Spatial distribution of block falls using volumetric GIS–decision-tree models

Block falls are considered a significant aspect of surficial instability contributing to losses in land and socio-economic aspects through their damaging effects to natural and human environments. This paper predicts and maps the geographic distribution and volumes of block falls in central Lebanon using remote sensing, geographic information systems (GIS) and decision-tree modeling (un-pruned and pruned trees). Eleven terrain parameters (lithology, proximity to fault line, karst type, soil type, distance to drainage line, elevation, slope gradient, slope aspect, slope curvature, land cover/use, and proximity to roads) were generated to statistically explain the occurrence of block falls. The latter were discriminated using SPOT4 satellite imageries, and their dimensions were determined during field surveys. The un-pruned tree model based on all considered parameters explained 86% of the variability in field block fall measurements. Once pruned, it classifies 50% in block falls’ volumes by selecting just four parameters (lithology, slope gradient, soil type, and land cover/use). Both tree models (un-pruned and pruned) were converted to quantitative 1:50,000 block falls’ maps with different classes; starting from Nil (no block falls) to more than 4000 m³. These maps are fairly matching with coincidence value equal to 45%; however, both can be used to prioritize the choice of specific zones for further measurement and modeling, as well as for land-use management. The proposed tree models are relatively simple, and may also be applied to other areas (i.e. the choice of un-pruned or pruned model is related to the availability of terrain parameters in a given area).     

Soil and bedrock distribution estimated from gully form and frequency: A GIS-based decision-tree model for Lebanon

Torrential rainfall and relatively sparse vegetation in the Mediterranean region result in the development of gully systems and land degradation, notably on lands with specific types of soil and bedrock. This paper proposes a decision-tree model to predict the distribution of soil and bedrock susceptible to gully erosion (white Rendzinas and marly rocks) from the form and frequency of gullies. The study area is located in Lebanon and the model is linked to GIS. V-fold cross-validation of the pruned model indicates that gully features including cross-section size and shape, network frequency, types of meandering, and catchment area can explain 80% of variance in soil/rock properties. The overall accuracy of the soil/rock map was estimated to be ca. 87%. The proposed model is relatively simple, and may also be applied to other areas. It is particularly useful when information about soil and rock obtained from conventional field surveys is limited.     

Decision-tree analysis on optimal release of reservoir storage under typhoon warnings

The wet and dry seasons are distinctive in Taiwan as the amount of precipitation in wet seasons accounts for over three-fourth of the total rainfall. And the water-resources management relies pretty much on the rainfall brought in by typhoons as it accounts for a significant portion of the precipitation during wet seasons. Furthermore, as the storage of reservoirs is limited due to topographical factors, the management of typhoon rainfall has always been an important issue in Taiwan. The technique of decision-tree analysis is applied in this article to determine the optimal reservoir release in advance upon the issuance of a typhoon warning by the Central Weather Bureau (CWB), and the proposed methodology may provide solution to the trade-off judgment of reservoir operations between flood control and water supply according to economic efficiency. In this article, the economic loss functions of flooding damage and water-supply shortage are assumed in linear and nonlinear conditions, and the respective expected optimal releases based on the predicted precipitation as issued by CWB are derived. The proposed methodology has been applied to the Shihmen Reservoir System, and the capabilities of the model as an aid to real-time decision-making as well as the evaluation of the economic worth of forecasts is presented.     

Comparison of landslide susceptibility based on a decision-tree model and actual landslide occurrence: The Akaishi Mountains, Japan

This paper proposes a statistical decision-tree model to analyze landslide susceptibility in a wide area of the Akaishi Mountains, Japan. The objectives of this study were to validate the decision-tree model by comparing landslide susceptibility and actual landslide occurrence, and to reveal the relationships among landslide occurrence, topography, and geology. Landslide susceptibility was examined through ensemble learning with a decision tree. Decision trees are advantageous in that estimation processes and order of important explanatory variables are explicitly represented by the tree structures. Topographic characteristics (elevation, slope angle, profile curvature, plan curvature, and dissection and undissection height) and geological data were used as the explanatory variables. These topographic characteristics were calculated from digital elevation models (DEMs). The objective variables were landslide occurrence and reactivation data between 1992 and 2002 that were depicted by satellite image analysis. Landslide susceptibility was validated by comparing actual data on landslides that occurred and reactivated after the model was constructed (between 2002 and 2004).This study revealed that, from 2002 to 2004, landslides tended to occur and reactivate in catchments with high landslide susceptibility. The landslide susceptibility map thus depicts the actual landslide occurrence and reactivation in the Akaishi Mountains. This result indicates that the decision-tree model has appropriate accuracy for estimating the probabilities of future landslides. The tree structure indicates that landslides occurred and reactivated frequently in the catchments that had an average slope angle exceeding ca. 29° and a mode of slope angle exceeding 33°, which agree well with previous studies. A decision tree also quantitatively expresses important explanatory variables at the higher order of the tree structure.     

岩石流变力学试验数据挖掘研究

介绍MSAnalysisServices的数据挖掘过程 ,分析决策树算法的基本原理。对泥板岩岩石力学的一组流变试验数据进行数据挖掘分析 ,所得结论反应出实验过程中影响岩石应变的重要因素以及它们影响力度的差异