Keywords: Image Fusion; ChingBai Mountain; IKONOS; Monitoring and structural significance of ground deformations at

G. Ventura*a, Vilardo G. b, V. Sepea

a Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Roma, Italy, 00143;

bOsservatorio Vesuviano- Istituto Nazionale di Geofisica e Vulcanologia, Via Diocleziano 328,

Napoli, Italy, 80124

ABSTRACT

Cities are growing around active volcanoes. Campi Flegrei supervolcano (CF, Italy) is a nested structure formed during two main collapses associated with two caldera-forming eruptions at 39 ka and 15 ka. The last event occurred in AD 1538 (Monte Nuovo volcano). CF hosts 350,000 people and two main uplift phases were recorded in 1968–1972 and 1982–1984 with deformations of about 2 m. The town of Pozzuoli was partially evacuated during the last crisis. Subsequent minor deformations (subsidence and uplift), seismicity, and diffuse CO2 degassing concentrate in the central

part of the caldera. Here, we apply the Permanent Scatterers Synthetic Aperture Radar Interferometry (PS-InSAR) to investigate the ground deformations of CF by using data acquired from 1993 to 2007 from ascending and descending tracks by ERS-1 ,ERS-2 and RADARSAT satellites. Deformation maps identify a subsidence interrupted by micro-uplift episodes. These maps are combined with digitized topographic, geological (faults and landslides), seismic, and urbanization data. The merged information allow us to identify the areas involved in the deformation and the volcanotectonic

structures activated during the uplift and subsidence episodes. We propose a structural-volcanological model for the unrest episodes. Data indicate that uplift episodes, which are associated to seismicity, are followed by subsidence episodes accommodated by pre-existing faults. The urbanized areas subjected to the higher deformations and shaking are also identified and mapped. A multi-hazard zonation including landslides is also provided. The approach used here may be utilized to (a) recognize the tectonic and/or volcanic structures activated during ground deformations, (b) to investigate structural models, (c) to evaluate and map multi-risk zonations, and (c) monitor other volcanic areas or nonvolcanic zones subjected to gravity instability or tectonics.

Xizhang Gao*, Tao Zhang, Chenghu Zhou, Xiaomei Yang

Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Science,

11 Datun Road, Beijing, China, 100101

Considering large, complex and multi-dimensional marine data, the general representation and analysis methods based on 2D GIS can not meet the requirements of Marine Research. For multi-dimensional marine sampling data of sea water properties (ARGO sampling data, ship measured data, etc.), three dimensional interpolation and visual analysis methods can be used to reveal the distribution of sea water properties (such as temperature and salinity), and it is an effective way

to detect regional anomaly of marine phenomenon and to integrate expert knowledge. In order to handle three dimensional sampling data of sea water properties, this paper developed a three dimensional volume visual analysis GIS component with OpenGL and C++. Aim at varieties of marine data, different visual analysis methods are designed and presented, including two-dimensional, three-dimensional, three dimensional volume visualization, three dimensional slicing and cutting analysis and three-dimensional contour surface analysis, intending to aid marine scholars’ research.

We conduct an example test in this paper, and ship measured data about temperature and salinity from 1958 to 1959 of a Chinese ocean is used in the test. The test results show that the approach proposed and component developed in this paper can be effectively used for visual analysis in marine research and aid marine scholars for ocean study.

Song Cuiyu*a Yang Fengjiea Wang Guoqinga Long Xuanyaob

aCollege of Geological Science and Technology, Shandong University of Science and Technology,

Qingdao, 266510

bPatent Examination Cooperation Center of Sipo, Beijing, 100190

ABSTRACT

Studies on rotation invariant texture in remote sensing image proceeding are relatively rare. Local Binary Pattern (LBP) is a relatively new rotation invariant texture measure which is theoretically simply but powerful. The existed LBP operator only measures the spatial variability of gray-level values within the single-band image. When used in remote sensing multispectral image processing, the multivariate version of LBP operator should be considered in order to characterize the multivariate spatial autocorrelation among those bands. In this paper, a multivariate rotation invariant

LBP operator was proposed to calculate multivariate texture for multispectral remote sensing image. The derived multivariate texture image was combined with the spectral data in image classification to evaluate the performance of the proposed texture measure. The result was compared to classifications using spectral data alone and plus single-band LBP texture images. Experiments demonstrate that compared to spectral classification, the classification overall accuracy can be significantly improved when the proposed multivariate texture is included. The results also show that the multivariate

rotation invariant texture results show high overall accuracy, which is comparable with the best result from the existed single-band LBP texture. Moreover, it avoids the band selection procedure which makes it more convenient to be used.

Qingyuan LI*a，Jiarong ZENGb, Xu CHENc , Gang LIAOd

aChinese Academy of Survering & Mapping, Beitaiping Road 16, Haidian District, Beijing, P.R.C.

+86-13511013028; bChinese University of Technolegy and Mining, Beijing, Ding 11, Haidian

District, Beijing P.R.C; cLiaoning Technical University, Zhonghua Road 47, Fuxin, Liaoning

Provence, P.R.C; dPeking University, Yiheyuan Road 5, Haidian District, Beijing, P.R.C

As a supplement to 3D vector model and 3D voxel model, this paper proposes a new data model ---- volume function model, with partitioned fitted volume functions to express ununiformity inside of solid feature in 3DGIS. Firstly, the paper gives the necessity and possibility of expression ununiformity inside of solid feature in 3D GIS. Secondly, the paper propose a concept of volume function model, i.e. from 3D discrete sampling points data ( values of the interestproperty ), fitting volume function to express the ununiformity interest-property field inside of 3D solid features. Thirdly,

the paper gives a theory frame of the volume function model. It includes 3 rules about the volume function data model: 1)the space should be sub-divided into many sub-volume to fitting volume function for every sub-volume, 2) volume functions and their derivative function should continuous on the interface between sub-volume inside of one layer rock, and 3) discontinuous on the interface between different layer. Fourthly, the paper gives some primary research work about one volume function fitting method based on tetrahedron volume coordinate. The paper sums up eight characteristics about the tetrahedron volume coordinate. A method of fitting on Bernstein spline-base is advised. A compare of Volume function model with vector model and voxel-base model is given in the paper.

Keywords: Volume function data model, Tetrahedron volume coordinate, Mine and geology, T3DGIS

A Dynamic Process-Based Cadastral Event-Cascaded Spatio-Temporal Data Model

ZHANG Feng^*a,b, TENG Longmei^a,c, LIU Nan^a, LIU Ren-yi^b

The key point of TGIS is to design an appropriate data model that can illustrate and describe changes of features. In the view of the characteristics of cadastral spatio-temporal database that is of frequent changes, large amount and abundant source. This paper presents a dynamic cadastral event-cascaded spatio-temporal data model process-based.

First, inspect the changes of cadastre features. Point out the deficiencies of the spatio-temporal data model based on event or sequence of events. It's difficult to describe the relations of spatio-temporal objects and causal relationships of events with these models. Furthermore, it is inefficient to realize the query and redivivus of the history information based on the model of base state with amendments though the modal can store change data with low redundancy.

Analyzing the parcel change characters of the object-oriented TGIS, a Process-Based Spatio-Temporal Data Model (PBSTDM) is proposed here. The model can retain the evolution characteristics of spatio-temporal objects and reflect the cause and effect events involved in the process. An organization frame work of PBSTDM is proposed here. It provides a method for designing and building spatio-temporal database based on dynamic multilevel base state with amendments.

Finally, a software package of cadastre change management system - ReGIS based on above model and methods are presented. The application proclaims that PBSTDM can facilitate the data access, advance the efficiency of spatio-temporal data processing, store and manage spatio-temporal data with low redundancy and trace history information efficiently.

Keywords: spatio-temporal data model; event; spatio-temporal process; cadastre


RS & GIS Based Assessment of Soil Erosion in Disaster Region of Wenchuan Earthquake

Hongjuan Zeng*^a, Shengtian Yang ^a, Donghai Zheng^b, Yunfei Gao ^a, Yu Zhang^b

Email: zhj_eve@126.com

A catastrophic earthquake struck the Sichuan Province in southwest China on May 12, 2008. The ecological environment of the disaster –stricken region (about 220 thousand km2) now has become very fragile, but that area plays an important role in the ecological security of Middle and Lower Reaches of Yangtze River. In this study, we develop a method for quick assessment of soil erosion in the quake-hit area by integrating expert knowledge and spatial information on erosion intensity influencing factors, and analysis the changes of soil erosion intensity from pre-quake to post-quake. Firstly, we construct a classification standard for soil erosion situation of these region according to the related expert knowledge, where the soil erosion intensity was influenced by erosion factors and the formula can be expressed as follow: E=f（L，S，C，SP）, where E is the soil erosion intensity (which is classified into six levels: non-erosion, light, moderate, strong, severe and very severe); L is the land-use type; S is the ground slope; C is the vegetation coverage and SP is special erosion types including landslide, rock fall mass, mud-rock flow and broken buildings caused by the earthquake. The soil erosion intensity influencing factors are extracted by integrating RS and GIS techniques. The land-use map is extracted from CEBERS images using a man-computer interactive interpretation method; the vegetation coverage and special erosion types are extracted from multi-spectrum bands of Beijing-1 Micro-satellite before and after the earthquake, a Digital Elevation Model (DEM) on a scale of 1: 50000 is used to compute slope gradient. Then the classification standard and spatial information are coupled and used to estimate the erosion intensity. The results showed that: 1) the coupling model of knowledge and spatial information can make a quick assessment of soil erosion effectively, and enjoys a great advantage in regional soil erosion monitoring and assessment; 2) the earthquake has deteriorated the soil erosion widely. Compared to the publicized result of remote sensing investigation of soil erosion in 2000, the area of soil erosion after the earthquake has increased by 19.5 thousand km2, or approximately 22.2% of the soil erosion area before the earthquake; 3) the area of soil erosion intensities had increased except the slight degree. The percentage of light, moderate, strong, severe and very severe in the total erosion area is 31.0%, 40.4%, 17.2%, 6.1%, 5.3% respectively. These results are anticipated to be helpful for decision making and policy planning for recovery and restruction in the disaster region. It also provides a quick methodology for quantifying soil erosion distributions at a regional scale.

Gang Cheng*a, Zengzhang Guo a, Jianjun Gao a

aSchool of Surveying and Land Information Engineering, Henan Polytechnic University, No.2001,

Century Avenue Jiaozuo, Henan, China, 454003

1. Purpose

Historical relic records a specific political, economic and cultural development of the process of a country and nation, they are not eternal life, but also non-renewable, and how to make long-term preservation of human treasure has become a global issue. Human must pay more attention to the scientific protection and development of existing heritage, and also fully exploit and reflect their cultural connotations. In this paper, three-dimensional scanning technology combined with the

traditional measurement method was applied into the fast, high-precision three-dimensional modeling and reverse engineering of large-scale cultural landscape, making the digital cultural heritage to the permanent preservation and doing good for the state monitoring and analysis, non-contact repair, the virtual display of the cultural heritage, so as to further establish a digital museum. 2．Research Program and Methods GPS technology was used to obtain accurate three-dimensional coordinates of high-level control points; high-precision total stations were employed for the establishment of local control network; while the three-dimensional laser scanner scanned the cultural landscape and access to laser point cloud data. The post-processing software undertook the processing of point cloud de-noising, the establishment of precise three-dimensional model and line model, and the data output for other software. The three-dimensional modeling process is shown as follows: 1) Data acquisition The local control by use of high-precision GPS and total station combined three-dimensional model with the WGS-84 or local coordinate systems. Data acquisition processes were carried out by stations connected by common targets. Cyclone was connected to Leica scanstation2 and used to control it.

2) Point cloud registerAll the original point cloud data collected by stations can be de-noised by the Cyclone, and the software provides thecomplete set of tools for aligning point clouds captured from different scanning positions, quickly and accurately.3) Data extraction and analysis

Soil Salinization in Angulinao wetland Based on Multi-remote Sensing Data

Lina Zhang*a, Jicang Wu a, Xiao Chengb

aDepartment of Surveying and Geomatics,Tongji University, Shanghai, China, 200092;

bBeijing Normal University, Beijing, China, 100875

ABSTRACT

Wetlands are natural reservoirs and natural species libraries, are important parts of natural resources and ecological environment. Angulinao Lake wetland is located in Hebei plateau region in the upper reaches of Beijing and Tianjin, only 180km straight-line distance from the capital. It is very important component in the capital ecosphere. Angulinao Lake’s water area is up to 6667 hm2 at most, and the lake Prairie area is 15333 hm2.There are lakes, rivers, marshes and other types of wetlands. The Angulinao Lake wetland are typical wetlands in Hebei province. In recent years, as a result of climate drought, over-cultivation, over-exploitation of groundwater and other reasons, the vegetation in this region is severely damaged; soil desertification and salinization are serious;the rapid rate of wetland degradation and valuable wetland resources are in danger of disappearing, which have seriously affected the regional ecological security environment. Since remote sensing has much advantage in obtaining information of the land surface，it has been more and more often

used in the land monitoring. In this paper, there are three frames of remote sensing data including one ETM data in 2001，one CBERS data in 2004 and one AVNIR data in 2007 of the study area.Using the method of decision tree,wetlands, water, vegetation, bare land, light salinity and heavy salinization are successfully separated from the surface features and redundant information.By calculating the area of the five classis, we found wetlands in Angulinao Lake have disappeared in recent years, and the vegetation cover have been in rapid declination,while the area and degree of the bare land and salinizition soil have gradually increased.There is a serious problem of soil degradation in

Angulinao Lake , which is coincident with the local situation.The reasons of soil salinization in this area have been analyzed..In Angulinao Lake area, the accumulation of soil salinization are caused by the following three factors.Arid climate is the major factor of the wetlands degradation and soil salinization in Angulinao Lake.This area belong to the semi-arid continental monsoon climate with the average temperature 2.8 ° C over the years, the multi-year average precipitation 401.6mm/yr and the evaporation is 1844.8mm/yr, evaporation is much larger than precipitation.Hydrologic geology provide conditions for the occurrence of soil salinity, with the gradual evaporation of water and gradually

concentration of salt solution, salts are separated and precipitated in different parts of the terrain. Human factors exacerbated the wetlands degradation, the construction of a large number of reservoirs, the excessive use of groundwater and the unreasonable land use mainly contribute to the increase of the wetlands degradation and soil salinization.

Wetland characterization and classification using polarimetric RADARSAT-2 data

Jingjuan Liao a, Qing Wang a,b

aCenter for Earth Observation and Digital Earth, Chinese Academy of Sciences, Beijing, 100101;

bGraduate University of Chinese Academy of Sciences, Beijing, 100049

Wetlands play a key role in regional and global environments and are critically linked to many major issues such as climate change, water quality, the hydrological and carbon cycles, and wildlife habitat and biodiversity. Mapping wetlands and monitoring their change are a long-term task. Remote Sensing characters with macrocosm, dynamics, quantity, and comparability will largely favor wetland research, especially radar remote sensing , which is not limited by climate conditions, has been proved an effective tool in wetland monitoring. In this paper, the unique polarimetric data of RADARSAT-2 is investigated for wetland classification. The target decomposition is used for optimum characterization