A 25.5 km long access road has been constructed in a hilly area in Cisokan region. Several slope instabilities occurred during the rainy season, particularly at the end of heavy rainfall. A comprehensive study was performed to understand the characteristics of rainfall-induced slope instability. The study consisted of field observation, analyses of field and laboratory test data, and numerical analyses. The study revealed that in general there were two categories of slopes with instability characteristics: (i) slopes with a significant groundwater level increase during rainfall; (ii) slopes with an insignificant groundwater level increase during rainfall. In the first category, the slope instability was caused by a loss of matric suction and eventually the pore-water pressure, uw became positive as indicated by an increase of the groundwater level. In the second category, the slope instability was caused by a loss of matric suction without a rise in pore-water pressure, uw, to a positive magnitude. Two empirical curves of slope stability were developed as a preliminary guidance to assess slope stability during rainfall in the region.

Sigi Biromaru is an area prone to landslides. This study aims to apply the statistical method of Weight of Evidence (WoE) in landslide susceptibility mapping using Geographic Information Systems (GIS). The 265 landslides that occurred 2009-2019 were randomly divided into two groups, 70% of the data were used as training dataset for susceptibility modelling and 30% of the data were used as test data for validation of the susceptibility model. Twenty-one parameters were tested for their influence on landslides. Based on the Area Under Curve (AUC), parameters that significant controlling the landslides are slope gradient, elevation, aspect, flow direction, peak ground acceleration, clay content (<0,002 mm), land cover, terrain ruggedness index (TRI), river density, soil type, lineament density, lithology, rainfall and stream power index (SPI) respectively. The validation results show that the AUC success rate is 0,811 using the training dataset and AUC prediction rate is 0,756 using the test dataset. These results indicate that the WoE method produces a good landslide susceptibility map in the Sigi Biromaru area.

Landslide is one type of geological disasters that frequently occurs during the rainy season. Rainfall infiltration can cause soil saturation that increases the positive pore water pressure, disturbing the slope stability. Therefore, knowledge of future landslide-triggering rainfall is required for mitigation efforts and reducing the risk of landslide hazards. This paper presents slope-stability modeling in the Cililin area using the well-established infinite slope model called the transient rainfall infiltration and grid-based regional slope-stability (TRIGRS). The modeling used the rainfall data obtained from the statistical analysis of the maximum daily rainfall by using the Gumbel distribution. The present study applied six scenarios in the modeling. Scenario I is the initial condition without rainfall, showing the slope stability influenced by topography, slope, and soil characteristics. TRIGRS modeling involves rainfall infiltration in scenarios II, III, IV, V, and VI. The maximum rainfalls used in the modeling are 66, 76, 101, 120, and 132 mm/d, showing that rainfall infiltration affected the slope stability. The result indicates that rainfall triggered an increase and expansion of the area distributions critical to the slope stability.

No abstract

ABSTRAKPeta kerentanan gerakan tanah sangat diperlukan sebagai dasar dalam perencanaan tata ruang, pemanfaatan lahan dan mitigasi bencana. Kerentanan gerakan tanah dipengaruhi oleh beberapa faktor seperti kemiringan lereng, arah lereng, litologi, tutupan lahan, elevasi, curah hujan, kelurusan, percepatan gempabumi, kurvatur, arah aliran, jarak dari sungai, dan jalan. Dalam penelitian ini dikembangkan metode pemetaan kerentanan gerakan tanah menggunakan metode kombinasi logistic regression (LR) – weight of evidence (WoE). Metode gabungan ini diharapkan dapat menghasilkan metoda yang menggabungkan kelebihan dari masing-masing metode serta sekaligus mengatasi kelemahan masing-masing metode. Wilayah studi kasus penelitian adalah Takengon, salah satu wilayah di Provinsi Aceh yang rawan terhadap bencana gempabumi dan gerakan tanah. Data yang digunakan dalam penelitian ini adalah 251 kejadian gerakan tanah secara acak yang terjadi pada tahun 2000 hingga tahun 2016. Data tersebut dibagi menjadi dua kelompok data, 70% data digunakan sebagai set data analisis untuk penyusunan model dan 30% data digunakan sebagai set data validasi untuk pengujian model. Tahapan penelitian meliputi pembobotan dua belas parameter yang mempengaruhi kerentanan gerakan tanah dengan menggunakan metode WoE. Analisis kombinasi LR-WoE menggunakan parameter hasil pembobotan metode WoE dan kemudian di analisis menggunakan statistik LR. Selanjutnya melakukan analisis perbandingan hasil pemetaan kerentanan gerakan tanah melalui pengujian kurva Receiver Operating Characteristic (ROC). Hasil validasi dan pengujian model menunjukkan bahwa metode kombinasi LR-WoE mempunyai nilai AUC 0,853 yang lebih tinggi dibandingkan menggunakan metode WoE (AUC 0,830). Berdasarkan hasil penelitian ini disimpulkan bahwa metode kombinasi LR-WoE memberikan tingkat akurasi yang lebih baik dari metode WoE untuk pemetaan kerentanan gerakan tanah. Metode kombinasi LR-WoE dapat terus dikembangkan dan dapat diusulkan menjadi metode pemetaan gerakan tanah yang akurat, efektif dan ekonomis. Kata kunci: Kerentanan gerakan tanah, Logistic Regression, Takengon, Weight of Evidence. ABSTRACTLandslide susceptibility map is an imperative basic tool for land use application, spatial planning and disaster mitigation. The susceptibility of landslide is influenced by factors such as slope, slope aspect, lithology, land cover, elevation, rainfall, linemeant, peak ground acceleration, curvature, flow direction, distance from rivers, and roads. In this research, a combined method of weight of evidence (WoE) and logistic regression (LR) was applied to assessed its advantages and overcome the limitation of each method. Takengon is an area prone to earthquake disaster and landslide. The 251 landslides from 2000 until 2016 were randomly divided into two groups of modelling/training data (70%) and validation/test data sets (30%).  The research stages include weighting of twelve parameters that affect the susceptibility of landslide using the WoE method. The combination LR-WoE analysis uses the weighted parameter of the WoE method and then analyzed using LR statistics. The validation results using Receiver Operating Characteristic (ROC) curve showed that the LR-WoE method had a better accuracy than the WoE methods, with values of 0,890 higher than that of the WoE method 0,830 prediction. Therefore, it is concluded that the combined method of LR and WoE can provide a promising level of accuracy for landslide susceptibility mapping. Combined LR-WoE method can be developed and proposed to be an accurate, effective and economical method of mapping the landslides susceptibility map. Keywords: Kerentanan gerakan tanah, Logistic Regression, Takengon, Weight of Evidence.

Indonesia is a risk prone country from landslide disasters. The occurrence of landslides followed by the debris flow often make a lot of casualties and very terrible destructions. Accordingly, debris flow modeling of some Java landslides has been conducted in this study to determine run-out distribution characteristics of the debris materials. The concept of debris flow modeling is based on the equations of momentum, continuation, riverbed deformation, erosion/deposition and riverbed shearing stress. From this modeling, it has been indicated the best fit simulation results. In this case, run-out distributions of Pacet landslide at Mojokerto, December 12, 2002 has been properly modeled with a scenario of 0.4 viscosity value for around 5 minutes. At Sijeruk landslide, Banjarnegara, January 4, 2006 run-out distributions have been modeled using viscosity value of 0.45 for 14 minutes 51 seconds. Meanwhile, the modeling for Tenjolaya landslide at southern Bandung, February 23, 2010 with viscosity value of 0.38 shows time needed for debris materials to reach depositional area is estimated for around 12 minutes 37 seconds. From this study, debris flow modeling had given better understandings regarding flow track, velocity and distribution of debris materials from some debris flows in Java.

Indonesia is one of the world’s most natural disaster-prone country from landslides. These landslides mostly occur at areas having steep to very steep slopes, intensive weathering processes and high to very high rainfall intensity. Generally, debris flows referred to mudflows or lahars. These types of landslide are common type of fast-moving landslide. In this regard, the occurrence of several landslides followed by the debris flow often make a lot of casualties and very terrible destructions in some areas of Java Island, Indonesia. For example, Jemblung Landslide on December 12, 2014 has 139 causalities. Accordingly, some debris flow modelling have been conducted to determine run-out distribution characteristics of debris materials at the depositional areas. The concept of debris flow modeling is based on the equations of momentum, continuation, riverbed deformation and erosion/deposition and riverbed shearing stress. From the modeling of Pasir Panjang landslide case, the viscosity value of 0.38 indicated the best fit simulation result. The flowing material of this landslide case has very long distance, 2.3 km in approximate. It occurs because 275,295 m3 volume material which flowing is supported with 25.58 km/hour of maximum flow rate and relatively straight flow track in 35°-45° of slope angle.

Landslide susceptibility modeling using neural network (ANN) are applied to semi detailed volcanic-sedimentary water catchment. Annually landslide occurred in catchment area frequently in unconsolidated and weathered material combined with uncertainty in rainfall pattern that complicated landslide occurrence. Data used for analysis including landslide inventory, geology, digital elevation related data, distance to stream, and several other available data. Results show that machine learning method yield fair result data based on evaluation on Area under Curve (AUC). Thus, it can be suggested that machine learning methods for landslide susceptibility model could still be develop to produce robust prediction model with different characterization of parameter data and machine learning parameters.