THREE DIMENSIONAL ANALYSIS OF SHALLOW LANDSLIDE ON VOLCANIC ROCKS IN TROPICAL REGION: APPLIED IN SINDANGKERTA, WEST BANDUNG REGENCY

• Main Author: DWI KARTIKO (NIM : 32010302), RENDY
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/30314
• Institution: Institut Teknologi Bandung

<p align="justify">Various studies have been conducted to understand landslide for its mechanism in nature. Landslide are often understood as movement process controlled by groundwater level rise to intersect with slope surface. However, if particular area has deeper local groundwater level without any intersection with slope surface in any particular peak rainy season, another mechanism should be explored. This mechanism will have to be explained by its lateral occurrence, depth of sliding surface, and also its time series development. <br /> <br /> Weathered volcanic rocks in southern Bandung usually have deeper groundwater. Many locations have difficulties in establishing water wells due to scarcity of groundwater. Steeper areas have deep groundwater table and high cost to build and maintain well compared than routing water from springs, although springs could be located far away from resident areas. Landslide can still be found throughout these locations, and periodically occurred in rainy season. <br /> <br /> Landslides within this vicinity dominantly have shallow depth of sliding surface. Landslide depth range from 30 cm to 2 m with materials consisted of mixture between residual soils with regolith of weathered volcanic rocks. Lateral distribution of these landslides is initial part of this study, especially within specific water catchment with weathered volcanic rocks at its base. <br /> <br /> Experimental approach are conducted to see possibility of shallow landslide occurrence within laboratory scale as a comparison with natural slope model. Landslide can occurred within well graded arenite quartz sand without clay content. Landslides occurred in second cycle artificial rainfall. First rain cycle intensity is close to sand material permeability coefficient. Rainfall is absorbed and flows under surface to the open area at the end of the model. Water pond is produced in front of the slope. Time delay then is given to lower the groundwater formed by first cycle rainfall until groundwater lowered to base of experimental apparatus. Second cycle rain is then conducted with higher intensity, several times of sand permeability coefficient. Landslide occurred just within minutes after administration of second cycle rainfall. Landslides that occurred within experimental slope are local, and concentrated on slope's toe, forming saturated transient zone due to excessive rain. <br /> <br /> Rainfall infiltration are modeled in one, two, and three dimension to see possible development of transient saturated zone. This zone could be formed by two dimension model approach. Meanwhile, three dimension analysis with finite depth basal boundary show that slope unstability timing occurred within range of actual historical landslides. The model with infinite depth boundary on the other hand predict that landslide will occurred in later time and not comparable with landslide inventory data. Therefore, determination of transient saturated zone's depth should be assessed because it control occurrence and timing of shallow landslide. This research gives depth range of transient saturated zone within 1 meter from surface in weathered Tuf of Beser Formation. <br /> <p align="justify">