Ming Chee Wu

Estimating landslide-induced riverbed roughness variation by using LiDAR data

With advancements in the efficiency and accuracy of investigation techniques and equipment, remote sensing technologies have been widely used to investigate river conditions. Quantifying the morphology along a river channel was difficult before airborne laser altimetry technology, light detection and ranging (LiDAR), was introduced, facilitating the collection of high-resolution, highly accurate topographical data. This study adopted airborne LiDAR data for analyzing and recognizing riverbed morphology. The roughness index was defined as the standard deviation of a residual topography. A variable moving-window was used to derive a smoothed digital elevation model (DEM). According to the roughness index, the residual topography was the difference between the original and smoothed DEMs. Roughness data derived from different reaches of a predisaster riverbed were compared with data derived from a postdisaster riverbed. The experimental results showed that the upper reaches exhibited higher roughness values than did the lower reaches. Thus, the relief of the postdisaster riverbed surface was near the derived smoothed relief. Such characteristics were reflected in the major differences evaluated through slope measurements in the riverbed morphological analysis; the position of the peak value changed after the disaster. An integrated plane-wise fluvial circumstance of a river watershed area was rapidly and accurately constructed, and this study concluded that these remote sensing techniques are vital in facilitating traditional surveys for regional investigations.

National airborne LiDAR mapping and examples for applications in deep-seated landslides in Taiwan

For the purposes of geohazard study, a national airborne LiDAR mapping program spanning 2010 to 2015 was launched with the aim of simultaneously capturing the territory (36,000km2) by airborne LiDAR and digital imagery. The results include very detailed digital elevation models (DEM) and digital surface models (DSM) of 1m grid and digital aerial photograph of 50 cm grid, as well as an inventory of the geological disastrous features with the acquired LiDAR data and images. In total, 400 deep-seated landslides are obtained in a preliminary interpretation. In this paper, an example in Chasan Tribal Settlement of the large-scaled landslides is selected for validation to further inspected with other geological investigation means. The results of this paper demonstrate the merit of the national airborne LiDAR survey and the effectiveness of simple manual interpretation approach for a census of deep-seated landslides with LiDAR-derived images of various types of enhancement such as shaded relief.

Satellite observation on coastal change in the Western Hunching Peninsula, Southern Taiwan

This study integrates multi-satellite images for investigating coastal erosion in the Western Hen-Chung peninsula, southern Taiwan. Coastal erosion was observed in the estuaries with erosion rate of 1.2 m/yr. Sediment slightly accumulated in the southern part the study area while a semi-closed bay was filed by inappropriate artificial construction. In addition, coastline erosion occurred in 2009 after typhoon passage, but it gradually silted through time. However, human activities and coastal engineering were mainly factors that change the coastal environment rather than nature weather dose. It is suggested that human actives and artificial constructions should be prohibited for the coastal zone development.

A comparative study on sorption and diffusion of Cs in crushed argillite and granite investigated in batch and through-diffusion experiment

The batch and through-diffusion experiments in this study were conducted and compared in order to investigate the sorption and diffusion of cesium (Cs) for two potential host rocks in Taiwan: argillite from Taitung and granite from Kinmen Island, with the purpose of establishing a reliable safety-performance assessment methodology for the final disposal of low level radioactive waste. The results of Cs mapping by scanning electron microscope equipping by energy dispersive spectrometer (SEM–EDS) showed that the distribution of Cs on argillite and granite were enriched in illite and biotite, respectively. In addition, it showed that higher sorption capacities were found for argillite than granite; due to the clay mineral content (illite) in the argillite. Experiments for diffusion of Cs is agreement to the values estimated for the diffusive results (Da) of Cs in argillite were revealed to be lower than those of granite.

Stream morphometry analysis by airborne laser altimetry

The river characteristics have played an important role for hydrological models. Sediments supplied from landslides may affect the river channel morphological changes in different reaches or magnitude; namely, the riverbed morphology can be related to disaster events. Measurements of riverbed morphology are made using high-resolution LiDAR DEM in this study to understand the relationship. Quantifying the morphology along a river channel has been proven difficult; till the airborne laser altimetry technology was implemented in the survey. In this study, the remote sensing techniques are employed to assist the studies on stream roughness distribution and morphological investigation. The results are proven to be valuable for hydrological and engineering applications.

Application of non-linear heterogeneity-based isotherm models for characterizing sorption of Cs and Se on mudrocks

It plays a very important role for characterizing sorption behaviors of cesium (Cs) and selenium (Se) on Taiwanese mudrocks to retard the migration of radionuclides from a radioactive waste repository. In this study, two non-linear heterogeneity-based isotherms, Langmuir–Freundlich model (LF) and generalized-Freundlich model (GF), were applied for the evaluation of the sorption characteristics of Cs and Se on Taiwanese mudrocks. The batch sorption experiments were carried out and the experimental data were simulated by LF and GF heterogeneity-based isotherm models. In addition, the results showed that both of the LF and GF models could fit the experimental data more perfectly than the Langmuir one. The heterogeneity of sorption behaviors for Cs and Se could be well characterized by the LF and GF models from the root mean square error calculation and plot of the affinity spectrum. The results demonstrated that the sorption mechanism of Cs and Se on mudrock is quite different and Cs sorption is more heterogeneous and complicated than that in Se.

Analysis of spatial characteristics for landslides vegetation restoration monitoring by LiDAR surface roughness data and multispectrum imagery

Vegetative cover can contribute to improving the stability of steep slopes by reducing erosion; the vegetation growing condition may reflect the landslides spatial characteristics. The landslides and vegetation restoration evaluation in this study was implement by high resolution air-photo, SPOT-5 satellite image and air-borne LiDAR data then acquired for calculated vegetation restoration rate. The vegetation restoration results shown different spatially distribution on landslide area, the vegetation restoration rate where nearby riverbank is less than where in mountain area; the growing rate also shows the same appearance, the mean value with 3.2 meter in mountain area and the riverbank area with 1.1 meter restoration rate. It indicated that the vegetation restoration of mountain area appear well growing condition than in riverbank area. The steep slope appears with low vegetation restoration rate, it may have relations with higher frequency of landslides and lower disturbed by human in this area.

Landslides volumetric estimation by using LiDAR data and Sediment Production Rate method

LiDAR DEM (Digital Elevation Model) data of two different periods, before and after a heavy rain event, were acquired to specify the exact boundary of the mass been moved, and to calculate the variations of ground elevation by using the DEM values for each photometric cellular element. With the modified Sediment Production Rate (SPR) equation, volumes of the mass wasted were as well being estimated. In addition, with the aids of spatial analysis tools of Geographic Information System (GIS), details of the landslide were well enhanced; such that, the magnitude and the peripheral extent of the landslide were characterized.

Automatic image classification of landslides improved with terrain roughness indices in various kernel sizes

Using spectral-only information for landslides classification is usually confusing with houses, roads, and other bare lands because these ground features have similar spectral patterns on images. The terrain roughness can be measured by significant wavelengths; some studies have linked the relationships between terrain roughness and the landslide by using numerical analyses of topography data. In this study, airborne LiDAR data of 1m grid are used to explore the possibility of improvement of landslide classification, the LiDAR-derived data include DEM slope and terrain roughness indices including diversity, dominance and relative richness with different grid size data are used to improvement classification accuracy. The improvement of accuracy when including DEM slope is 22% in producers accuracy and 27% in users accuracy. The accuracy of diversity, dominance and relative richness indices all are improved when kernel sizes enlarge in Maximum Likelihood and Mahalanobis Distance algorithms.

Coastal wetland monitoring and evaluation by integrated multitemporal remote sensing images and historical maps

Monitoring and evaluation are the base work for environment protection and management. Maps are the most directly and lucid measure to represent and record geographic environment. As a result, By overlaying maps with different ages and processes of geographic change can be quantitative analyzed and should provides long-term scale, on the other hand, satellite image can be use for short-term investigation to know the wetland change which cause by hazard event. It is shown in this study that local spatial autocorrelation should provide objective classification in training stage, and the spatial autocorrelation also can enhance the pattern of vegetation and water body on image.