*5.3. Impact on the LA*<sup>300</sup>*<sup>s</sup> of the ANE Subcategories*

In this paper, the impact on the computation of *LA*<sup>300</sup>*<sup>s</sup>* has been evaluated for every individual ANE every 5 min. The analysis presents interesting results to be discussed considering the SNR and duration of individual ANEs. The results in Section 4 show the existence of several ANEs with high impact, which present high SNR and long duration. After the individual analysis of the impact of ANEs on the *LA*<sup>300</sup>*<sup>s</sup>* computation, future work will also take into account the fact that dynamic acoustic mapping in real-life conditions face a more complex operating scenario. In a real-operation scenario, several ANEs can occur in a predefined integration time, so the ANE impact must be evaluated in an aggregated way for each period.

Another relevant result of the individual ANE analysis is the presence of ANEs with negative SNR. As detailed in Section 3.3.1, the SNR is evaluated taking into account each ANE in relation to its surrounding RTN signal level. In certain cases, the RTN decreases as the ANE occurs, so a negative SNR is obtained. Therefore, only those events with positive SNR should be removed from the *LAeq* computation, as also concluded in [52].

From this work, it can be concluded that working with data recorded in a real operating scenario is crucial to obtain a reliable modelling of the nodes' acoustic environment, according to the differences observed between the expert- and WASN-based datasets. Moreover, the analysis of the WASN-based collected data again shows the important role played by SNR and duration of individual ANEs in their impact on the *LAeq* computation, obtaining ANEs that should be considered for their high impact on the equivalent level.

**Author Contributions:** R.M.A.-P. has written part of the paper and has participated in the recording and labeling of the dataset. F.O. implemented the SNR and impact calculations, conducted the part of the dataset analysis and participated in the audio labeling and writing of the paper. F.A. has written the related work and reviewed the entire paper, with a special support on the contributions. J.C.S. has worked in the technical part, the recording, the labeling, and the analysis of the dataset.

**Funding:** The research presented in this work has been partially supported by the LIFE DYNAMAP project (LIFE13 ENV/IT/001254). Francesc Alías thanks the Obra Social La Caixa for grant ref. 2018-URL-IR2nQ-029. Rosa Ma Alsina-Pagès thanks the Obra Social La Caixa for grant ref. 2018-URL-IR2nQ-038. Ferran Orga thanks the support of the European Social Fund and the Secretaria d'Universitats i Recerca del Departament d'Economia i Coneixement of the Catalan Government for the pre-doctoral FI grant No. 2019\_FI\_B2\_00168.

**Acknowledgments:** The authors would like to thank Marc Hermosilla, Ester Vidaña, Alejandro González and Sergi Barqué for helping in the audio labeling. The authors would like to thank ANAS S.p.A. for the picture of the sensor installed in the portal and Bluewave for the recording of the audio files in the sensors of the WASN.

**Conflicts of Interest:** The authors declare no conflict of interest.
