Future Urban Growth Modelling for Vietnam based on SLEUTH and EO-products

Dies ist die aktuellste Version dieses Eintrags.

Kurzfassung

Urbanization is an environmental transformation process that the conversion of rural and natural areas into urban settlements. Accompanied by a rural-urban migration, this process has been rapidly accelerated since the second half of the last century. In 1950, 30 percent of the world's population lived in urban agglomerations. This proportion has increased to 55 percent in 2018 (United Nations Department of Economic and Social Affairs, 2019). In the coming decades, the trend of urbanization is expected to continue, notably in small and mid-sized cities in Asia and Africa. Vietnam belongs to the fastest urbanizing countries in the Asian-Pacific region, mainly driven by a significant rural-urban migration. Besides the large urban centres like Ho Chi Minh City, Ha Noi or Da Nang, more than 50 medium-sized cities like Hue, Nha Trang or Dong Hoi play an important role in the urbanization process. These cities show diverse growth patterns and are facing significant challenges that come with the urbanization. Besides land-use conflicts, the need for rapid planning approaches in various sectors, increased development of a sustainable supply infrastructure, many cities are also facing an increase in climate-related extreme events and hazards such as flooding (coastal, fluvial and flash floods) (Nguyen et al., 2021) leading to increasing exposure of urban populations, assets and infrastructures. Earth observation (EO)-based information has been widely used for urban modelling for decades and have proven their potential for urban growth simulation and prediction (Donnay et al., 2014). Nowadays, EO data from satellites with shorter repeat cycles and higher resolution enable spatially explicit urban modelling with a high temporal information density. The World Settlement Footprint Evolution (WSF Evolution) (Marconcini et al., 2021), developed by German Aerospace Center (DLR) and ESA, is a comprehensive annual dataset that outlines settlements and urban agglomerations with a global coverage from 1985 to 2015. For the derivation of dataset, training data from the WSF2015 layer with 10 m resolution (based on Sentinel-1 and Landsat-8) has been used. An iterative backward classification approach based on machine learning, exploited the entire Landsat archive from 1985-2015 (Marconcini et al., 2020). The dataset describes the development of human settlements on an annual base and therefore provides enormous potential for future urban growth modelling. In the proposed study, a national scale urban growth projection until 2050 under a business-as-usual scenario has been conducted for Vietnam. The analysis is based on the SLEUTH cellular automata model using a genetic algorithm (GA) (Clarke, 2017). The analysis is based on the WSF Evolution in full spatial resolution and conducted for the entire Vietnamese territory, using the WSF settlement information starting from 1985, road infrastructure data, topographic data from the Copernicus Digital Elevation Model (DEM) and areas that are excluded from urban development, like protected natural areas. To overcome the computing cost for such a large data volume and improve the efficiency of the SLEUTH calibration process, a tiling processing strategy based on local growing speed and k-mean clustering was applied. Results confirm the potential of the WSF Evolution for urban growth prediction. The predicted future urban growth illustrates the most dynamic development hotspots across the entire country for the upcoming decades and demonstrates different trends of urban developments in various regions. Based on the current model configuration, different growth scenarios will be developed and modelled in order to complement the business-as-usual scenario. Findings of this study can support sustainable planning and will support research activities e.g. in the Vietnamese-German research projects FloodAdaptVN (https://floodadapt.eoc.dlr.de/) and Drought-ADAPT (https://www.bmbf-client.de/en/projects/drought-adapt).

Verfügbare Versionen dieses Eintrags

• Future Urban Growth Modelling for Vietnam based on SLEUTH and EO-products. (deposited 22 Sep 2022 09:36) [Gegenwärtig angezeigt]