IIIT Hyderabad Publications |
|||||||||
|
MODELING LAND USE-LAND COVER TRANSITION INDUCED MICROCLIMATE CHANGES: A CASE STUDY OF VIJAYAWADA CITY, ANDHRA PRADESH, INDIAAuthor: M. Vani Date: 2020-06-10 Report no: IIIT/TH/2020/49 Advisor:Ramachandra Prasad Pillutla AbstractUrbanization is a rapid and inevitable trend around the world with seemingly no sign of abating. The way in which urban sprawls are sustainably managed can be a major force for global adverse climate change mitigation and adaptation strategies; that entails better mass transit, more efficient lighting and heating systems for buildings, improved waste management and cleaner sources of energy production. All can ultimately have large positive effects for reducing climate change but are rooted in locally oriented strategies. The thesis thus aims to analyze the relationship between urban expansion and microclimate temperature fluctuations and develop an algorithm to forecast the land surface temperature based on the predicted land use-land cover changes. The thesis also extends to develop and test green measures on the adaptation front to reduce the intensifying daily surface temperature. Future increase in the size of the world's urban population is expected to be highly concentrated in developing countries that include India as well. The total number of towns and urban agglomerations were reported as 393 and 465 as per the 2001 and 2011 India census. In this regard, the thesis examines the Vijayawada urban agglomeration, Andhra Pradesh, India as the case study to validate the developed algorithm. The city, Vijayawada is one of the educational hubs of the state and is witnessing rapid population migration as well as temperature fluctuations both during summer and winter. Over the period, the city expanded at the cost of natural vegetation, hills, and water bodies leading to an increase in greenhouse gas emissions and loss of green spaces and wetlands existing within the city. The city expansion and the associated temperature fluctuations are studied with the aid of remote sensing technology. Temporal Landsat satellite images of four years viz., 1990, 2000, 2010 and 2018 are used to generate Land Use/Land Cover maps with four major classes viz.; built-up, vegetation, water body and others. Change detection and transition of the natural land cover to man-made land use was computed for the study area. Sprawl analysis of the city was carried out by generating multiple buffer rings over the study region to evaluate the urban density and annual urban growth rate. Shannon‘s entropy was employed to identify the nature of city expansion. The seasonal variation of the land surface temperature was studied using Mono-window algorithm. The temperature variation over individual classes was computed with the aid of a self-designed random point method. Results showed a steady increasing trend in the urban density and land surface temperature with the distinct formation of a heat island over the city especially during winters throughout the study period. The built-up area has increased from 28.20 km2 in 1990 to 138.01 km2 in 2018. The directional growth analysis captured the pattern of city growth as tentacle type development in conjunction with infill development. The prepared land use/land cover maps were taken forward to model future scenarios of the built-up dynamics of the city in cellular automata (CA) based environment. Accurate identification of the dominant driving factors in the expansion of a city is essential for CAbased urban expansion modeling. The drivers of change for the city were examined and then applied to model the future scenario of the city. Further, the study also aims at comparing the efficiency of the renowned CA-based urban growth model, SLEUTH (Slope, Land use, Exclusion, Urban extent, Transportation, Hill shade), and a self-designed CA-based hybrid model developed in combination with genetic algorithm (GA) to model future scenario of the city. Analysis of the statistical significance of the driving factors shows that the terrain and population density are the two dominant factors influencing the expansion of the city. Both the models' output on predicting the urban growth indicate that the available open spaces within the existing city extent get further converted to built-up indicating infill development, and more growth occurs along the fringes of the existing city. Validation of both the models showed overall accuracy higher than 80%. The inclusion of the demographic variable and utilization of the GA ensured that the hybrid model performs better than the SLEUTH model. Analysis of the built-up densification of the cityscape showed a shift in the pattern of built-up dynamics over time. An algorithm to forecast the future land surface temperature for the modeled scenario of the city in 2020 and 2030 has been developed based on multivariate kernel ridge regression. The land surface temperature is predicted for the modeled Land Use/Land Cover of the city by deriving the constraining ratio between the spatial and climatic variables. Development of this algorithm adds to the existing knowledge on microclimate change studies as seldom studies discusses on forecasting the microclimate temperature fluctuations associated with the geographical changes. Test for the stability and computational efficiency of the algorithm was performed over varying image resolutions. The forecast exhibited an increasing trend in the 8 land surface temperature for both summer and winter seasons irrespective of the Land Use/Land Cover type. In order to bring down the intensifying LST over years, greenery strategy as a soft measure has been tried and tested in the 3D perspective of a congested street in the core of the city study area. The temperatures observed at an interval of three hours for 21 hours showed a distinctive pattern for the different test cases viz., actual scenario, road-side vegetation, and combination of road-side and roof-top vegetation. The analysis concluded that the role of vegetation is significant in combating the increasing temperatures of the environment. The choice of the vegetation type and their location ie., whether road-side or roof-top or combination is purely dependent on the willingness of the inhabitants and the availability of the resources. Many studies have been performed to analyse the Land Use/Cover changes of a location over time using Remote Sensing technology. Modeling the future scenario of a region has also been done quite broadly using readily available software in recent years. However, the studies often restrict themselves within analysing the changes and modeling the future scenarios. Though climate change has been a debatable topic over the past decade, seldom studies relate the actual geographical changes and the associated microclimate fluctuations. This thesis thus stands out by bridging the gap between them by developing an algorithm that relates the spatial variable (land use/cover features) and the climatic variable (land surface temperature). This further adds to the existing knowledge on microclimate change related studies. Further the thesis also extends to recommend practical solution for the increasing temperatures observed in the chosen city scale congested street canyon. On an extended note, the thesis also covers the development of a graphical user interface for an established urban growth model; which otherwise has bugging issues and complexities in implementation. This contributes towards the development of an existing system thereby improving the efficiency of the model available. To overcome certain drawbacks observed in the model, a self-designed advanced hybrid model has further been developed to improve the efficiency and user experience. Full thesis: pdf Centre for Spatial Informatics |
||||||||
Copyright © 2009 - IIIT Hyderabad. All Rights Reserved. |