Detection of Flooded Areas using Machine Learning Techniques: Case Study of the Ljubljana Moor Floods in 2010

Lamovec Peter1*, Mikoš Matjaž2 and Oštir Krištof 1,3

Floods are one of the most frequent and severe natural disasters. In addition to causing a huge economic damage (to the human property) they cause a substantial loss of human lives. It is therefore crucial to use all available technology, including Earth observation, in their prevention and mitigation. Machine learning techniques can reduce the time needed for flood detection from optical satellite images. In the paper a case study of the September 2010 Ljubljana Moor floods is presented. RapidEye satellite images were combined with digital terrain model (DTM) and hydrological network to detect flooded areas. Two training samples and a number of different attributes were tested with points and segments. Afterwards different machine learning methods were used to find the most influential attributes which compose the model for classification of flooded areas. The model with the highest classification accuracy was chosen and used to define the floods in the whole observed area.

Full Text

Deformation Behavior of Reinforced Embankment Slopes under Seismic Excitation

Lin Yu-liang

Residual deformation caused by earthquake is main reason that results in seismic damage of embankments. In order to study deformation behaviors of different patterns of reinforced embankment slopes under seismic excitation, shaking table tests on three embankment slope models were carried out including unreinforced embankment slope, 2-layer reinforced embankment slope and 4-layer reinforced embankment slope. By performing Wenchuan excitations of different intensities, dynamic deformation responses of embankment slopes were investigated. Both lateral residual deformation and subsidence deformation behaviors were studied. And then, the deformation modes of embankment slopes were discussed. Results show that the baseline of dynamic deformation response fluctuates obviously and embankment slope shows large seismic residual deformation when the input peak acceleration is greater than 0.4 g. Seismic residual deformation increases significantly with the increasing of input peak acceleration, especially when the input peak acceleration is greater than 0.4 g. Reinforced embankment slope shows less seismic residual deformation than unreinforced embankment slope. The reinforcing effect is more obvious under strong seismic excitation. The deformation modes of different patterns of reinforced embankment slopes are all in seismic subsidence modes.

Full Text

Formation mechanism and stability analysis on the Hekou sugar-refinery landslide in Xinping County, Yunnan Province

Xu Wei1,2*, Hu Ruilin1 and Li Tiefeng2

Based on the field geological investigation and the slope mass strength test, formation mechanism of the Hekou sugar-refinery landslide which occurred in August 2007 was ascertained. The test result showed that the mass strength including cohesion and internal friction angle decreased with the increasing mass water content. The slide mass strength value was bigger than that of the slide zone at the same mass water content. The mass cohesion variation trend with the water content represented a relationship of exponential function. The mass internal friction angle variation with the mass water content represented linear relation. The three-year displacement monitoring data showed that the landslide was stable from 2009 to 2011. The monitoring is of great importance in future for early warning of the neighborhood safety.

Full Text

Real-time Response of Yuhuangge Landslide on the Reservoir Water Fluctuations in the Three Gorges Reservoir Area and Inversion on its Permeability Coefficient

Zhu Dapeng1, 2*, Yan Echuan3, Dong Shier1 and Huang Lu1

The detailed response analysis of landslides on the reservoir water fluctuations has important significance for the resurrection prediction of landslide. To accurately determine the permeability coefficient is the key to evaluate stability and seepage field of the landslide affected by the reservoir water. Taking the Yuhuangge accumulation landslide in Three Gorges Reservoir Area for example, the response of landslide including displacement, pore water pressure with changing reservoir water was analyzed and then the permeability coefficient was inversed based on the relationship of the real-time monitoring ground water and reservoir water by interpolation and finite element method.

Full Text

Catastrophe Progression Method on Forecast of Rock Burst

Jin Pei-Jian, 1,2* Wang En-Yuan,1,2 Huang Ning3 and Wang Si-Heng1,2

The catastrophe progression method combining the catastrophe theory and the fuzzy mathematics was introduced and raised as a new index on forecast of rock burst. A new model on forecast of rock burst danger was built based on comprehensive considering of the influence factors of rock burst danger. Then, following the non-dimensionalization and normalization processing, sample learning on the model was conducted according to coal mine field data. Catastrophe progressions of different burst danger grade were ascertained and the model was verified by existing data. The results show that this method could forecast exactly and has high accuracy and good practical value.

Full Text

The Post-Earthquake Economic Resilience on System Perspective

Xu Jiuping* and Liu Zhujun

Resilience is an important part of the study of natural hazard. Qualitative/quantitative study of environment, economy and society's different reaction after disaster risk, as well as different receptors' change process after disasters lay the foundation for better prevention, control and repair the loss caused by disaster. This paper based on the concept of resilience, defined post-earthquake economic resilience as disaster region economic system’s ability to adjust, adapt, recover and reconstruct successfully from earthquake attack to its pre-earthquake state and to transit the economy system quickly to a “better” equilibrium. The economic resilience is treated as an open dynamic system. Three factors are selected to characterize the static economic resilience and 4R dimensions to measure the dynamic economic resilience after earthquake.

Full Text

Ionospheric Precursor Detection for the M= 7.6, Costa Rica earthquake under large Geomagnetic Storm Activity: Tw0 Dimensional Principal Component Analysis

Jyh-Woei Lin

Two-dimensional principal component analysis (2DPCA) is performed to detect ionospheric total electron content (TEC) anomaly related to earthquake undergoing geomagnetic storm activity. The time period of examined TEC data are 5 days before the M=7.6 Costa Rica earthquake at 14: 42:07 on 05th September 2012 (UT). A TEC precursor of this earthquake with large principal eigenvalues6 (>0.5) is found nearby the epicenter during the time period from 20:05 to 20:10 on 03rd September 2012 (UT) with the duration time of at least 5 minutes by using 2DPA while TEC anomaly related to other smaller earthquakes is not detectable. TEC anomaly caused by the geomagnetic storm activity during the examined time period results in the small principal eigenvalues, therefore, the detection of TEC precursor is regardless of the geomagnetic storm activity.

Full Text

Study on Landfill Leachate Plumes with coupled Liquid-Solid Model

Liu Xiaoli,1* Hu Liming,1 Wang Enzhi1 and Xue Qiang2

There are potential public health and environmental implications associated with solid waste landfills. Deformation of landfill changes the void space (pores and fractures) in it and then affects the pathway of leachate flows. Inversely, the leachate flows movement induces seepage force and affects the deformation of the landfill. Therefore, the interaction of landfill and leachate flows in it is coupled process, which obviously can affect the transport and migration of contaminants in the landfill. In this paper, the coupled liquid-solid mathematical model was developed to describe the full process of liquid seepage, solid deformation and contaminant transport and migration. The model was adapted to simulate the process of landfill leachate polluting groundwater. The results make clear that the differences considering coupling process and uncoupling process are obvious and the effects of coupling cannot be ignored.

Full Text

GIS based Geosystem response modelling for landslide vulnerability mapping parts of Nilgiris, South India

Muthukumar M.

Landslide is one of the major disasters causing damages to man and his properties. Though this phenomenon was restricted to the selected mountain regions around the world in the past, now it is occurring in almost all the mountains. The Indian subcontinent too is not an exception to it. But, though the rainfall and the improper interaction of the man remained the major tutoring factors for landslides, the vulnerability of the terrain to landslides depends upon the various geosystems which only assign landslide vulnerability grades. This stimulated many Geoscient¬ists and the technocrats to study the geological parameters not only for landslide vulnerability mapping but also for mitigation and management. Various methods are being adopted for Landslide Vulnerability Mapping(LVM) resulting into differing types of outputs. So a new method was attempted for Nilgiri mountains, Western Ghats, South India by assigning weightages to various geosystem parame¬ters viz: Lithology, Lineament / fracture density, geomorphology, Slope, Land use / Land cover and their sub classes or subvariables, on the basis of their responses to landslides and integrated such weighted five geosystems to arrive finally at a landslide vulnerability map.

Full Text

Assessment of Forest Damage caused by Ice Storm based on MODIS Data- A Case Study of Jiangxi Province, China

Shi leilei1,2

Ice storm as a natural disaster influences the structure, function and dynamics of many forest ecosystems. Between 11 January and 5 February 2008, an unexpected ice storm hit southern China. The objective of this study is to evaluate the damage of subtropical forest from the severe ice storm in Jiangxi Province. The study quantified forest damage by using changes in Enhanced Vegetation Index (EVI) from moderate resolution imaging spectroradiometer (MODIS). The result showed a northeast-to-southwest pattern of forest damage in the basin. Results indicated that the ice storm caused EVI abruptly decreased from 0.28 to 0.23. The reduction on EVI by the ice storm is quite similar for mixed broadleaf-conifer forests (0.037) and evergreen broad-leaf forests (0.039) but is the highest for the coniferous forests (0.047). This result suggests that the artificial pure coniferous forests are most vulnerable to the ice storm disturbance.

Full Text

Analysis of Seepage Field in Multistage Tailing Dams Area based on Drainage System

Lu Ruili1, Wei Wei1* and Sun Dongpo2

In order to study seepage properties in areas of tailing pond, we simulated the integrated drainage system composed numerically by constructing a specific FEM element mesh (substructure) and generalizing boundary condition. The inversion values and measured values fit great which indicate reasonable applied model parameters. According to the numerical simulation for seepage in tailing pond, seepage gradient along the flow way changes strongly. Water head in slag accumulation areas changes gently, however, in the initial dam toe the water head falls faster. Drainage facilities will decrease the seepage gradient of dam toe area significantly. A combined drainage system of drainage pipes and longitude-horizontal blind ditch performs better than a single drainage system. Single horizontal pipe drainage system decreases the seepage gradient by 8.7%-12.6%, while combined drainage system can homogenize the phreatic surface behavior and in addition, it decreases the overall gradient of 8.6% to 26.3% more than only setting the horizontal drainage pipe. The seepage gradient is influenced by size and arrangement of drainage system significantly; in addition, it presents monotone increase relevance with the effective water head of multistage dam. To meet the stability requirement for tailing dam, the drainage system should be arranged to match the height of multistage dam.

Full Text