| Author: | Faridatul, Mst Ilme |
| Title: | Research on the hydrological effects of urbanization based on multiple-source geospatial data |
| Advisors: | Wu, Bo (LSGI) |
| Degree: | Ph.D. |
| Year: | 2020 |
| Subject: | Hydrology Urban hydrology Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Land Surveying and Geo-Informatics |
| Pages: | xvi, 161 pages : color illustrations |
| Language: | English |
| Abstract: | Urbanization results in the physical growth of cities. It is therefore an important worldwide concern, as excess urbanization creates various problems, such as water scarcity, urban sprawl, and environmental pollution. It encroaches upon land available for other types of development, thereby reducing available developable land and expediting the reclaiming of land via filling bodies of surface water. Moreover, urbanization induced land cover changes affect watersheds and the hydrologic system. It is important to investigate and understand the hydrological effects of urbanization, both to improve watershed management and planning and to develop adaptation strategies to address the challenges posed by urbanization. Remote sensing (RS) geospatial data is invaluable for quantifying and monitoring urbanization and its hydrological effects on a global scale. Various satellite RS techniques have been developed to facilitate the quantification of urban growth based on land cover classification, and the monitoring of changes in surface water and other important hydrologic components, e.g., evapotranspiration (ET). In this research, multi-year RS data were used, i.e., Landsat 8 Operational Land Imager (OLI), Landsat 5 Thematic Mapper (TM) and Landsat 7 Enhanced Thematic Mapper (ETM). Two high-resolution images i.e., SPOT-6 and ZY-3, and the SRTM DEM were also used in this research. This study focused on improving the assessment of two key hydrological effects of urbanization: (1) changes in surface water, and (2) changes in the amount of water transfer from the urban environment to the atmosphere, in the form of ET. A detailed summary of the method and findings is given below. First, an approach for mapping major urban land covers based on three novel spectral indices was developed, i.e., a modified normalized difference bare-land index (MNDBI), a tasseled cap water and vegetation index (TCWVI) and a shadow index (ShDI). RS data from multiple sensors were used to assess the performance of the proposed indices for multi-temporal land cover mapping. The indices were developed to enhance the classification of major land covers and facilitate the accurate mapping of urbanization and surface water. The MNDBI was developed to improve the detection of urbanization by separating impervious urban areas from bare lands. The TCWVI was formulated to facilitate the simultaneous detection of vegetation and water, and the ShDI was developed to separate dark building shadows from water for the enhancement of water detection in urban areas. An approach to optimizing thresholds was also developed, based on the spectral reflectance of the indices. Next, the optimized thresholds were inputted to classify the land covers. The results indicated that the indices developed in this research characterized similar patterns in the distribution of land covers. However, the spectral reflectance and typical statistics of the indices showed variations in sensors and study areas. The experimental evaluation confirmed that the novel spectral indices facilitate a reliable mapping of major urban land covers using Landsat TM, ETM and OLI data that provides between 93-96% accuracy with a kappa coefficient of 0.91-0.96. Second, an assessment was made of the changes in surface water in Hong Kong and Dhaka in relation to urbanization using multi-year land cover maps spanning 27 years. Various factors affect surface water, such as urbanization, changes in land cover types, seasons and topography. The influence of all these factors was evaluated to facilitate the sustainable management of urban watersheds. The influence of seasonality on the distribution of water was evaluated using pre- and post-monsoon satellite images. Then, multi-year land cover maps and a contingency matrix were used to evaluate the influence of urbanization on the occurrence of and transitions in surface water. The subsequent application of an advanced geo-statistical tool enabled investigation of the spatially varying relationships of changes in surface water in relation to changes in various driving factors. The results showed that urbanization induced land cover changes largely affected surface water. However, the changes were variable to urban areas. In the Dhaka study site, surface water reduced to 8.12% from 23.26%. In contrast, in the Hong Kong study site, it reduced to 4.21% from 5.03%. The experimental evaluation confirmed that the changes in surface water area showed stronger variability in Dhaka than in Hong Kong, because the topography of the Dhaka study area is relatively flat with a mean terrain elevation of 14 m. The availability of flat land facilitated horizontal urban expansion towards low-lying wetlands and surface water in Dhaka. In contrast, a large area of Hong Kong covers an elevation of more than 100 m, which acted as a barrier for the horizontal expansion thus the changes in surface water were least in Hong Kong. Third, the impact of urbanization and land cover changes on the amount of water transfer from the urban environment to the atmosphere was examined, which required multi-temporal ET mapping to be developed. An improved surface energy balance algorithm for urban areas (uSEBAL) was developed in this research, which was suitable for the estimation of ET in urban environments. Urban land cover compositions were considered in the delineation of the land surface parameters, and anthropogenic heat was delineated and incorporated in the land surface energy budget. The results of the uSEBAL were compared with the traditional method of SEBAL, and a sensitivity analysis was performed to evaluate the impact of uncertainties in ET estimates. The model performance metrics indicated that the uSEBAL is better than the SEBAL for estimating ET in a heterogeneous urban environment. However, the errors in the estimation of solar radiation and surface albedo can result in the highest uncertainty in the estimation of ET. The findings of this study also indicated that the variability in urban land cover types impacts spatial variability in energy fluxes and ET. The results showed a seasonal influence on ET for different land covers, but no significant influence of seasonality was observed on urban impervious areas, which produced the lowest ET nearly zero mm/day. An investigation of factors of changes in ET indicated that several important factors, including urbanization, significantly affected ET, due to related factors such as the expansion of non-evaporative impervious land cover and the reduction of evaporative surfaces. The results showed a strong correlation between urbanization and a decrease in ET. During the 27 years, the number of pixels of low ET (< 0.2 mm/day) increased to 147,528 from 49,988. Pearson's correlation coefficient (r = 0.82) indicated that solar radiation and air temperature highly influenced ET in urban areas. Moreover, a strong negative correlation (r = -0.98 ~ -0.99) was observed between ET and surface albedo. A strong negative correlation (r > -78) was also observed between ET and land surface temperature (LST). However, the experimental evaluations showed the least influence of NDVI on ET in the heterogeneous urban environment. In view of the complex characteristics of the urban environment, the methods developed in this research provide quantitative ways of assessing the changes in surface water and ET, and will also be applicable to analogous assessments in other cities in the world. The results presented are invaluable in increasing our understanding of the influence of various driving factors of changes in urban hydrology. Particularly, this research is important in enabling the development of a long-term perspective on urbanization and its impact on the transformation of major land covers, including surface water, and the consequent effect of any changes on atmospheric water transfer. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/10516