We present an algorithm for daytime radiation fog detection (ADRFD) to characterize the fog distribution over land areas through two main steps: (1) detection of areas with clouds and radiation fog based on edge pixels and (2) separation of radiation fog from clouds based on object properties. The algorithm is tested in southeast China over an area of 2,640,000 km2 (longitude extent: 105°E to 122°E, latitude extent: 23°N to 40°N), with 1 km resolution moderate resolution imaging spectroradiometer/TERRA images, digital elevation model of China (grid size is 1×1 km2 and accuracy of elevation is 25 m), and the fog detection accuracies are evaluated using the observation results from 3590 ground observation stations. Results show that ADRFD can detect areas with clouds and radiation fog and effectively differentiate radiation fog from clouds based on object properties with a Kappa coefficient of 0.89 and critical success index at 0.87. It is concluded that ADRFD is a promising approach for daytime radiation fog detection over large land surfaces based on the condition that fog objects do not intersect with cloud objects or are not located under cloud objects. Extensions of this current study will be on improving the parameters determination method of ADRFD. ADRFD can be used to detect other types of fog in different regions and different seasons theoretically, but it should be tested further. In addition, more improvements are needed to allow for the detection of smaller areas of fog and fog regions intersecting with or under clouds.
Urbanization transforms the natural landscape to anthropogenic urban land use and changes surface physical
characteristics. Accurate information on the extent of urban growth and its impacts on environment are of great interest
for diverse purposes. As a result, increased research interest is being directed to the mapping and monitoring of urban
land use using remote sensing techniques. However, there are many challenges in deriving urban extent and development
densities quantitatively. This study utilized remote sensing data of Landsat TM/ETM+ to assess urban sprawl and its
thermal characteristics in Changsha of central China. A new approach was proposed for quantitatively determining urban
land use extents and development densities. Firstly, impervious surface areas were mapped by integrating spectral index
derived from remotely sensed data. Then, the urban land extents and development densities were identified by using
moving window calculation and selecting certain threshold values. The urban surface thermal patterns were investigated
using Landsat thermal band. Analysis results suggest that urban extent and development density and surface thermal
characteristics and patterns can be identified through qualitatively based remotely sensed index and land surface
temperature. Results show the built-up area and urban development densities have increased significantly in Changsha
city since 1990s. The differences of urban development densities correspond to thermal effects where higher percent
imperviousness is usually associated with higher surface temperature. Remotely sensed index and land surface
temperature are demonstrated to be very useful sources in quantifying urban land use extent, development intensity, and
urban thermal patterns.
Geographical Cellular Automata (GCA) approach is based on complexity theory and is widely used in geospatial
modeling. A reason for the increasing attention given to GCA models is that they can easily be integrated with rasterbased
GIS environment. However, the behavior of the GCA models is affected by uncertainties arising from the
interaction between model elements, structures, and the quality of data sources used as model input. The objective of
this study is to examine the impacts of model elements on the generated outputs of a GIS-based GCA land-use growth
model using sensitivity analysis (SA) approach. The proposed SA method consists of KAPPA index with different
spatial metrics. A stochastic GCA model was built to model land use change in the changsha region (Hunan,China). The
transition rules were empirically derived from four Landsat-TM (30m resolution) images taken in 1996,1999, 2002 and
2005 that have been resampled to four resolutions (30, 60, 90, 120m). Five different neighbourhood configurations were
considered (Moore, Von Neumann, and circular approximations of 2, 3 and 4 cell radii). Simulations were performed
for each of the twenty spatial scale scenarios. Results show that spatial scale has a considerable impact on simulation
dynamics in terms of both land use area and spatial structure. The spatial scale domains present in the results reveal the
nonlinear relationships that link the spatial scale components to the simulation results.
This paper analyzes and simulates the land use changes in the Pearl River Delta, China, using Longgang City as a case study. The region has pioneered the nation in economic development and urbanization process. Tremendous land use changes have been witnessed since the economic reform in 1978. Land use changes are analyzed and simulated by using stochastic cellular automata model, land use trajectories analysis, spatial indices and multi-temporal TM images of Longgang City (TM1987, TM1991, TM1995, TM1999, TM2003, TM2005) in order to understand how urbanization has transformed the non-urban land to urban land and estimate the consequent environment and ecological impacts in this region. The analysis and simulation results show that urban land continues to sprawl along road and fringe of towns, and concomitant to this development is the loss of agricultural land, orchards and fish ponds. This study provides new evidence with spatial details about the uneven land development in the Pearl River Delta.
Urbanization is the complex process of converting rural land uses to urban land uses, which has caused significantly land cover changes and associated surface characteristics. Therefore, researches on land cover and its landscape pattern change under urbanization are essential for analyzing the impacts of human activities on environment. This study firstly detected land cover (i.e., forestland, non-forest vegetation, built-up area, and water) changes in Changsha City from 1973 to 2005 by using the multi-temporal Landsat images (TM1973, TM1993, TM1998, ETM+2001) and land use map (2005); and then analyzed the spatiotemporal changes of landscape pattern at landscape-level and class-level by using FRAGSTATS, respectively. At last, the class-level metrics of each land cover class were further regressed to the degree of urbanization. The results indicated that: (1) in the context of urbanization, the built-up area and the non-forest vegetation experienced a significant changes, while the forestland and water remained relatively unchanged, and the non-forest vegetation cover bore the major burden of urbanization; (2) with the advance of urbanization, the change of overall landscape pattern of Changsha represented a complex dynamic process; (3) obvious differences of impacts of urbanization on landscape patterns of various land cover classes existed, i.e., along with the decrease of MPS of non-forest vegetation, the AI of built-up area increased dramatically; (4) some class-level metrics of various land cover classes were strongly correlated to the degree of urbanization, but the correlated extend varied along with the various land cover classes. To sum up, this study demonstrated the differences of impacts of urbanization on various land cover patterns. The results have the potential to assist land-use planning and management.