A noise monitoring campaign has been performed in an urban park of Milan (Italy) called Parco Nord. The area of study is a large peri-urban park in the northern part of the city, characterized by wooded land rich in biodiversity and exposed to different sources and degrees of anthropogenic disturbances, such as road traffic noise and artificial light.
The acoustic environment is rather complex due to the contemporary presence of different noise sources, leading to the difficult task of discriminating them in audio data. Due to these multifactorial characteristics, we evaluated different eco-acoustic indices in the attempt to derive a methodology to evaluate the potential of sound ecology indicators to discriminate the different types of sounds present in medium-large urban parks. Time series of about two-week recordings have been transformed into eco-acoustics indices and statistically analysed. The results show a redistribution of recordings into each cluster associated with different sound components and different period of the day. This allowed the identification of different degree of biophonic and/or anthropogenic activities throughout the day.