東華大學圖書館 |

Dynamic Exposure, Inequality and Urbanization Effects: A Multidimensional Evaluation of Urban Greenspace in China.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Dynamic Exposure, Inequality and Urbanization Effects: A Multidimensional Evaluation of Urban Greenspace in China./
作者:	Song, Yimeng.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2019,
面頁冊數:	200 p.
附註:	Source: Dissertations Abstracts International, Volume: 81-10, Section: B.
Contained By:	Dissertations Abstracts International81-10B.
標題:	Geographic information science. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=27784075
ISBN:	9781392897744

Dynamic Exposure, Inequality and Urbanization Effects: A Multidimensional Evaluation of Urban Greenspace in China.
Song, Yimeng.

Dynamic Exposure, Inequality and Urbanization Effects: A Multidimensional Evaluation of Urban Greenspace in China. - Ann Arbor : ProQuest Dissertations & Theses, 2019 - 200 p.

Source: Dissertations Abstracts International, Volume: 81-10, Section: B.

Thesis (Ph.D.)--The Chinese University of Hong Kong (Hong Kong), 2019.

As one of the most critical factors in the urban environment, urban greenspace plays a key role in mitigating a series of urban problems. Beyond that, serving as the main place for residential activities and social interactions, urban greenspace has been proven to improve people's physical and mental health. Over the past decades, however, rapid urbanization has greatly reshaped the land covers in and around urban areas, dramatically changing urban greenspace in both quality and quantity. As a result, the opportunity to enjoy urban greenery differs considerably for residents who live in different regions. Therefore, a scientific and thorough understanding of different urban green environments within cities and how people interact with greenspaces is urgently needed. However, most existing urban greenspace assessment methods do not incorporate human-related information-such as human distribution, human mobility, and so on-which may increase uncertainty of assessment results and lead to imprecise assessment. To address these issues, this dissertation provides an improved three-stage assessment framework based on multi-source big spatiotemporal data. The three-stage assessment framework is applied to 290 major Chinese cities and allows for a comprehensive assessment regarding the urban green environment of this big developing country.The three stages of the assessment framework are (i) urban areas extraction, (ii) satellite-image-based greenspace mapping, and (iii) dynamic greenspace exposure assessment.Stage (i): For the purpose of focusing the assessment on urban areas, a more appropriate urban-area definition method is proposed. Notably, most existing satellite-based "urban area" extraction models merely aim to separate built-up areas with other land covers rather than also consider the spatial extent of human activities. That is, directly using the boundaries of "urban area" from such models or products will exclude part of the frequently accessed greenspace from the assessments. By integrating nighttime light (NTL) images (i.e., DMSP-OLS NTL data and NPP-VIIRS NTL data) and land-cover data (i.e., ESA-CCI land cover data), a method for urban-area extraction is developed, considering both the spatial extent of built-up surfaces and the areas with high human activities. The urban areas of China in 1992 and 2015 are successfully extracted via this method. Two different accuracy assessment methods (i.e., POI-based and land-cover-based) are then used to evaluate the results and demonstrate that the proposed method is effective and robust.Stage (ii): The second stage aims to map the distribution of greenspace within the received urban areas. Due to the advantages in free access, acceptable data size, and so on, medium-resolution remotely sensed images from Landsat-8 OLI (30-m) and Sentinel-2A (10-m) are widely used for green vegetation extractions. However, which is more suitable for urban-greenspace mapping and assessments remains unclear. Therefore, based on a linear spectral unmixing model and pixel-based static population data, a series of comparisons are conducted in terms of normalized difference vegetation index, greenspace extraction accuracy, greenspace coverage rate, and greenspace exposure level. The results not only demonstrate that Sentinel-2-based greenspace maps have better performance than maps derived from Landsat-8 images in both providing detailed information and generating higher accuracies, but also suggest that over-estimation will widely exist in Landsat-8-based urban greenspace assessments, especially when the population distribution is considered. Therefore, Sentinel-2A images are selected as appropriate data for urban greenspace mapping and assessments in this dissertation.Stage (iii): An improved model is proposed to dynamically measure people's ambient urban green environments (i.e., greenspace exposure) by integrating mobile-phone location-based service data and Sentinel-2A-based greenspace maps. In order to test the validity of the proposed method, it is applied to several major cities in China to evaluate residents' ambient greenspace with different buffer scales (i.e., 0.5 km, 1 km, and 1.5 km). The results demonstrate that the proposed method can effectively identify the greenspace exposure levels with different buffer-scale and their diurnal and daily variations, which makes it possible able to effectively quantify how residents enjoy greenspace during their daily lives. Moreover, the assessment model also provides an effective way to reduce the uncertainties caused by the modifiable areal unit problem (MAPU) and the uncertain geographic context problem (UGCoP).Based on the proposed assessment framework, the urban greenspaces of 290 major cities in China are assessed from three dimensions: (i) greenspace exposure level, (ii) inequality in greenspace exposure and the generation mechanism, and (iii) urbanization effects on urban greenspace (i.e., differences between old and new urban regions).Dimensions (i-ii): City-scale greenspace exposure levels are directly received by using a dynamic exposure assessment model, and the corresponding spatial features are further identified for different regions of China. In addition, relying on a modified greenspace-exposure-based Gini index, inequality in greenspace exposure during three periods (i.e., all-day, daytime, and night) are assessed for all selected cities. Then the generation mechanism of the inequality is explored, relying on two geographical detector models (i.e., factor detector model and interaction detector model). The results show that, in terms of greenspace exposure and the corresponding inequality, significant differences can be found among different cities and regions in China. Among the selected potential explanatory variables, higher greenspace coverage rate and lower land surface temperature are identified as the factors that play the most critical roles in reducing the inequality level in greenspace exposure. This result suggests that increasing green infrastructure (e.g., green roofs, green buildings, and green public space) can be an effective and practical way to not only improve greenspace exposure level but also reduce the inequality.Dimension (iii): Given how urbanization (particularly urban expansion) contributes to differing greenspaces in different urban areas has been limitedly addressed in existing researches, the last dimension of the assessment was trying to explore how urbanization affects urban greenspace and exposure level. The three-stage assessment framework was used to quantitatively estimate the effects of urban expansion (1992-2015) on people's exposure to green environments for 290 cities in China. Results showed that the urban expansion process directly led to differences in green environments between old/original urban areas ("old urban areas" hereafter) and newly built urban areas ("new urban areas" hereafter) in China. These differences were not only observed with the general evaluation index as the greenspace coverage rate but also captured using the dynamic exposure assessment model. For most of China's large cities, people could enjoy more greenspace in new urban areas than in old urban areas, except for a few cities located in Western China. Meanwhile, significant day-to-night variations in people's exposure to greenspace were identified between old and new urban areas.

ISBN: 9781392897744Subjects--Topical Terms:

3432445
Geographic information science.
Subjects--Index Terms:

Urban greenspace

Dynamic Exposure, Inequality and Urbanization Effects: A Multidimensional Evaluation of Urban Greenspace in China.
LDR:09403nmm a2200289 4500 001 2275344
005 20210119090706.5
008 220723s2019 ||||||||||||||||| ||eng d
020 $a 9781392897744
035 $a (MiAaPQ)AAI27784075
035 $a AAI27784075
040 $a MiAaPQ $c MiAaPQ
100 1 $a Song, Yimeng. $3 3553592
245 1 0 $a Dynamic Exposure, Inequality and Urbanization Effects: A Multidimensional Evaluation of Urban Greenspace in China.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2019
300 $a 200 p.
500 $a Source: Dissertations Abstracts International, Volume: 81-10, Section: B.
502 $a Thesis (Ph.D.)--The Chinese University of Hong Kong (Hong Kong), 2019.
520 $a As one of the most critical factors in the urban environment, urban greenspace plays a key role in mitigating a series of urban problems. Beyond that, serving as the main place for residential activities and social interactions, urban greenspace has been proven to improve people's physical and mental health. Over the past decades, however, rapid urbanization has greatly reshaped the land covers in and around urban areas, dramatically changing urban greenspace in both quality and quantity. As a result, the opportunity to enjoy urban greenery differs considerably for residents who live in different regions. Therefore, a scientific and thorough understanding of different urban green environments within cities and how people interact with greenspaces is urgently needed. However, most existing urban greenspace assessment methods do not incorporate human-related information-such as human distribution, human mobility, and so on-which may increase uncertainty of assessment results and lead to imprecise assessment. To address these issues, this dissertation provides an improved three-stage assessment framework based on multi-source big spatiotemporal data. The three-stage assessment framework is applied to 290 major Chinese cities and allows for a comprehensive assessment regarding the urban green environment of this big developing country.The three stages of the assessment framework are (i) urban areas extraction, (ii) satellite-image-based greenspace mapping, and (iii) dynamic greenspace exposure assessment.Stage (i): For the purpose of focusing the assessment on urban areas, a more appropriate urban-area definition method is proposed. Notably, most existing satellite-based "urban area" extraction models merely aim to separate built-up areas with other land covers rather than also consider the spatial extent of human activities. That is, directly using the boundaries of "urban area" from such models or products will exclude part of the frequently accessed greenspace from the assessments. By integrating nighttime light (NTL) images (i.e., DMSP-OLS NTL data and NPP-VIIRS NTL data) and land-cover data (i.e., ESA-CCI land cover data), a method for urban-area extraction is developed, considering both the spatial extent of built-up surfaces and the areas with high human activities. The urban areas of China in 1992 and 2015 are successfully extracted via this method. Two different accuracy assessment methods (i.e., POI-based and land-cover-based) are then used to evaluate the results and demonstrate that the proposed method is effective and robust.Stage (ii): The second stage aims to map the distribution of greenspace within the received urban areas. Due to the advantages in free access, acceptable data size, and so on, medium-resolution remotely sensed images from Landsat-8 OLI (30-m) and Sentinel-2A (10-m) are widely used for green vegetation extractions. However, which is more suitable for urban-greenspace mapping and assessments remains unclear. Therefore, based on a linear spectral unmixing model and pixel-based static population data, a series of comparisons are conducted in terms of normalized difference vegetation index, greenspace extraction accuracy, greenspace coverage rate, and greenspace exposure level. The results not only demonstrate that Sentinel-2-based greenspace maps have better performance than maps derived from Landsat-8 images in both providing detailed information and generating higher accuracies, but also suggest that over-estimation will widely exist in Landsat-8-based urban greenspace assessments, especially when the population distribution is considered. Therefore, Sentinel-2A images are selected as appropriate data for urban greenspace mapping and assessments in this dissertation.Stage (iii): An improved model is proposed to dynamically measure people's ambient urban green environments (i.e., greenspace exposure) by integrating mobile-phone location-based service data and Sentinel-2A-based greenspace maps. In order to test the validity of the proposed method, it is applied to several major cities in China to evaluate residents' ambient greenspace with different buffer scales (i.e., 0.5 km, 1 km, and 1.5 km). The results demonstrate that the proposed method can effectively identify the greenspace exposure levels with different buffer-scale and their diurnal and daily variations, which makes it possible able to effectively quantify how residents enjoy greenspace during their daily lives. Moreover, the assessment model also provides an effective way to reduce the uncertainties caused by the modifiable areal unit problem (MAPU) and the uncertain geographic context problem (UGCoP).Based on the proposed assessment framework, the urban greenspaces of 290 major cities in China are assessed from three dimensions: (i) greenspace exposure level, (ii) inequality in greenspace exposure and the generation mechanism, and (iii) urbanization effects on urban greenspace (i.e., differences between old and new urban regions).Dimensions (i-ii): City-scale greenspace exposure levels are directly received by using a dynamic exposure assessment model, and the corresponding spatial features are further identified for different regions of China. In addition, relying on a modified greenspace-exposure-based Gini index, inequality in greenspace exposure during three periods (i.e., all-day, daytime, and night) are assessed for all selected cities. Then the generation mechanism of the inequality is explored, relying on two geographical detector models (i.e., factor detector model and interaction detector model). The results show that, in terms of greenspace exposure and the corresponding inequality, significant differences can be found among different cities and regions in China. Among the selected potential explanatory variables, higher greenspace coverage rate and lower land surface temperature are identified as the factors that play the most critical roles in reducing the inequality level in greenspace exposure. This result suggests that increasing green infrastructure (e.g., green roofs, green buildings, and green public space) can be an effective and practical way to not only improve greenspace exposure level but also reduce the inequality.Dimension (iii): Given how urbanization (particularly urban expansion) contributes to differing greenspaces in different urban areas has been limitedly addressed in existing researches, the last dimension of the assessment was trying to explore how urbanization affects urban greenspace and exposure level. The three-stage assessment framework was used to quantitatively estimate the effects of urban expansion (1992-2015) on people's exposure to green environments for 290 cities in China. Results showed that the urban expansion process directly led to differences in green environments between old/original urban areas ("old urban areas" hereafter) and newly built urban areas ("new urban areas" hereafter) in China. These differences were not only observed with the general evaluation index as the greenspace coverage rate but also captured using the dynamic exposure assessment model. For most of China's large cities, people could enjoy more greenspace in new urban areas than in old urban areas, except for a few cities located in Western China. Meanwhile, significant day-to-night variations in people's exposure to greenspace were identified between old and new urban areas.
520 $a The results also found that urbanization not only plays a negative role in changing urban environments but also brings some positive effects to improve green environments for cities located in harsh natural conditions (e.g., semiarid/arid and desert regions).Overall, the potential highlights and contributions of this dissertation can be concluded as follows: (i) Information regarding human activities was well incorporated into the urban greenspace assessment framework; (ii) The proposed dynamic greenspace exposure model provided an effective way to reduce the uncertainties caused by the MAPU and the UGCoP; (iii) The generation mechanism of the city-scale inequality in greenspace exposure was first discussed from the angle of geographical features via geographical detector models; (iv) The differences in greenspace exposure between old and new urban regions were identified; (v) The proposed assessment framework could hold potential utilities in supporting urban environmental studies and health studies.
590 $a School code: 1307.
650 4 $a Geographic information science. $3 3432445
653 $a Urban greenspace
690 $a 0370
710 2 $a The Chinese University of Hong Kong (Hong Kong). $3 1017547
773 0 $t Dissertations Abstracts International $g 81-10B.
790 $a 1307
791 $a Ph.D.
792 $a 2019
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=27784075