City Evaluation Of Ecological Livability Based On Urban Sensing Multi-source Data

With the rapid development of new industrialization and urbanization,issues such as environmental pollution,dried up resources,and poor living environment have become hot spots for urban residents.More and more people paid attention to an important indicator of the urban ecological environment,the ecological livability of cities.However,there are many factors influencing urban ecological livability,and there are various evaluation methods and systems,each of which has its own advantages and disadvantages.There is still no complete and unified standard system to evaluate the ecological environment of the city.Due to the rapid development of Internet technology,it is no longer difficult to obtain data resources.Big data technology has been widely used in the research and analysis of urban problems,and has achieved good results.Focusing on the integration of spatiotemporal multi-source big data as the data source,this paper uses Shenzhen as the study area and blocks as the basic scale of urban research.Based on the LDA model,high-level semantic features were extracted,and the random forest model was used to extract the real Shenzhen land use results.These results were further used as housing units for subsequent evaluations.Subsequently,for the first time,using the extracted residential areas and urban villages as the evaluation targets,an urban ecological environment livability assessment index system was established,and the Shenzhen ecological livability index was designed and calculated.Based on multi-source big data and environmental factors that are closely related to people,five factors were selected: residential area density,thermal environment,air quality,population density,and urban green space.The TOPSIS model based on entropy weight method was used to fuse multi-dimensional indicators,and finally the comprehensive evaluating of the ecological livability of residential areas in Shenzhen was obtained.This research shows:(1)Multi-source big data has an incomparable advantage in traditional field investigations in the analysis of urban problems.In the extraction of Shenzhen’s urban functional area results,the mapping accuracy obtained by combining the LDA model and the random forest model can reach 82.1%.In the evaluation of urban ecological livability,various living environment indicators were calculated and generated from the data,so that the results was more objectively and truly reflect the ecological livability of the city.(2)According to the classification and prediction results of urban functional zones,the main residential areas of Shenzhen are concentrated in the south,the urban villages and industrial areas are mostly concentrated in the north,and the government and public services occupy little land,mostly airports,terminals,customs or railway stations and other places,while judging from the results,the overall urban green space coverage is good;the final ecological livability assessment results show that most of the top scores for urban ecological livability are residential neighborhoods near Shenzhen Park;the lower rankings are mainly urban neighborhoods with poor living environment conditions.(3)The Shenzhen functional zoning results extracted in this study can objectively reflect the real land use in Shenzhen and the set of ecological livability evaluation index system can reflect the comprehensive ecological livability of various residential areas in Shenzhen to a certain extent.The analysis of urban functional land use,and the pros and cons of residential areas in terms of ecological environment,climate factors,and other aspects can provide reliable theoretical basis and data results for government departments and urban planners.