Our research project titled “Integrated study on mitigation of multimodal disasters caused by ejection of volcanic products” began in 2014 under SATREPS (Science and Technology Research Partnership for Sustainable Development) and is now coming to an end in 2019. Indonesia has 127 active volcanoes distributed along its archipelago making it a high risk location for volcano-related disasters. The target volcanoes in our study are Guntur, Galunggung, Merapi, Kelud, and Semeru in Java, and Sinabung in North Sumatra. Guntur and Galunggung are currently dormant and are potentially high-risk volcanoes. Merapi generated pyroclastic flows along the Gendol River in 2010, which resulted in over 300 casualties and induced frequent lahars. New eruptive activity of Merapi began in 2018. The 2014 eruption of Kelud formed a gigantic ash plume over 17 km high, dispersing ash widely over the island of Java. Semeru continued minor eruptive activity, accompanying a risk of a dome collapse. The aim of our research includes disaster mitigation of the Sinabung volcano, whose eruption began to form a lava dome at its summit at the end of 2013, followed by frequent pyroclastic flows for approximately 4 years, and the deposits became the source of rain-triggered lahars. Our goal is to implement SSDM (Support System for Decision-Making), which would allow us to forecast volcano-related hazards based on scales and types of eruptions inferred from monitoring data. This special issue collects fundamental scientific knowledge and technology for the SSDM as output from our project. The SSDM is an integrated system of monitoring, constructed scenarios, forecasting scale of eruption, simulation of sediment movement and volcanic ash dispersion in the atmosphere. X-band radars newly installed by our project in Indonesia were well utilized for estimation of spatial distribution not only of rain fall in catchments but also of volcanic ash clouds. Finally, we hope the SSDM will continue to be utilized under a consortium in Merapi, which was newly established in collaboration with our projects, and extended to other volcanoes.
“Integrated Study on Mitigation of Multimodal Disasters Caused by Ejection of Volcanic Products” Project was launched in March 2014 for the Galunggung, Guntur, Kelud, Merapi, and Semeru volcanoes. The objectives of the project include the development of an observational system for the prediction and real-time estimations of the discharge rate of volcanic products. Under the project, a team from the Sakurajima Volcano Research Center, Center for Volcanology and Geological Hazard Mitigation (CVGHM) and the Balai Penyelidikan dan Pengembangan Teknologi Kebencanaan Geologi (BPPTKG) initiated the installation of a digital seismic and global navigation satellite system (GNSS) observational network for the volcanoes in December 2014, and finished the installation in September 2015. The seismic and GNSS data are transmitted by wireless local area networks (WLANs) from the stations to an observatory at each target volcano. We introduced three Windows PC software for data analysis: the first for estimating the equivalent rate of ejected ash from a volcano, the second for continuous smoothing of tilt data and detecting inflation and deflation in the volcanic sources, and the third for continuously evaluating eruption urgency to predict the eruption time. The seismic and GNSS data were routinely transmitted to the Support Systems of Decision Making (SSDM) at CVGHM or BPPTKG. Data completeness varied from volcano to volcano; for example, the data acquired for Kelud volcano were relatively stable, while those for Merapi volcano were problematic, owing to a communication disruption in the WLAN. We obtained the seismic and GNSS data at the target volcanoes in the observation period since 2015 when they have been relatively quiet.
Merapi, the dangerous active volcano in Indonesia, has been monitored since the 1920s by applying several methods and tools. The monitoring data from earlier times are stored well and can be used as reference for any precursors and signs before each eruption. This article evaluates the long-term activity of Merapi from the monitoring data for 1992–2011 to obtain the trends and patterns before the eruption period by combining the seismicity, deformation, volcanic gas, and temperature data in the same time span. Several characteristics are exhibited before effusive and explosive eruptions, i.e., a significant level up in volcano-tectonic energy and increased CO2 gas concentration indicating an explosive eruption. Effusive eruption is characterized by a significant multiphase earthquake with less occurrence of deep and shallow volcano-tectonic events. Deformation data from a tiltmeter and electronic distance measurement are important in understanding the dynamics of the lava dome and the eruption direction.
Kelud Volcano is among the most active volcanoes in Indonesia, with repeated explosive eruptions throughout its history. Here, we reconstructed the relationship between the repose period and the cumulative volume of erupted material over the past 100 years and estimated the long-term magma discharge rate and future eruptive potential and hazards. Tephra data and eruption sequences described in historical documents were used to estimate the volume and mass discharge rate. The volumes of the 1901, 1919, 1951, 1966, 1990, and 2014 eruptions were estimated as 51–296 × 106 m3. The mass discharge rates were estimated to be on the order of 107 kg/s for the 1919, 1951, and 2014 eruptions and the order of 106 kg/s for the 1966 and 1990 eruptions. Based on a linear relationship between the repose period and cumulative erupted mass, the long-term mass discharge rate was estimated as ∼ 1.5 × 1010 kg/year, explaining the features of the larger eruptions (1919, 1951, and 2014) but not those of the smaller eruptions (1966 and 1990). This estimate is relatively high compared to other typical basaltic-andesitic subduction-zone volcanoes. This result provides important insights into the evolution of magmatic systems and prediction of future eruptions at Kelud Volcano.
Eruption scenarios were prepared as possible sequences in event trees for six active volcanoes in Indonesia, that are located near populated areas or have erupted in recent years (Galunggung, Guntur, Kelud, Merapi, Semeru, and Sinabung). The event trees prepared here show sequences of possible eruption phenomena without probabilities on branches and cover sequences experienced in historical and pre-historical eruptions based on archives and field research results. Changing magma discharge rates during eruption sequences were considered for the event tree of Merapi. This conceptual event tree can also be used as a short-term event tree in which forecasting the coming eruption became possible with geophysical and geochemical monitoring data. Eruption event trees prepared for selected time windows cannot illustrate all plausible hazards and risks associated with an eruption. Therefore, hazards and risks generated from an eruption should be considered in different domains from the event tree.
We propose a method to evaluate the potential volume of eruptive material using the seismic energy of volcanic earthquakes prior to eruptions of Merapi volcano. For this analysis, we used well-documented eruptions of Merapi volcano with pyroclastic flows (1994, 1997, 1998, 2001, 2006, and 2010) and the rates and magnitudes of volcano-tectonic A-type, volcano-tectonic B-type, and multiphase earthquakes before each of the eruptions. Using the worldwide database presented by White and McCausland , we derived a log-linear formula that describes the upper limit of the potential volume of erupted material estimated from the cumulative seismic energy of distal volcano-tectonic earthquakes. The relationship between the volume of pyroclastic material and the cumulative seismic energy released in 1994, 1997, 1998, 2001, 2006, and 2010 at Merapi volcano is well-approximated by the empirical formula derived from worldwide data within an order of magnitude. It is possible to expand this to other volcanic eruptions with short (< 30 years) inter-eruptive intervals. The difference in the intruded and extruded volumes between intrusions and eruptions, and the selection of the time period for the cumulative energy calculation are problems that still need to be addressed.
After a volcanic eruption, in situations where pyroclastic material generates thick ground cover, even small amounts of rainfall can trigger lahars; this effect sometimes continues for many years. For hazard mitigation against lahar disasters after an eruption, it is essential to evaluate the current risk of occurrence and estimate any possible temporal changes for the future. Putih River is one of the rivers where lahars occurred frequently after the 1984 eruption of Mt. Merapi. In this study, the characteristics of lahars and floods in the Putih River after the 1984 eruption and their change over the years were analyzed, focusing on the runoff characteristics of lahars. Irrelevant rainfall and discharge data for analyzing runoff characteristics of lahars were excluded in preprocessing. The magnitude and occurrences of lahars decreased annually during the four years following the eruption. The maximum runoff rate of lahars was approximately 12 during the 1984–1985 rainy season and decreased yearly after this. A judgement graph was employed to track the temporal changes of lahar- triggering rainfall characteristics. For the 1984–1985 and 1985–1986 rainy seasons discriminant lines, which discriminate between rainfall events triggering lahar flow with peak discharge > 900 m3/s and other rainfall events, were drawn on the judgement graph.
Rainfall monitoring is important for providing early warning of lahar flow around Mt. Merapi. The X-band multi-parameter radar developed to support these warning systems provides rainfall information with high spatial and temporal resolution. However, this method underestimates the rainfall compared with rain gauge measurements. Herein, we performed conditional radar-rain gauge merging to obtain the optimal rainfall value distribution. By using the cokriging interpolation method, kriged gauge rainfall, and kriged radar rainfall data were obtained, which were then combined with radar rainfall data to yield the adjusted spatial rainfall. Radar-rain gauge conditional merging with cokriging interpolation provided reasonably well-adjusted spatial rainfall pattern.
An X-band radar was installed in 2014 at Merapi Museum, Yogyakarta, Indonesia, to monitor pyroclastic and rainfall events around Mt. Merapi. This research aims to perform a reliability analysis of the point extracted rainfall data from the aforementioned newly installed radar to improve the performance of the warning system in the future. The radar data was compared with the monitored rain gauge data from Balai Sabo and the IMERG satellite data from NASA and JAXA (The Integrated Multi-satellitE Retrievals for GPM), which had not been done before. All of the rainfall data was compared on an hourly interval. The comparisons were conducted based on 11 locations that correspond to the ground rainfall measurement stations. The locations of the rain gauges are spread around Mt. Merapi area. The point rainfall information was extracted from the radar data grid and the satellite data grid, which were compared with the rain gauge data. The data were then calibrated and adjusted up to the optimum state. Based on January 2017–March 2018 data, it was obtained that the optimum state has a NSF value of 0.41 and R2 value of 0.56. As a result, it was determined that the radar can capture around 79% of the hourly rainfall occurrence around Mt. Merapi area during the chosen calibration period, in comparison with the rain gauge data. The radar was also able to capture nearby 40–50% of the heavy rainfall events that pose risks of lahar. In contrast, the radar data performance in detecting drizzling and light rain types were quite precise (55% of cases), although the satellite data could detect slightly better (60% of cases). These results indicate that the radar sensitivity in detecting the extreme rainfall events must receive higher priority in future developments, especially for applications to the existing Mt. Merapi lahar early warning systems.
Merapi has become one of the most enticing volcanoes due to its activity over the past century. Although we have to agree that the 2010 VEI = 4 (Volcanic Explosivity Index, ) eruption is the greatest in its recorded history, Merapi is more famous for its shorter cycle of smaller scale, making it one of the most active volcanoes on Earth. Many mechanisms are involved in an eruption, and pyroclastic flow is the most dangerous occurrence in terms of volcanic hazard. A pyroclastic flow is defined as a high-speed avalanche consisted of high temperature mixture of rock fragments and gas, resulted from lava dome collapse and/or gravitational column collapse. Researchers have studied Merapi’s history and behavior, and numerical simulations are an important tool for future hazard mitigation. By utilizing numerical simulation on basal part of pyroclastic flow, we investigated the applicability of the simulation on pyroclastic flows from historical eruptions of Merapi (1994, 2001, and 2006). Herein, we present a total of 32 simulations and discuss the areas affected by pyroclastic flows and the factors that affect the simulation results.
A pyroclastic flow is one of the most dangerous hazardous phenomena. To escape a pyroclastic flow, the influenceable area must be evacuated before the flow occurs. Therefore, to predict the inundation area of a pyroclastic flow is important, and numerical simulation is a helpful tool in this prediction. This study simulated a pyroclastic flow by reproducing the pyroclastic flow of Mt. Merapi that occurred in 2010. However, necessary detailed information of the flow to conduct the simulation, such as total volume and the property of the pyroclastic flow material, flow rate, etc., were not available. Therefore, 20 simulations were conducted, varying the important conditions, such as the volume of pyroclastic material, inter-granular friction factor, and duration of the flow. The results showed that the volume of the pyroclastic material and inter-granular friction factor strongly control the flow characteristics. However, the friction factor does not result in a wide range of values; therefore, volume is the most influencing factor. The most suitable condition is a total volume of pyroclastic material of 30 × 106 m3, a 5 min duration of flow, and a 0.6 friction factor.
Mt. Semeru (3676 m asl.) is an active volcano in Indonesia. Mt. Semeru has a specific topography i.e., a large straight scar in its south-east flank. The geometry of the scar is approx. 2 km in length and 300–500 m width. The scar is connected to three major drainage channels: the Kobokan River, the Kembar River, and the Bang River. On December 29, 2002, a pyroclastic flow (PF) with an approximate volume of 3.25 × 106 m3 was generated and it traveled 9–11 km along the Bang River. This pyroclastic flow was the largest among the ones generated from 2002–2003 eruptions of Mt. Semeru. All prior recorded pyroclastic flows traveled 1–2.5 km along the Kembar channel. Thus, this pyroclastic flow suddenly changed its flow path, and it traveled more than three times longer than its antecedents. To investigate the cause of the sudden change, a simulated reproduction of this pyroclastic flow was carried out by employing the numerical simulation method proposed by Yamashita and Miyamoto (1993). Due to the uncertainty of the volume of each pyroclastic flow and the temporal change of deposition thickness, a total of 12 simulation cases were set up, with variations in the number of sequence events, the duration of inflow at the upper reach of the flow, and the inter-granular friction factor. The simulation results showed that to explain the sudden change in flow path, the Kembar channel, around 3 km from the vent, has to be buried by antecedent pyroclastic flows. Furthermore, the individual volumes of the prior flows must be less than 0.25–1× 106 m3, with an inflow duration of less than 1 min. The friction factor must be set to be 0.5. By using the most acceptable case, the simulated pyroclastic flows were in good agreement with observed results. The results implied that careful investigation and continuous monitoring of the area at 1500–2000 m asl. on the south-east flank of Mt. Semeru are important to prepare for future pyroclastic flows.
An estimation method for debris flow potential is proposed to evaluate the possibility of the occurrence of rain-triggered debris flows. Sakurajima volcano has repeatedly erupted (Vulcanian type) and has continuously emitted volcanic ash at the Minamidake summit crater or Showa crater east of the summit since 1955, and debris flows have frequently occurred at rates of 10 to 111 events per year. Ground deformation associated with debris flows along the Arimura River were analyzed for the period from 2009 to 2016. Downward tilt (10–450 nrad) in the direction of the river and extensional strain (3–138 nstrain) were detected during occurrence of the debris flows. The tilt and strain changes were modeled using a point load caused by debris flow deposition beside a sabo dam. Depositional weights of individual debris flow events were estimated to range from 6 to 276 kt. The total weight of the debris flows was 2,154 kt, which is approximately 5% of the total weight of volcanic ash ejected from the craters during the study period. Debris flow potential (DFP) was defined as the difference in the volcanic ash deposits along the upper stream of the river (5% of the total) and the lower stream of the river, and the temporal change of the debris flow potential was investigated. When the debris flow potential reached a level of 0.4 Mt resulting from an increase in eruptive activity, debris flows frequently occurred or large debris flows were induced during rainy seasons. The concept of debris flow potential was applied to volcanoes in Indonesia as lahar potential. After the 2010 eruption at Merapi volcano, lahar potential, perhaps, quasi-exponentially decays during the dormant period. The lahar potential of Sinabung volcano complicatedly varies because of long-term eruptivity beginning in 2014.
This paper presents a theoretical method for estimating volcanic ash fall rate from the eruption of Sinabung Volcano on February 19, 2018 using an X-band multi-parameter radar (X-MP radar). The X-MP radar was run in a sectoral range height indicator (SRHI) scan mode for 6° angular range (azimuth of 221°–226°) and at an elevation angle of 7° to 40° angular range. The distance of the radar is approximately 8 km in the Southeastern direction of the vent of Mount Sinabung. Based on a three-dimensional (3-D) image of the radar reflectivity factor, the ash column height was established to be more than 7.7 km, and in-depth information on detectable tephra could be obtained. This paper aims to present the microphysical parameters of volcanic ash measured by X-MP radar, which are the tephra concentration and the fall-out rate. These parameters were calculated in a two-step stepwise approach microphysical model using the scaled gamma distribution. The first step was ash classification based on a set of training data on synthetic ash and its estimated reflectivity factor. Using a naïve Bayesian classification, the measured reflectivity factors from the eruption were classified into the classification model. The second step was estimating the volcanic ash concentration and the fall-out rate by power-law function. The model estimated a maximum of approximately 12.9 g·m-3 of ash concentration from the coarse ash class (mean diameter Dn= 0.1 mm) and a minimum of approximately 0.8 megatons of volcanic ash mass accumulation from the eruption.
This paper reports a preliminary attempt to determine volcanic ash particle size distribution using the video drop size detector (VDSD) for estimating volcanic ash amount with X-band radar. The VDSD records an image showing the size and number of particles falling into the aperture by a charge coupled device camera. Size distribution spectra of a range of particles from fine ash to small lapilli were derived in discrete form from the VDSD observation. The parameterization of the particle size distribution following Gamma function was done using volcanic ash of eruptions at the Sakurajima Volcano between December 13–21, 2014. Three Gamma distribution parameters were determined analytically. The analytical results revealed a continuous distribution of particles characterized by shape, intercept, and slope. The distribution was used to determine volcanic mass concentration, ground deposit weight, and reflectivity. Verification of these results with X-band radar observations showed that the reflectivity obtained from analytical results is similar to that from radar observation. However, the ground deposit weight from analysis was overestimated, compared with the real weight of ash deposit on the ground. The algorithm proposed in this study is expected to provide a practical method for estimating ash distribution in the aftermath of a volcanic eruption using radar-reflectivity for cases where direct measurement at the location is not possible. An overview of the algorithm for volcanic ash retrieval from X-band radar observations is also presented.
In this study, a real-time volcanic ash dispersion model called PUFF is applied to the Sakura-jima volcano erupted on 16 June 2018 to assess the performance of the new system connected with a real-time emission rate estimation. The emission rate of the ash mass from the vent is estimated based on an empirical formula developed for the Sakura-jima volcano using seismic monitoring and ground deformation data. According to the time series of the estimated emission rate, a major eruption occurred at 7:20 JST indicating an emission rate of 1000 t/min and continued for 15 min showing a plume height of 4500 m. It is observed that we need to introduce an adjusting constant to fit the model prediction of the ash fallout with the ground observation. Once the particle mass is calibrated, the distributions of ash fallout are compared with other eruption events to confirm the model performance. According to the PUFF model simulations, an airborne ash concentration of 100 mg/m3 extends to a wide area around the volcano within one hour after the eruption. The simulation result quantitatively indicates the location of the danger zone for commercial airliners. The PUFF model system combined with the real-time emission rate estimation is useful for aviation safety purposes as well as for ground transportation and human health around active volcanoes.
A consensus has emerged that decentralization of the disaster management sector improves disaster risk governance effectiveness and responsiveness. While many researchers contend that decentralization creates institutional capacity building and disaster management regulation opportunities, few studies have measured or analyzed both decentralization and disaster management. We examined changes to the disaster management system and the opportunities and challenges arising following decentralization, as well as how vertical and horizontal relationships between government actors have changed in Indonesia. First, we found that decentralization had a positive effect on the implementation of disaster management with respect to regulation, institutional establishment, budgeting, and planning. Second, despite general improvements, challenges remain, including regulatory inconsistencies, a lack of funding and capacity for local institutional establishments, a lack of participation of experts, a strong dependence on the central government, and an increased corruption rate. Third, while a decentralized disaster management system framework has been established, the local government’s capacity and the overall network remain limited, with national institutions playing a leading role. These findings suggest that empowering the Regional Disaster Management Agency (BPBD) and strengthening the vertical and horizontal provincial/municipal networks of the BPBD would both enhance the disaster management system and allow local actors to play a more critical role in disaster management.
When modeling by IFAS it becomes necessary not only to obtain the model parameters but also to define the cell size, which influences both the tank model and the kinematic wave model. Since PWRI-DH model, the model in which is based IFAS, is a distributed model, the cell size defines the discretization of the computational domain. On the other hand, PWRI-DH model use altitude data such as GTOPO30 and Hydro1k, both with resolution of 1 km and IFAS is restricted to a minimum cell size of 100 m. Because of the restriction on cell sizes, an error on the predicted discharge is obtained, since this size of cell is not small enough to capture any details. As results, it is necessary to find a cell size able to predict discharge correctly, and more important, to quantify the error produce by the model and taking it into account in the analysis of the results. In this paper, an analysis of the influence of cell size on the IFAS predicted discharge is performed. The effect of cell size on the delimitation of the river as well on the definition of the land use, on the definition of the vegetation cover, on the definition of the topographic of the river basin is evaluated in detail. From the results of this study, the authors have been able to improve the accuracy of the PWRI-DH model and therefore to predict discharge using IFAS, more accurately. Finally, conclusions of this study are presented.