Zhongchang Sun

Estimating urban impervious surfaces from Landsat-5 TM imagery using multilayer perceptron neural network and support vector machine

In recent years, the urban impervious surface has been recognized as a key quantifiable indicator in assessing urbanization impacts on environmental and ecological conditions. A surge of research interests has resulted in the estimation of urban impervious surface using remote sensing studies. The objective of this paper is to examine and compare the effectiveness of two algorithms for extracting impervious surfaces from Landsat TM imagery; the multilayer perceptron neural network (MLPNN) and the support vector machine (SVM). An accuracy assessment was performed using the high-resolution WorldView images. The root mean square error (RMSE), the mean absolute error (MAE), and the coefficient of determination (R2) were calculated to validate the classification performance and accuracies of MLPNN and SVM. For the MLPNN model, the RMSE, MAE, and R2 were 17.18%, 11.10%, and 0.8474, respectively. The SVM yielded a result with an RMSE of 13.75%, an MAE of 8.92%, and an R2 of 0.9032. The results indicated that SVM performance was superior to that of MLPNN in impervious surface classification. To further evaluate the performance of MLPNN and SVM in handling the mixed-pixels, an accuracy assessment was also conducted for the selected test areas, including commercial, residential, and rural areas. Our results suggested that SVM had better capability in handling the mixed-pixel problem than MLPNN. The superior performance of SVM over MLPNN is mainly attributed to the SVMs capability of deriving the global optimum and handling the over-fitting problem by suitable parameter selection. Overall, SVM provides an efficient and useful method for estimating the impervious surface.

Long-term effects of land use/land cover change on surface runoff in urban areas of Beijing, China

Abstract The objective of this paper is to present a case study to derive land use/land cover (LULC) maps and investigate the long-term effects of LULC change on surface runoff in the fast urbanizing Beijing city. The LULC maps were derived from Landsat TM/ETM+ imagery (acquired in 1992, 1999, 2006, and 2009) using support vector machine method. A long-term hydrologic impact assessment model was applied to assess the impact of LULC change on surface runoff. Results indicated that the selected study area experienced rapid urbanization from 1992 to 2009. Because of urbanization, from 1992 to 2009, modeled runoff increased 30% for the whole area and 35% for the urban portion. Our results also indicated that the runoff increase was highly correlated with urban expansion. A strong relationship ( R 2 = 0.849 ) was observed between the impervious surface percent and the modeled runoff depth in the study area. In addition, a strong positive relationship was observed between runoff increase and percentage of urban areas ( R 2 = 0.997 for the whole area and R 2 = 0.930 for the urban portion). This research can provide a simple method for policy makers to assess potential hydrological impacts of future urban planning and development activities.

Unequal Dual-Band Rat-Race Coupler based on Dual-Frequency 180 Degree Phase Shifter

This paper presents a dual-band rat-race coupler with arbitrary power dividing ratio. The dual-band and unequal power dividing characteristics are obtained by using a dual-frequency 180 degree phase shifter with the conventional coupler. Closed-form design equations are derived. The proposed structure can support large power dividing ratio with a wide frequency ratio range. The measured results show that good matching, isolation, power transmission and phase balance can be achieved at two operating frequency bands.

Spatiotemporal analysis of urban environment based on the vegetation–impervious surface–soil model

Abstract This study explores a spatiotemporal comparative analysis of urban agglomeration, comparing the Greater Toronto and Hamilton Area (GTHA) of Canada and the city of Tianjin in China. The vegetation–impervious surface–soil (V–I–S) model is used to quantify the ecological composition of urban/peri-urban environments with multitemporal Landsat images (3 stages, 18 scenes) and LULC data from 1985 to 2005. The support vector machine algorithm and several knowledge-based methods are applied to get the V–I–S component fractions at high accuracies. The statistical results show that the urban expansion in the GTHA occurred mainly between 1985 and 1999, and only two districts revealed increasing trends for impervious surfaces for the period from 1999 to 2005. In contrast, Tianjin has been experiencing rapid urban sprawl at all stages and this has been accelerating since 1999. The urban growth patterns in the GTHA evolved from a monocentric and dispersed pattern to a polycentric and aggregated pattern, while in Tianjin it changed from monocentric to polycentric. Central Tianjin has become more centralized, while most other municipal areas have developed dispersed patterns. The GTHA also has a higher level of greenery and a more balanced ecological environment than Tianjin. These differences in the two areas may play an important role in urban planning and decision-making in developing countries.

New approaches to urban area change detection using multitemporal RADARSAT-2 polarimetric synthetic aperture radar (SAR) data

With the current pace of regional and global urbanization, change detection and monitoring of urban areas using multitemporal synthetic aperture radar (SAR) datasets is becoming an important research topic for SAR Earth observation. Many SAR change detection methods have been proposed, but the change detection methods based on full polarimetric SAR and polarimetric SAR interferometry have not been extensively studied. In addition, due to the complexity of urban environments, most of the change detection methods were focused on natural targets. In this study, three urban area change detection methods are proposed using multitemporal RADARSAT-2 polarimetric SAR datasets; the Stationary Index Method (SIM) of urban area change detection using polarimetric SAR interferometry, the Coherence Index Method (CIM) of urban area change detection using polarimetric SAR interferometry, and the Two Scattering Components Method (TSCM) of manmade target change detection using polarimetric SAR. Three methods were applied to detect changes in an urban area after the 2008 Wenchuan earthquake. Through filed observation and high-resolution optical datasets, the true positives ratio (the ratio between the number of correctly detected “changed” pixels and the number of “changed” pixels) of manmade targets of the SIM, CIM, and TSCM were 81.34%, 89.00%, and 78.75%, respectively. The true negatives ratio (the ratio between the number of correctly detected “unchanged” pixels and the number of “unchanged” pixels) of manmade targets of the SIM, CIM, and TSCM were 87.17%, 71.42%, and 74.89%, respectively. The true positives ratio of the crops field for the SIM and CIM were 75.87% and 97.02%, respectively. The results indicated that these three methods can be effectively used to detect urban area change and manmade target change, demonstrating that RADARSAT-2 multitemporal datasets have great potential for urban area change detection.

Extracting distribution and expansion of rubber plantations from Landsat imagery using the C5.0 decision tree method

Abstract. Natural tropical rainforests in China’s Xishuangbanna region have undergone dramatic conversion to rubber plantations in recent decades, resulting in altering the region’s environment and ecological systems. Therefore, it is of great importance for local environmental and ecological protection agencies to research the distribution and expansion of rubber plantations. The objective of this paper is to monitor dynamic changes of rubber plantations in China’s Xishuangbanna region based on multitemporal Landsat images (acquired in 1989, 2000, and 2013) using a C5.0-based decision-tree method. A practical and semiautomatic data processing procedure for mapping rubber plantations was proposed. Especially, haze removal and deshadowing were proposed to perform atmospheric and topographic correction and reduce the effects of haze, shadow, and terrain. Our results showed that the atmospheric and topographic correction could improve the extraction accuracy of rubber plantations, especially in mountainous areas. The overall classification accuracies were 84.2%, 83.9%, and 86.5% for the Landsat images acquired in 1989, 2000, and 2013, respectively. This study also found that the Landsat-8 images could provide significant improvement in the ability to identify rubber plantations. The extracted maps showed the selected study area underwent rapid conversion of natural and seminatural forest to a rubber plantations from 1989 to 2013. The rubber plantation area increased from 2.8% in 1989 to 17.8% in 2013, while the forest/woodland area decreased from 75.6% in 1989 to 44.8% in 2013. The proposed data processing procedure is a promising approach to mapping the spatial distribution and temporal dynamics of rubber plantations on a regional scale.

A Modified Normalized Difference Impervious Surface Index (MNDISI) for Automatic Urban Mapping from Landsat Imagery

Impervious surface area (ISA) is a key factor for monitoring urban environment and land development. Automatic mapping of impervious surfaces has attracted growing attention in recent years. Spectral built-up indices are considered promising to map ISA distributions due to their easy, parameter-free implementations. This study explores the potentials of impervious surface indices for ISA mapping from Landsat imagery using a case study area in Boston, USA. A modified normalized difference impervious surface index (MNDISI) is proposed, and a Gaussian-based automatic threshold selection method is used to identify the optimal MNDISI threshold for delineating impervious surfaces from background features. To evaluate its effectiveness, comparison analysis is conducted between MNDISI and the original NDISI using Landsat images from three sensors (TM/ETM+/OLI-TIRS) acquired in four seasons. Our results suggest that built-up indices are sensitive to image seasonality, and summer is the best time phase for ISA mapping. With reduced uncertainties from automatic threshold selection, the MNDISI extracts impervious surfaces from all Landsat images in summer with an overall accuracy higher than 87% and an overall Kappa coefficient higher than 0.74. The proposed method is superior to previous index-based ISA mapping from the enhanced thermal integration and automatic threshold selection. The ISA maps from the TM, ETM+ and OLI-TIRS images are not significantly different. With enlarged data pool when all Landsat sensors are considered and automation of threshold selection proposed in this study, the MNDISI could be an effective built-up index for rapid and automatic ISA mapping at regional and global scales.

Improved alpine grassland mapping in the Tibetan Plateau with MODIS time series: a phenology perspective

The Tibetan Plateau is primarily composed of alpine grasslands. Spatial distributions of alpine grasses, however, are not well documented in this remote, highly uninhabited region. Taking advantage of the frequently observed moderate resolution imaging spectroradiometer (MODIS) images (500-m, 8-day) in 2010, this study extracted the phenological metrics of alpine grasses from the normalized difference vegetation index time series. With the Support Vector Machine, a multistep classification approach was developed to delineate alpine meadows, steppes, and desert grasses. The lakes, permanent snow, and barren/desert lands were also classified with a MODIS scene acquired in the peak growing season. With ground data collected in the field and aerial experiments in 2011, the overall accuracy reached 93% when alpine desert grasses and barren lands were not examined. In comparison with the recently published national vegetation map, the alpine grassland map in this study revealed smoother transition between alpine meadows and steppes, less alpine meadows in the southwest, and more barren/deserts in the high-cold Kunlun Mountain in the northeast. These variations better reflected climate control (e.g. precipitation) of different climatic divisions on alpine grasslands. The improved alpine grassland map could provide important base information about this cold region under the pressure of rapidly changing climate.

Study of RADARSAT-2 synthetic aperture radar data for observing sensitive factors of global environmental change

Abstract Global environmental change has gained widespread global attention. It is a complex system with special spatial and temporal evolutionary characteristics. Sensitive factors are indicators of global environmental change, and some can be observed with Earth observation technology. RADARSAT-2 is capable of polarimetric and interferometric observations, which can provide an effective way to document some sensitive factors of global environmental change. This study focuses on the usage of RADARSAT-2 data for observing sensitive factors of environmental change and building highly accurate application models that connect synthetic aperture radar data and observable sensitive factors. These include (1) extracting spatiotemporal distribution of large-scale alluvial fan, (2) extracting vegetation vertical structure, (3) detecting urban land cover change, and (4) monitoring seasonal floods. From this study, RADARSAT-2 data have been demonstrated to have excellent capabilities in documenting several sensitive factors related to global environmental change.

A novel image-fusion method based on the un-mixing of mixed MS sub-pixels regarding high-resolution DSM

ABSTRACT A major reason for the spectral distortions of fused images generated by current image-fusion methods is that the fused versions of mixed multispectral (MS) sub-pixels (MSPs) corresponding to panchromatic (PAN) pure pixels remain mixed. The MSPs can be un-mixed spectrally to pure pixels having the same land cover classes in a fine classification map during the fusion process. Since it is difficult to produce such a land cover classification map using only MS and PAN images, a Digital Surface Model (DSM) derived from airborne Light Detection And Ranging data were employed in this study to facilitate the classification. In a novel fusion method proposed in this paper, MSPs near and across boundaries between vegetation and non-vegetation are identified using MS, PAN, and normalized Digital Surface Model (nDSM). The identified MSPs then are fused to pure pixels with respect to the corresponding land cover class in the classification map. In a test on WorldView-2 images over an urban area and the corresponding nDSM, the fused image generated by the proposed method was visually and quantitatively compared with fused images obtained using common image-fusion methods. The fused images generated by the proposed method yielded minimal spectral distortions and sharpened boundaries between vegetation and non-vegetation.