels below 45 dB Laeq.
We focus on on those who are highly annoyed by road traffic noise (“Väldigt mycket”,“Mycket”). Then we group all of the answers into two categories: “Annoyed” (1) and “Not Annoyed” (0) and calculated proportion of those highly annoyed by different sources of noise (road, train, aircraft, wind turbine, industry, any noise) within different noise exposure bins.
Generally, the proportion of highly annoyed people increases with increasing noise exposure levels. It was more prompt for road traffic annoyance (<1% to aprox 5% in exposure range from 20 to 45 dB). Some tendencies for aircraft noise and railway noise annoyance. These trends are easier to see in the “smoothed” 5 dB band plot.
Modelled annoyance probabilities
With restricted cubic splines with 5 knots, we modeled the annoyance probabilities for different noise sources. The modeled probabilities are shown in Figure 4. Generally it shows the same trends with higher probabilities for road traffic, aircraft and railway over the exposure range but wind turbine noise. With of confidence intervals vary differently for different sources.
