Determining CNOSSOS-EU meteorological correction factors in Ire- land Simon Shilton 1 Acustica Limited 3000 Aviator Way, Manchester, M22 5TG Joshua Nunn 2 Noise Consultants Limited 23 Coldharbour Road, Bristol BS6 7JT

ABSTRACT Ahead of the Round 4 strategic noise mapping under the END, TII commissioned a research project to determine the meteorological correction factors required for CNOSSOS-EU road and railway traffic noise calculations across Ireland. Methodologies for determining the percentage favourable propagation were identified under NMPB2008 and NORD2000, and the input data requirements of the methods assessed. Meteo data available from Met Eireann and TII weather stations was collected and collated and compared with the requirements of the two methodologies. The data available in from Irish met stations led to the selection of the methodology from NORD2000 being selected. Thirty years of hourly data was analysed for 26 counties to provide long term weather data for temperature, relative humidity, mean sea level pressure, and percentage of favourable propagation suitable for calculations under CNOSSOS-EU. This paper will present an overview of the methodology, available data and results obtained.

1. INTRODUCTION

Transport Infrastructure Ireland (TII) is a state agency in the Republic of Ireland. The primary func- tion of TII is to provide an integrated approach to the future development and operation of the national roads network and light rail infrastructure throughout Ireland.

Directive 2002/49/EC [1] (commonly referred to as the Environment Noise Directive ‘the END’) relates to the assessment and management of environmental noise. It is the main instrument of the European Union (EU) to quantify noise pollution levels, and trigger action within both Member States and at EU level. The END has the aim of establishing a common approach to the management of noise within the EU.

The Environmental Noise Regulations 2018 (ENR) [2] (as amended) [3] gives effect to the END in Irish law. Under the ENR, the National Roads Authority (now TII) are designated as the Noise Mapping Body (NMB) responsible for the development of strategic noise maps for all national roads carrying in excess of 3 million vehicles a year. The relevant local authority is responsible for the

1 simon.shilton@acustica.co.uk

2 joshuanunn@noiseconsultants.co.uk

worm 2022

development of strategic noise maps for non-national major roads. Under the ENR, TII are also the NMB for Luas light rail services in Dublin.

In accordance with the ENR, the first three rounds of strategic noise mapping (2007, 2012 and 2017) in Ireland were undertaken using an ‘interim’ method. The ‘interim’ method selected for these rounds was the Calculation of Road Traffic Noise (CRTN, 1988). However, under the END, a com- mon noise assessment methodology has been developed and is now required for the preparation of strategic noise maps going forward, commencing with the fourth round (2022).

Commission Directive (EU) 2015/996 [4] replaces Annex II of the END and describes the com- mon noise assessment methodology for the END. The Directive describes methodology for the cal- culation for noise from roads, railway, industry and aircraft. Directive 2015/996 has since been sub- sequently amended by a Corrigenda in January 2018 [5], and a Commission Delegated Directive in December 2020 [6]. This assessment method is referred to as CNOSSOS-EU:2020.

worm 2022

2. METEOROLOGICAL CORRECTIONS IN CNOSSOS-EU

The last three rounds of strategic road traffic noise mapping undertaken by TII has adopted the ‘in- terim’ assessment method, CRTN 1988 [7]. This method assumes downwind outdoor sound propa- gation in all directions. Under the methodology there is no specific requirement to account for mete- orological data and the associated effect on outdoor sound propagation.

The CNOSSOS-EU:2020 methodology requires the propagation between each source and receiver to be assessed twice: • Once with straight rays for homogeneous conditions, Figure 1; and • Once with curved rays for conditions favourable to propagation between the source and the re- ceiver, Figure 2.

5 =SO, + 0,0) + 0,03 +03R —SR

Figure 1: Example of calculation of the path difference in homogeneous conditions, in the case of multiple diffractions

Figure 2: Example of calculation of the path difference in favourable conditions, in the case of mul- tiple diffractions

The calculated long term average noise level at the receiver, L LT , to be reported for the purposes of the END is then the logarithmic average of the two calculated levels, L F and L H , according to the ratio determined by the occurrence of ‘favourable’ propagation. This is expressed as a percentage value, p , in 20° increments around a 360-degree meteorological rose, as shown in Equation 1.

(1)

The attenuation due to atmospheric absorption is governed by air temperature and humidity. This is calculated in line with ISO 9613-1:1996 [8], the attenuation due to atmospheric absorption A atm during propagation over a distance d is given in dB by Equation 2:

worm 2022

(2)

where: d is the direct 3D slant distance between the source and the receiver in m; α atm is the atmos- pheric attenuation coefficient in dB/km at the nominal centre frequency for each frequency band, in accordance with ISO 9613-1; and the values of the α atm coefficient are given for a temperature of 15 °C, a relative humidity of 70% and an atmospheric pressure of 101 325 Pa. They are calculated with the exact centre frequencies of the frequency band. These values comply with ISO 9613-1. Meteoro- logical average over the long term shall be used if meteorological data is available.

For the purposes of preparing strategic noise maps, the calculation of α atm requires the availability of long term annual average air temperature and relative humidity datasets.

3. CALCULATING FAVOURABLE PROPAGATION

CNOSSOS-EU:2020 does not include a description on how the occurrence of favourable propagation is to be determined. The methodology to calculate outdoor propagation within CNOSSOS-EU:2020 is based upon the French method described within NMPB2008 [9], which includes occurrence of favourable propagation for a selection of weather stations across France, and whilst supporting doc- uments are mentioned, it does not contain a description on how the occurrence of favourable propa- gation was determined for each of the weather stations.

3.1. Methodologies identified Based on a literature review, two methodologies were identified which would enable the calculation of sound speed gradient and the occurrence of favourable propagation in 20-degree increments around a 360-degree meteorological rose. These are published in the context of NMPB2008 by Sétra in France [9], and NORD2000 by Vtt Technical Research Centre of Finland [10].

The Sétra document describes a methodology to calculate outdoor propagation from which the CNOSSOS-EU:2020 point-to-point propagation methodology is based. Under this methodology, the temperature gradient and wind speed gradient are required at two heights, preferably from 2 metres and 10 metres positions above ground level. To facilitate this, the following meteorological data is

required: average wind speed m/s at 6m at 6 m above the ground; wind direction; cloud cover; air temperature; solar radiation (W/m 2 ); and terrain moistness.

The TRC Finland document describes a methodology which is also cited within the NMPB2008 method as a valid methodology for the calculation of the percentage occurrence of favourable prop- agation. This method requires the following meteorological data: average wind speed m/s at 10 m above the ground; wind direction; cloud cover; and air temperature.

3.2. Data available from Met Éireann and TII As two valid approaches to calculating the occurrence of favourable propagation have been identified, the decision as to which approach can be used, is somewhat determined by the parameters which are gathered in the available meteorological data.

Meteorological data in Ireland is primarily collected and managed by Met Éireann, who are Ire- land’s National Meteorological Service, for the purposes of weather forecasting and provision to rel- evant stakeholders. Hourly average data from Met Éireann has been made available to this project in a columnar timeseries format with embedded metadata, contained within a single file for each met site, spanning the period 1990 to 2020. Hourly data was also available from TII weather monitoring stations along the national roads network.

Following the review, it was identified that the following parameters were available within the metrological data: wind direction; wind speed; temperature; cloud cover; relative humidity; and mean sea level pressure.

The review also identified that most of the data required for both of the assessment methodologies, was available from most of the meteorological stations. However, some of the data required by the methodology described in the Sétra document for use with NMPB2008 is difficult to obtain from the meteorological stations. For example, the review confirmed that solar radiation and terrain moistness are rarely measured at meteorological stations in Ireland. Whilst terrain moistness can be calculated using a meteorological model, as described in (D. Courault et al.1996) [11] this is complex and con- sidered beyond the scope of this project. While it may be possible to estimate hourly solar irradiance by derivation from cloud cover, the data were found to be inconsistently captured and therefore any derived solar radiation data may not be fit for purpose.

worm 2022

Due to the unavailability in solar radiation data, and the complexity in calculating terrain moist- ness in place of unavailable measured data, both required for the Sétra methodology for NMPB2008, the NORD2000 methodology as described in the TRC Finland document has been identified as the most feasible and practical way to calculate the occurrence of favourable propagation in Ireland.

3.3. NORD2000 methodology Within the NORD2000 methodology, the effect of the meteorological conditions on sound speed gradient is calculated using Equation 3:

(3)

where: z height above ground surface; z 0 the roughness length; A coefficient of the logarithmic term (m/s); B coefficient of the linear term (1/s); and C(0) sound speed at height z = 0 m (m/s).

The A and B coefficients required for the calculation of sound speed gradient are themselves a function of the aerodynamic and thermal conditions which can be calculated as function of three other

parameters with wind direction: friction velocity (u*); Monin-Obukhov length (L); and temperature scale (T*).

These three parameters can be derived, providing three characteristics are available from annual average meteorological data, for local meteorological stations. These characteristics are: wind speed at 10 m above ground, V (z = 10 m); cloud cover in octas; and time of the day.

The three characteristics above can be used to derive two meteorological propagation classes (wind speed class and atmospheric stability class). These are reproduced in Table 1 and Table 2 below, respectively:

worm 2022

Table 1: Wind speed classification ((Raimo Eurasto, 2006) [10], Table 1, Page 7)

Table 2: Classification of atmospheric stability ((Raimo Eurasto, 2006), Table 2, Page 7)

‘wind speed ‘lass WI w2 w3 wa ws.

With the wind speed class and stability class, the friction velocity (u*), the Monin-Obukhov length (L) and the temperature scale (T*) parameters can then be derived using Tables 3, 4 and 5 from Raimo Eurasto, 2006.

(4)

time of day cloud cover ‘day 018 10 28 day 3/8 10 5/8 day 68 10 88 night 518 10 88 night 08 10 4/8

(5)

(6)

Once all the parameters are calculated, the A and B constants can then be calculated through the formulas reproduced in Equations 4, 5 and 6 above.

Where: u* is the friction velocity in m/s; T* is the temperature scale in K; L is the Monin-Obukhov length in m; C vk is the Von Karman constant = 0,4; g is the Newton's gravity acceleration = 9,81 m/s;

= daring day (stability cesses S,, S2 and $5)

C p is the specific heat capacity of air at constant pressure, 1005 J/kg K; T ref is the reference tempera- ture = 273 K; and α is the angle between wind direction and the direction of sound propagation.

With the A and B coefficients it is possible to generate 25 different sound propagation conditions based on the sound speed gradient. These 25 possible scenarios can be re-classified into average acoustic propagation classes using the method indicated by Birger Plovsing in a paper prepared for the Danish minister of environment [12]. In this paper, 25 classes are combined into the 4 propagation classes, namely: upward propagation; homogeneous; downward propagation; and very downward propagation.

4. APPLYING DEFINITIONS FOR FAVOURABLE PROPAGATION TO IRISH MET DATA

4.1 Extracting Meteorological Data from Irish Met Stations Historical hourly datasets are available directly from Met Éireann [13], and were obtained for all available years spanning 1990 to 2020. Furthermore, TII maintain a network of over 100 meteoro- logical stations across Ireland, as part of their national road network monitoring system. Measured data at a sub-hourly resolution have been made available via Vaisala for the purposes of this study, who operate the sites on behalf of TII.

Before this raw data could be used, it was subjected to a multistage processing procedure, and quality assurance (QA) process. Hourly average data from Met Éireann is available in a columnar timeseries format with embedded metadata, contained within a single file for each site, spanning the period 1990 to 2020. The datasets provided by TII are in a similar columnar timeseries format, sepa- rated by calendar year, and are generally measured at sub-hourly resolution. While the overall pro- cessing steps are broadly similar for the datasets from each provider, some differences in the datasets mean that bespoke pre-processing is required in each case.

Following the meteorological data processing workflow described above, resulted in a total of 133 meteorological stations in a consistent format suitable for subsequent use.

The original request from TII was to prepare meteo data for each of the eight Forecast Regions in Ireland. On the basis of having suitable meteorological data available for 133 stations distributed around the country, it was determined that it was possible to develop the outcomes of this study for each of the 26 Irish counties. The hourly data from these stations were aggregated into the counties within which they are situated. Table 3 provides an example of the long term averages for key mete- orological variables in each county, as derived from the post-processed hourly datasets. The process of county level aggregation has the effect of reducing the impact of any missing variables.

Table 3: Example of long term means for relevant meteorological variables in Ireland, by county

County Temperature

Temperature (K)

Wind speed (m/s)

Cloud cover (oktas)

Relative humidity (%)

Mean Sea Level Pressure (kPa)

(°C)

Carlow 9.9 283.0 3.4 3.5 83.0 101.4 Cavan 9.3 282.5 2.5 3.7 83.2 101.2 Clare 10.6 283.8 4.2 5.4 82.7 101.3 Cork 9.8 283.0 3.8 4.8 84.6 101.4

worm 2022

4.2 Applying the NORD2000 methodology Once the county-level meteorological data was extracted and processed, a computer program was written to implement the NORD2000 methodology, which has been based on the process described above. The meteo-classes were then re-classified as described above. The yearly average number of occurrences were summed across the seven classes identifying conditions favourable to propagation. These values were then divided by the total number of occurrences and presented in percentages, which are the percentages of favourable propagation (%Pf) required by the CNOSSOS-EU:2020 methodology.

The %Pf calculated for the average of meteorological stations in County Dublin are shown in Figure 4 below. The data are presented as a percentage value, in 20° increments around a 360-degree meteorological rose.

worm 2022

Figure 4: %Pf presented in 20-degree increments around a 360-degree meteorological rose of the

mean of Meteorological Stations in County Dublin, over the period 1990 to 2020

Similar results were prepared for all 26 counties in Ireland. All the results show how the evening and night-time period have generally higher probability of favourable propagation when compared to the day-time period. This phenomenon is due to the different meteorological conditions across the different period of the day, during the night as the temperature gradient tends to be positive (temper- ature increase with altitude), which improves the sound propagation.

Despite that the %Pf values are derived from many distinct meteorological variables, examination of the patterns and values shown in the %Pf meteorological roses highlights that the counties appear to fall into several categories. For example, the western counties of Donegal, Mayo, Galway, Clare, Kerry and Cork, which each have significant exposed coastal areas, show a similar pattern of low %Pf values, with relatively small variation between day, evening and night-time periods. In contrast, the vast majority inland regions exhibit a very different and more evenly distributed pattern, with much greater average %Pf values for the evening and night-time periods.

5. CONCLUSIONS

Meteorology data available from Met Éireann has been reviewed to identified approaches to devel- oping meteorological corrections in Ireland for use with CNOSSOS-EU.

Under CNOSSOS-EU, the percentage of favourable propagation in 20-degree increments is re- quired. To determine this, it was identified that given the data available from Met Éireann that the

‘%PF County Dublin, PF not (2300.07.00) —xeF Da¥(07:0025:00) ° 340 100% 2 x20 20% cox 200 260 240 —H0F eve 19:0023.00) 100 0

approach adopted for computing favourable propagation for use with NORD2000 was the most ap- propriate methodology.

Using the available meteorological data across Ireland, it has been possible to implement the NORD2000 methodology for the calculation of %Pf, at a county level. Several approximations have been used to address shortcomings in the provided meteorological datasets, which are considered appropriate and sufficiently robust, given the spatial and temporal resolution of the aggregated da- tasets.

Long term results for temperature ( o C), Relative humidity (%), mean sea pressure (kPa), and per- centage of favourable propagation in 20-degree increments for the day, evening and night periods (%), have been generated for all 26 administrative counties in Ireland to support Round 4 strategic noise mapping in Ireland using CNOSSOS-EU under the END.

6. ACKNOWLEDGEMENTS

We gratefully acknowledge the support of Transport Infrastructure Ireland for commissioning the research under the TII Open ResearchCall 2021 please see https://www.tii.ie/technical-services/re- search/ for further information on TII Research. 7. REFERENCES

1. Directive 2002/49/EC of the European Parliament and of the Council of 25 June 2002 relating

to the assessment and management of environmental noise, Official Journal of the European Communities, L 189/12-25, 18 th July 2002. 2. S.I. No. 549 of 2018 – European Communities (Environmental Noise) Regulations 2018. 3. S.I. No. 663/2021 - European Communities (Environmental Noise) (Amendment) Regulations

2021. 4. Commission Directive (EU) 2015/996, The European Commission, May 2015. 5. Corrigendum to Commission Directive (EU) 2015/996, OJ L168 of 1st July 2015, L5/35 to

L5/46. 6. Commission Delegated Directive (EU) 2021/1226 of 21.12.2020 amending, for the purpose of

adapting to scientific and technical progress, Annex II of Directive 2002/49/EC of the European Parliament and the Council as regards common noise assessment methods. 7. Department of Transport, Calculation of Road Traffic Noise (CRTN), HMSO, 1988. 8. ISO 9613-1, Acoustics – Attenuation of sound during propagation outdoors – Part 1 – Calcula-

tion of the absorption of sound by the atmosphere, International Organisation for Standardisa- tion, 1996. 9. Sétra, Road noise prediction - Noise propagation computation method including meteorological

effects (NMPB 2008), Sétra, June 2009. 10. Raimo Eurasto, NORD2000 for road traffic noise prediction: Weather classes and statistics, Vtt

Technical Research Centre of Finland, 2006. 11. Courault, D., Lagouarde, J.P., Aloui, B., Evaporation for maritime catchment combining a me-

teorological model with vegetation information and airborne surface temperatures, Agricultural and Forest Meteorology 82 (1996) 93- 117, 1996 12. B. Plovsing, Noise mapping by use of Nord2000 Reduction of number of meteo-classes from

nine to four, Danish Ministry of the Environment, working Report No. 18, 2007. 13. https://www.met.ie/climate/available-data/historical-data

worm 2022