Research On The Characteristics And Evaluation Of Soundscape In Suburban Forests

With the rapid development of China’s economy and society,there are the accelerating frequency of urban renewal and the increasing density of cities,which not only promotes the process of urbanization,but also leads to a series of living environment problems and a negative impact on the health and daily life of citizens.Based on the Ecological Civilization,we need to focus on the elements of physical environment such as sound,light,heat and so on in the urban areas for exploring a way to develop the high quality city with ecological priority and green development.The soundscape is related to the perception of human beings,focusing on the relationship between psychology and physics in the human settlement,and the requirements of environmental construction,perceived landscape and environmental quality.In addition,the inherent nature of suburban forests is beneficial to producing the natural soundscape,which contributes to produce suitable soundscape to citizens and keep them healthy.Thus,I systematically conducted this thesis,including experiments,analyses,discussions and conclusions,on the characteristics and evaluation of soundscape in suburban forests.I conducted a study on the differences of soundscape between broad-leaved and coniferous forests in suburban areas.Results suggest that:1)when L_Aeqis higher than 44.2 d BA and 47.9 d BA in the broad-leaved forest,respectively,the soundscape harmonious degree of geophony and biophony is close to the maximum.;2)when L_Aeqis higher than 40.2 d BA and 39.4 d BA in the coniferous forest,respectively,the soundscape harmonious degree of geophony is close to slightly harmonious,and that of biophony gets the maximum;3)there are three stages of soundscape harmonious based on HCA,including a reduction difference,a stationary difference and an increasing difference;and 4)Stevens’law and Weber-Fechner’s law of psychophysics can be applied to the soundscape data in suburban forests.Based on the differences of soundscape in suburban forests,I conducted a study on the perceived occurrences and uncertainty of soundscape in suburban forests.Results indicate that:1)the sum of WPOS of geophony and biophony exceed 85%in suburban forests;2)the highest WPOS is birdsong which gets 58.6%and 53.6%of the biophony in broad-leaved and coniferous forests,respectively;3)the WPOS of leaves rustling gets 45.6%and 30.6%of the geophony,and the WPOS of branches swaying gets 26.4%and 39.7%in broad-leaved and coniferous forests,respectively;4)there is a positive correlation between the WPOS of geophony and biophony,and perceived pleasantness of soundscape;and 5)the perceived occurrences of soundscape is beneficial to connecting the elements of soundscape in evaluation models,which promotes to reflect and simulate humans’perception.To explore the drivers of soundscape cognition,I conducted a study on the cognitive persistence of soundscape in suburban forests.Results show that:1)in the 6th second after the end of the perceived soundscape samples,the ACS of loudness with 10s soundscape samples are greater than 0.8,but the ACS of harmony and direction with 30s soundscape samples are below 0.5;2)for the 18th second and 30th second conditions,the ACS of all samples excluding perceived loudness with 10s soundscape samples fall below 0.2;3)the optional cognitive persistence of soundscape is 20s to maintain the cognitive soundscape in suburban forests;4)loudness,harmonious and direction cognition in suburban forests explains 36.61%,24.75%and 16.55%of the variance respectively based on PCA.These are determined by soundscape information,space information,and both soundscape and lightscape information,respectively;5)high attenuation intervals of soundscape and lightscape are difficult to maintain the cognitive soundscape based on HCA and correlation analysis;and 6)BP Neural Network contribute a model to predict the cognitive attenuation of soundscape in suburban forests,which could be optimized by pre-classification of PCA.Based on the cognitive persistence of soundscape in suburban forests,I conducted a study on the applicability of artificial neural network with temporal memory to soundscape in suburban forests.Results suggest that:1)the prediction accuracy of perceived loudness and harmoniousness is78.6%and 61.3%respectively by NARX Neural Network,and 81.9%and 55.1%respectively by Elman Neural Network;and 2)NARX Neural Network contribute a better mean of prediction accuracy for soundscape in suburban forests.Overall,these findings reveal the characteristics of soundscape in suburban forests and supplement the theoretical basis of the field of soundscape knowledge,which contributes references and guidances for researches and constructions of suburban forests and human settlement.