Landslide susceptibility assessment using GIS has been done for part of Uttarakhand region of Himalaya (India) with the objective of comparing the predictive capability of three different machine learning methods, namely sequential minimal optimization-based support vector machines (SMOSVM), vote feature intervals (VFI), and logistic regression (LR) for spatial prediction of landslide occurrence. Out of these three methods, the SMOSVM and VFI are state-of-the-art methods for binary classification problems but have not been applied for landslide prediction, whereas the LR is known as a popular method for landslide susceptibility assessment. In the study, a total of 430 historical landslide polygons and 11 landslide affecting factors such as slope angle, slope aspect, elevation, curvature, lithology, soil, land cover, distance to roads, distance to rivers, distance to lineaments, and rainfall were selected for landslide analysis. For validation and comparison, statistical index-based methods and the receiver operating characteristic curve have been used. Analysis results show that all these models have good performance for landslide spatial prediction but the SMOSVM model has the highest predictive capability, followed by the VFI model, and the LR model, respectively. Thus, SMOSVM is a better model for landslide prediction and can be used for landslide susceptibility mapping of landslide-prone areas. 相似文献
In order to optimize ship navigation in the macrotidal Gironde Estuary, a recent project funded by the port of Bordeaux aims at better understand and forecast hydrodynamic and fine sediment transport within the estuary. In the framework of this project, a two-dimensional hydro-sedimentary model is built. The model includes hydrodynamic forcings, mixed-sediment transport, and consolidation processes. The harmonic analysis of the astronomical tides reveals a strong distortion of the tidal wave inducing the growth of overtide constituents and the non-significant effect of tide-surge interactions in annual-scale prediction. Depending on hydrological conditions, river discharge can considerably alter the model accuracy due to the migration of the turbidity maximum zone modifying the bottom roughness. Comparison with measurements shows the ability of the model to reproduce suspended-sediment concentrations in the central Estuary. Sensitivity of the model to sediment features has also been discussed in regard of suspended-sediment concentrations and fluid mud deposits. The model will be further coupled with ship squat predictions and a morphodynamic model. 相似文献
Existing research on DEM vertical accuracy assessment uses mainly statistical methods, in particular variance and RMSE which are both based on the error propagation theory in statistics. This article demonstrates that error propagation theory is not applicable because the critical assumption behind it cannot be satisfied. In fact, the non‐random, non‐normal, and non‐stationary nature of DEM error makes it very challenging to apply statistical methods. This article presents approximation theory as a new methodology and illustrates its application to DEMs created by linear interpolation using contour lines as the source data. Applying approximation theory, a DEM's accuracy is determined by the largest error of any point (not samples) in the entire study area. The error at a point is bounded by max(|δnode|+M2h2/8) where |δnode| is the error in the source data used to interpolate the point, M2 is the maximum norm of the second‐order derivative which can be interpreted as curvature, and h is the length of the line on which linear interpolation is conducted. The article explains how to compute each term and illustrates how this new methodology based on approximation theory effectively facilitates DEM accuracy assessment and quality control. 相似文献