Zhiqiang Lv

Quantifying different types of urban growth and the change dynamic in Guangzhou using multi-temporal remote sensing data

Abstract There is a widespread concern about urban sprawl. It has negative impacts on natural resources, economic health, and community character. Without a universal definition of urban sprawl, its quantification and modeling is difficult. Traditionally, urban sprawl was described using qualitative terms, and landscape patterns. Quantitative methods are required to help local, regional and state land use planners to better identify, understand and address it. In this study, an integrated approach of remote sensing and GIS was used to identify three urban growth types of infilling growth, outlying growth and edge-expansion growth at the city of Guangzhou, China. Spatial metrics were used to characterize long-term trends and patterns of urban growth. Result shows that the proposed method can identify and visualize different urban growth types. Infilling growth is the dominant expansion type. Edge-expansion is concentrated at suburban areas. Outlying growth mainly occurs relatively far from the urban core. The analysis shows that initially the urban area expands mainly as outlying growth, causing increased fragmentation and dispersion of urban areas. Next, growth filled in vacant non-urban area inwards, resulting into a more compact and aggregated urban pattern. The study shows an improved understanding of urban growth, and helps to provide an effective way for urban planning.

Evaluation of urban sprawl and urban landscape pattern in a rapidly developing region

Urban sprawl is a worldwide phenomenon happening particularly in rapidly developing regions. A study on the spatiotemporal characteristics of urban sprawl and urban pattern is useful for the sustainable management of land management and urban land planning. The present research explores the spatiotemporal dynamics of urban sprawl in the context of a rapid urbanization process in a booming economic region of southern China from 1979 to 2005. Three urban sprawl types are distinguished by analyzing overlaid urban area maps of two adjacent study years which originated from the interpretation of remote sensed images and vector land use maps. Landscape metrics are used to analyze the spatiotemporal pattern of urban sprawl for each study period. Study results show that urban areas have expanded dramatically, and the spatiotemporal landscape pattern configured by the three sprawl types changed obviously. The different sprawl type patterns in five study periods have transformed significantly, with their proportions altered both in terms of quantity and of location. The present research proves that urban sprawl quantification and pattern analysis can provide a clear perspective of the urbanization process during a long time period. Particularly, the present study on urban sprawl and sprawl patterns can be used by land use and urban planners.

Landscape Pattern Metrics: An Empirical Study from 2-D to 3-D

Analysis of landscape patterns, as an important means of studying landscape ecological functions and processes, is a key part of landscape ecology. The traditional patch-corridor-matrix paradigm, ignoring the differences between real three-dimensional (3-D) landscapes and two-dimensional (2-D) landscape maps, quantifies and describes 2-D landscape maps using landscape pattern metrics. Such simplification of the real landscape will lead to differences between 2-D landscape pattern metrics and real values, especially in mountainous areas. This study seeks to examine whether these differences are significant in natural landscape patterns. Fourteen small watersheds in a mountainous area were selected as sample landscape units. Two metrics at the patch level, five metrics at the class level, and four metrics at the landscape level were chosen as representative metrics. The results indicated that there were significant differences between the 2-D and 3-D landscape pattern metrics at the patch level, landscape level, and class level. We interpret the differences as errors in the 2-D metrics. The errors depend on the metric and the landscape type, and the errors for different landscape types were inconsistent. Whether these errors affect further interpretation of the role of pattern in landscape ecology remains uncertain.

Network transmission of hidden data using smartphones based on compromising emanations

In this paper, we proposes a method to obtain the stored in the air-gapped computer with the help of nearby smartphones by intentionally generating loud physical emanations. We can construct a big transmission network connecting a computer and many cell phones by using different kinds of physical emanations. We further demonstrate the scenario by implementing a system using electromagnetic waves and acoustical signals. In this system, convert transmission channels are established between the computer and the cell phones. The microphone and FM receiver of the mobile phone are used to obtain the information. This kind of hidden data transmission method has the characteristics of strong concealment.

AdvRefactor: A Resampling-Based Defense Against Adversarial Attacks.

Deep neural networks have achieved great success in many domains. However, they are vulnerable to adversarial attacks, which generate adversarial examples by adding tiny perturbations to legitimate images. Previous studies providing defense mostly focus on modifying DNN models to mitigate adversarial attacks. We propose a resampling-based defense, AdvRefactor, which aims at transforming the inputs of models, and thereby eliminates the adversarial perturbations. We explore two resampling algorithms, proximal interpolation and bilinear interpolation, which cost less, suit more models and are combinable with other defenses. Our evaluation results demonstrate that AdvRefactor can significantly mitigate the adversarial attacks.

Evaluation of urban sprawl and urban landscape pattern in a rapid developing region

Urban sprawl is a worldwide phenomenon especially in fast developing regions of China which is a typical representative. The study on the spatiotemporal characteristics of urban sprawl and urban pattern is useful for land sustainable management and urban land planning. This research explores the spatiotemporal dynamics of urban sprawl in the context of a rapid urbanization process in a booming economic region of southern China from 1979 to 2005. Three urban sprawl types are distinguished based on analyzing overlaid urban area maps of each two adjacent study years which originates from the interpretation of remote sensed images and the vector land use maps. The landscape metrics are used to analyze the spatiotemporal pattern of urban sprawl for each study period. The study results show that urban areas have expanded dramatically, and the spatiotemporal landscape pattern configured by three sprawl types changed obviously.

Pattern analysis and intensity evaluation of urban land use in east peral river dleta, China

Using four classified Landsat images and two vector land use map that covered the study area, the change of urban land use, urban land patterns, and urban land intensity were quantized. After urban expansion analysis, changes of landscape pattern were then quantified by a series of landscape metrics. Additionally, two indices, IULU and LOE, were proposed by us to evaluate the sustainable efficiency of urban land use. The analysis showed that the urban land of the study area increased dramatically among the study period, meanwhile urban patterns for the six years changed obviously, and the intensity of the urban land use increased a lot during these study years. This study indicated that the urban landscape and the sustainable analysis for urban configuration can give us a clear perspective of the process of the urbanization in a long time span, and the studies about urban sprawl and urban pattern can be effectively utilized by the land planners and the urban planners.

Monitoring of urban sprawl using remote sensing technology in the rapid urbanized region, China

Analyzing spatio-temporal characteristics of land use is essential for understanding ecological consequences of urbanization. More importantly, such analyses can provide basic information for sustainable management of land resources. The East Pearl River Delta (EPRD) in southern China, which has experienced a rapid urbanization process from 1979 to 2005, was taken as a study region to analyze the spatio-temporal dynamics of urban land use. Land use map with all of its land uses classified into 4 types which was generated based on Landsat images, together with local basic land-use information layers, were used to study the regions urban land use pattern and intensity. This analysis showed that the urban land area and intensity within the region increased dramatically during the study period.

Arable land spatial pattern analysis along the middle-lower of Dongjiang River, China

Regional land use change has been and will continue to be one of the hugest human impacts on the environment. Understanding the spatial configuration of urban areas will be important for quantifying dynamic change of land use. Arable land plays an important role in agricultural and economic development of China especially in Guangdong province. It is not only the basic material of crop production, but also the foundation and guarantee for implementing crop security. Based on the analysis of present situation of arable land, a land management policy for dynamic equilibrium of total arable land can be implemented. In order to understand the landscape pattern and the gradient characteristic of arable land along the middle-lower of Dongjiang River, we quantify the arable pattern with the combination of gradient analysis and landscape metrics. It is found that: (1) paddy land is the dominating land use type, the pattern of paddy land has strong effect on the whole landscape pattern, (2) The same metric value about different arable land use type has the obvious difference (3) Through out the gradient analysis, the metric chart of the arable land pattern has the inflexion near the borderline of the cities.