Review of Local and Global Impacts of Volcanic Eruptions and Disaster Management Practices: The Indonesian Example

Abstract

This paper discusses the relations between the impacts of volcanic eruptions at multiple-scales and the related-issues of disaster-risk reduction (DRR). The review is structured around local and global impacts of volcanic eruptions, which have not been widely discussed in the literature, in terms of DRR issues. We classify the impacts at local scale on four different geographical features: impacts on the drainage system, on the structural morphology, on the water bodies, and the impact on societies and the environment. It has been demonstrated that information on local impacts can be integrated into four phases of the DRR, i.e., monitoring, mapping, emergency, and recovery. In contrast, information on the global impacts (e.g., global disruption on climate and air traffic) only fits the first DRR phase. We have emphasized the fact that global impacts are almost forgotten in the DRR programs. For this review, we have extracted case studies from Indonesia, and compared them to those of other regions, because Indonesia is home to >130 volcanoes and experienced several latest volcanic eruptions with VEI > 5.

1. Introduction

Volcanic eruptions are considered major (very large) when the Volcanic Explosivity Index (VEI) ≥ 5 [1,2]. The nature of the impacts of a VEI ≥ 5 eruption ranges from the destruction of a city, an entire region, climate disturbances as well as to air travel [3,4,5]. Even an eruption with VEI < 5 may also have the potential to modify the environment and landscape particularly in proximal and medial facies, as well as surrounding human societies [2,6]. The variety of environmental destructions due to volcanic eruption differs from primary (e.g., summit collapse, vegetation burning, death), secondary (e.g., atmospheric cooling, global warming), and tertiary (e.g., flood, famine, disease) effects [6]. It is mostly generated by gas emissions, ashes, lava flow, pyroclastic flow, lahar, debris flow, and landslide which results in local and global impacts [6].

The variety of impacts can be simplified at the local scale by their consequence in various geographical features. Geographical features are any features on the Earth’s surface, whether natural or not, which can be characterized spatially and temporally [7]. It may be classified into four categories: (i) the drainage system, (ii) the structure morphology, (iii) the water bodies, and (iv) the society and environment (e.g., Refs. [2,8]). Several studies of local impacts have helped to design the disaster risk reduction (DRR) program on volcanic disasters, e.g., monitoring and evacuation [9], total vulnerability assessment [10], livelihood and resettlement [11,12]. At the global scale, the DDR issues such as air traffic disruptions or climate disturbance are more complex because a major eruption may impact anyone regardless of their geographic position [13]. Unfortunately, the interaction of both local and global impact with DRR is undeveloped. DRR is describe as a systematic effort to reduce the various types of causative factors of disasters by using various techniques or scientific approaches [14].

In this study, we present Indonesian cases both of local and global impacts along with volcanoes around the world as a comparison. We also describe how they relate to disaster management practices in order to provide informed data for decision making. Indonesia is an archipelago country with the highest number and highest density of active volcanoes in the world. Several active volcanoes have become open-laboratories such as Merapi in Central Java, Agung in Bali, and Sinabung in Sumatra (Figure 1). Among 130 volcanoes at present, major eruptions have also repeatedly occurred during the historical period: e.g., Samalas (1257 CE), Tambora (1815 CE), and Krakatoa (1883 CE) [5,15,16].

2. Local Impacts of Volcanic Eruptions

2.1. Impacts on the Drainage Systems

Volcanic eruptions can lead to geomorphological transformations in valleys, talweg, and the hydrographic network due to erosion and sedimentation processes. For example, at Merapi volcano, the sediment supply from pyroclastic density currents into the river network leads to riverbed aggradation and subsequently riverbed incision [19]. The river valley morphology has changed from the old Merapi (before 4.8 ka BP) to the new Merapi (after 4.8 ka BP) edifice by detrital fan deposits, andesitic lava flows, tephra, lapilli, and ash layers [20]. These materials buried the eastern drainage system of Merapi during a major flank collapse. During the 2006 CE eruption, Merapi’s block-and-ash flow covered the interfluve region and the main valley on the southern flank as long as ~7 km [21]. It indicates that the materials produced by volcanic eruptions are substantial in shaping morphology and channel stability. In the case of lahar/pyroclastic input, the most hazardous location along a drainage system is river confluence. For instance, the Mapanuepe and Marella rivers’ confluence widened up to 1.3 km after lahar input from the 1991 CE Pinatubo eruption in Philippines [22].

In the tropics, sediment-discharge in volcanic rivers increases during and after eruptions due to lahars, which can be either eruption-induced or rain-triggered during the rainy season [23,24,25]. Channels that are buried by lahars is often creating difficulties for extracting the necessary water for everyday uses. For example, lahars on the southern flank of Merapi obstructed the irrigation channels for three years after the 2010 event, and this led to the worst harvest in the last decade [26]. Furthermore, when volcanic materials settle in the drainage system, they reduce the system capacity and can cause overbank floods during heavy rainfall [27]. The morphology of a river valley can transform dramatically due to lahars, predominantly in two ways: (1) burial of the main channel and (2) scouring of a new valley and, therefore, abandon of the former one [28]. During the 2011 Merapi lahar events, a new 1.5 m-wide river channel formed as the lahar came through and eroded the main road of Sirahan village on the western flank [29]. The example of a river valley on Merapi that has been buried by lahars and now abandoned is presented in Figure 2.

2.2. Impacts on the Volcanic Structure

An eruption may alter the structure of a volcano and its evolution, depending on the type of eruption. On andesitic and rhyolitic volcanoes, structural changes during the growth of lava domes are frequently observed using geophysical monitoring systems, or other methods such as aerial photographs, which allow detecting even minor changes [30]. Hydrothermal alteration during the dome growth is often accompanied by structural weakening that potentially causes sector collapse [31]. Dome growth during a pre-eruptive event may also control a crater displacement, as identified in the 2007–2008 CE Kelud eruption in East Java, Indonesia [32]. However, dome growth can be associated with the VEI, i.e., a faster dome growth most likely creates a high VEI [33].

Volcano edifice can drastically evolve in two ways, i.e., caldera-forming eruption and sector collapse. Several volcanoes in Indonesia formed a caldera after a major eruption. There are three infamous volcanoes that resulted in caldera during the historic time in Indonesia, i.e., Samalas in Lombok (1257 CE), Tambora in Sumbawa (1815 CE), and Krakatoa in the Sunda Strait (1883 CE) [15,16,34,35]. Compared to the two others, caldera in Krakatoa is non-visible due to total collapse of the volcano in 1883 CE [36]. Samalas caldera is the largest (~6 km-wide) (Figure 3A–C). The caldera-forming eruption of Samalas was also responsible for extreme morphological changes in its eastern part of the island. The 1257 CE eruption ejected at least 4.4 × 10⁶ m³ of pumice rich PDCs (pyroclastic density currents) forming deposits with a thickness up to ~30 m, and it has significantly changed the valley pattern of eastern Lombok Island back in the year 1257 to the one existing today [37].

Many factors can lead to sector collapse during the process of eruption. Most of them are destabilizing factors such as an earthquake (tectonic faulting), magma intrusion, hydrothermal activity, gravitational loading, and progressive shearing [38,39]. Hummocky hills are typically formed by debris-avalanches, which induces morphological transformation through two ways: obstacle-free spreading and valley-filling distribution [40]. The textural materials (wet and dry) of hummocky hills is valuable to characterize their triggering factor, such as in Colima volcano (Mexico): dry materials most probably result from volcano-tectonic deformation, whereas wet materials are likely due to phreatic activity [41].

In Indonesia, three locations with extensive hummocky terrain resulting from debris avalanches can be observed from DEMs and aerial imagery. (1) In Tasikmalaya, West Java, the hummocky hills formed after an eruption of Galunggung volcano in 4200 + 150 BP that ejected 20 km³ of materials, creating a horse-shoe caldera (Figure 3D). Debris avalanche at that time formed hummocky hills 6.5–23 km away from the summit [42]. (2) The hummocky hills located in Jember, East Java, resulted from a flank collapse of Gunung Gadung, a composite volcano of Raung, which also left a horse-shoe shape with 13 km long and 8.5 km wide at the branch [43] (Figure 3E). Abundant hummocks can be found in the proximal-medial facies, and smaller hummocks with size < 20 m-height scattered in distal facies [43]. (3) Another hummocky hills system has been recently identified on Lombok Island. Debris avalanche deposits on Lombok Island covers an area of ~195 km² with total number of hummocks is >750 [44].

2.3. Impacts on the Water Bodies

The impacts of volcanic eruptions on water bodies include impacts on lakes, sea or ocean, and manmade reservoirs. At some volcanoes, water bodies are distributed at the volcano’s foot. They can also be the results of volcanic eruption and collapse, such as Mt. Bandai in Japan, where a group of lakes was formed by the sector collapse [45]. At the summit of the volcano, another type of water body is the crater lake and the caldera lake. In turn, the water-bodies can affect the eruptions, with notably the frequent breakout floods from crater lakes that can turn into lahars [46,47]. In Indonesia, access to water is important as in Merdada caldera lake, Central Java, a massive water extraction for agricultural purposes [48]. The modification of the chemistry of a crater and a caldera lake can have significant implications, notably in the pre and post-eruption [49]. Volcanic ash deposits in a lake system can increase water toxicity to living biota [50]. Tephra and ashfall can contaminate the lacustrine ecosystem [51], e.g., in Lake Van (Eastern Anatolia, Turkey) [52]. However, hazards and managements related to volcanic lake’s water quality are poorly developed [49]. For instance, water quality monitoring of volcanic lakes is useful for determining water’s common uses [48].

One of the well-known impacts of volcanic eruptions on water bodies is tsunamis. Volcanic eruptions can trigger tsunami through several processes: underwater explosion, volcanic blast, pyroclastic flow, underwater caldera collapse, subaerial failure, and submarine failure [53,54]. At least 24 known volcanic eruption-induced tsunamis in Indonesia are listed [54]. The latest one, which occurred in the Sunda Strait between Java and Sumatra Island in December 2018, was triggered by the eruption of Krakatoa volcano and its partial collapse, as shown from the Sentinel imagery (Figure 4). Prior to this event, Ref. [55] modeled a flank collapse with a volume of 0.28 km³ that would generate a tsunami with an initial wave height of 43 m and reach the Java coastline in 35–45 min. There are plenty of similar examples in other of the world, notably along the subduction-arcs archipelago, e.g., the 3500-year B.P. tsunami triggered by the eruption of Aniakchak volcano in Alaska [56] and the Late-Bronze age tsunami in Crete and Turkey triggered by the eruption of Santorini volcano in Greece [57].

2.4. Impacts on Societies and the Environment

Major volcanic eruptions can also cause fatal incidents, as indicated by the number of victims. The seven eruptions with the most significant death tolls globally are over 5000 fatalities [59], among of them, four events are located in Indonesia (

Table 1. List of most enormous fatalities due to volcanic eruptions since 1500 (source: Ref. [59]).

Volcano (Year)	Country	Fatalities
Krakatoa (1883)	Indonesia	36,000
Mount Pelée (1902)	Martinique	28,000
Nevado del Ruiz (1985)	Colombia	24,000
Tambora (1815)	Indonesia	12,000
Unzen (1792)	Japan	10,139
Kelud (1586)	Indonesia	10,000
Kelud (1919)	Indonesia	5110

). The largest number of victims ever recorded in modern history is 36,000 casualties during the 1883 Krakatoa eruption. However, the victims were mostly killed by the eruption-induced tsunami. The highest fatalities caused by a direct impact are recorded in the 1902 Pelée eruption that involved a PDC surge which killed 28,000 people. We assume that the 1257 CE Samalas eruption may also have large death tolls in Indonesia’s modern-historical eruptions. Indeed, a written source named Babad Lombok mentions that the city of Pamatan had an estimated population of more than 10,000 inhabitants when the eruption destroyed it.

In the event of an eruption, people living on and near the volcano are likely to be directly impacted, which for the farming population means loss of livelihood and access to its means, e.g., in Merapi (Indonesia) and San Vicente (El Salvador) [60,61]. In Indonesia, numerous ancient routes and cities are located on and near active volcanoes. For instance, the ancient Mataram Kingdom of Central Java had routes that climbed on Merapi and Merbabu volcanoes, with temples along the road and associated settlements, to avoid the swamp areas of the Borobudur basin [62]. Thus, despite being fertile, many cities and villages are located close to active volcanoes for economic and historical reasons. Similar cases are also figured in Lombok. Based on the written sources (Babad Lombok), the colossal eruption of Samalas has buried Pamatan, the capital city of Lombok, which is located in the foot-slope of the volcano. In addition, this ancient city was occupied by thousands of inhabitants, predominantly working on agriculture and aquaculture activities. The written source also described how they responded and conducted an evacuation to avoid the hazards and finally built a new civilization that completely differed from the former. This example emphasizes that ancient civilization also faced livelihood shifting due to volcanic eruptions.

The fragmentation of the volcanic material into fine ash and lapilli creates soils that are rich yet light to work with, and the volcanoes of Indonesia all have vegetated slopes. This land-cover often gets disrupted by eruption processes and there is a balance of vegetation destruction and revegetation occurring on volcanoes (e.g., [63]). This changing balance is often used to determine the extent of the volcanic deposits [64]. This results from the burying process and the burning process. It was shown at Mt. Shiveluch in Kamchatka and Sarychev Peak in the Kuril Islands (Russia), the pyroclastic surge and lava flows traveled through forest areas burned hectares of trees [65,66]. The prediction analysis shows that vegetation recovery requires several decades after the eruption event [66]. On volcanic islands, where most of the land-cover can be lost, the return of natural vegetation succession is controlled by seed dispersals and soil conditions, as it has been shown on the Krakatoa Volcano [67]. Human intervention is also known to influence vegetation successions and alter the vegetation species that used to grow before an eruption. The vegetation on the flanks of Mt. St. Helens volcano started to increase in number in the first 14 years after the eruption, notably accelerated by management practices [68].

Much of the research classifying volcanic impacts has been done for single events. It now appears as a logical continuity, with benefits for both the local inhabitant and the scientific community. The geographical division of local impacts are more feasible to develop concerning to DRR programs. However, at the global scale, the issues concerning DRR are more complex if only consider the geographical division. The interaction into DRR program for global impacts is more feasible to develop by involving two different factors, i.e., volcanic gases/aerosols and volcanic ashes.

3. Global Impacts of Volcanic Eruptions

We define here the global impacts of volcanic eruption is a scale of the globe, with an eruption having the potential of impacting anyone on the planet, regardless of their geographic position. These impacts may be (i) climatic disturbances related to the aerosols dispersion in the stratosphere and/or (ii) economic and social perturbations caused by volcanic ashes, which may disturb the air traffic thousands of miles away from the source volcano.

Volcanic gases (e.g., SO₂, H₂S, and CO₂) contribute more significantly to atmospheric-warming than volcanic ashes, and they exist in relatively high concentrations in the atmosphere. At high VEI, the volcanic column can eject material up to the stratosphere, in which case ejecta can travel rapidly worldwide. In the short term, SO₂ is one of the most ejected gases into the atmosphere. SO₂ become the significant factor in climate change, especially when converted to H₂SO₄ (Figure 5). It has been demonstrated to affect the global climate, potentially triggering temporary or long-term climate change [69]. In addition, the sulfate particle will stay between one and three weeks within the troposphere, whereas this duration may reach 1 to 3 years for stratospheric eruptions. In the long term, large VEI volcanic eruptions can alter variations in the global carbon cycles on Earth [70]. For example, the Samalas eruption in Lombok in 1257 CE [15], released 1.58 ± 0.12 gt (gt: gigatonnes) of sulfur dioxide, 2.27 ± 0.18 gt of chlorine, and 0.013 gt of bromine in the stratosphere [71]. It would have had far-reaching effects compared to contemporary eruptions for which data is available: the 1982 CE eruptions of El Chichon in Mexico and the 1991 CE eruption of Pinatubo in the Philippines that led to increase greenhouse gases and anthropogenic aerosols [72]. Not only a major volcanic eruption, even VEI 4 eruptions or less are also possible to affect the global climate anomalies [73].

The global impacts of the Samalas eruption 1257 CE have disturbed the global climate. As it is reported by Ref. [75] that the years 1258 and 1259, part of the Northern Hemisphere experienced the coldest summers in the past millennium. The effects were also aggravated by societal impacts, e.g., environmental and societal crises in England, France and Japan [75]. The nature of climate effects that were caused by Samalas led to the proposition that this eruption was one of the eruptions responsible for the Little Ice Age, part of a cluster of eruptions in the mid-13th century [76].

In Indonesia, another major eruption from Tambora volcano occurred in West Nusa Tenggara (Sumbawa Island) in 1815 CE. It produced 0.06 gt of sulfur in the stratosphere, causing a global spread of sulfate aerosols and, subsequently climatic perturbations [77]. This eruption modified air temperatures and precipitation in Europe, as recorded by the Czech Meteorological Station; the year 1816 is commonly known as the “year without summer” [78]. These anomalies returned to pre-eruption conditions in 1818, but many glaciers in the Alps continued to grow until 1820 as the glacial-systems reaction is much slower [79]. The schematic impact of the 1815 CE Tambora eruption on the global atmosphere is schematized in Figure 6. As the aerosols and gas spread near the equator, they impacted both the Northern and the Southern hemispheres, while atmospheric circulation usually confines most of the ejecta to the half-hemisphere, where the volcano is located. For this reason, Indonesia has a crucial position in term of global volcanic impacts, as the effects can rapidly reach both the Northern and Southern hemispheres.

For the historical period, global volcanic impacts are often recorded in written sources, such as the eruption of Eldgja in 934 CE [80]. This eruption has been suggested as a reason for the sweltering followed by hundreds of victims in China. It was affected by drought, plague, and famine that continued until 942–943 CE [81]. Despite the essential role these eruptions have played in history, many older eruptions that have impacted the climate remain debatable as to their origin. For example, two mystery volcanoes which erupted closely in 535–536 CE and 539–540 CE, injected SO₂ into the atmosphere at a similar volume to Samalas [82,83]. Potential sources for these eruptions have been recently proposed: Illopango in Salvador for the 535–536 CE event [84]; and El Chichon in Mexico for the 539–540 CE event [85]. These cluster eruptions have been suggested to be responsible for societal crises, pandemics, and human migration [86]. The 539–540 CE events were also reported to have disturbed the Maya’s society and environment, and led to massive migration on the core of Maya civilization [85].

The global volcanic impacts on climate are significant and are supposedly an essential concern for DRR and the contemporary human activities. The characteristics of these impacts are relatively similar across the globe: it starts with unusual weather variability and may result in societal crises, famine, and fatalities. Various examples have shown that societal impacts can occur everywhere on Earth, and may be devastating. Unfortunately, the risks associated with global environmental, economic and societal disruptions are not considered in DRR programs. This deficiency could result from the uncertainty in the volume of chemical elements ejected and the method of modeling. The development of research and technology on the volcanic study must be able to accommodate this issue for future mitigation.

If the volcanic gases strongly impact the climate conditions and recent human activities in the case of stratospheric eruptions, the volcanic ash is mainly responsible for modern air traffic disruption. Volcanic ash is ejected into the atmosphere and is sucked into the high-temperature turbine causing it to melt and solidify as glass that can cause the engine to stall, resulting in severe damage or failure [87]. Ref. [88] published a review of known aircraft incidents due to volcanic ash clouds from 1953–2009. The famous one occurred in 1982 when a plane from a British company flying from Australia to UK almost crashed due to the eruption of Galunggung volcano in West Java, Indonesia. Among other recent events, the Eyjafjallajökull eruption in Iceland, and the eruption of Merapi in Yogyakarta, Indonesia, both in 2010 also resulted in essential disruptions in global human activities. In April-May, the volcanic ash released from Eyjafjallajökull interrupted the air traffic throughout Europe, canceling over thousand flights [89]. In November the same year, Merapi eruption disturbed the air traffic for 15 days in the Central Java region, and perturbed 1950 flights departing from and to Yogyakarta [90]. At that time, one of vulnerable airline companies in Indonesia declared bankruptcy on January 2011 [90]. The societal impacts were also severe because of the timing of the eruption of Merapi, which happened at the same period as the period of pilgrimage to Mecca for the largest Muslim country in the world: Indonesia. With the occurrence of those eruptions, the air traffic warning system related to volcanic activity has then been improved to include real-time warnings through the International Airways Volcano Watch (IAVW) and the Volcanic Ash Advisory Centre (VAAC) [3].

4. Disaster Management Practices Related to Volcanic Impacts

Disaster management practices lie at the junction between individuals and communities and the natural environment processes, creating many cases as there are communities and volcanoes. In other words, no one model fits it all, and one could argue that it is a simplification brought by various researchers when working in spheres they do not fully understand [91]. However, in the last few decades, the paradigm has been shifted from focusing on response to mitigation strategies. General disaster management practices mainly consist of two stages, namely pre-disaster and post-disaster management. Another stage that requires attention is the emergency stage, i.e., when a disaster event has just occurred [92]. In this section, we discuss on how to correlate the knowledge of local and global impacts of volcanic disasters with disaster management practices. We have taken examples from the Indonesian DRR case, which is regulated in both national and regional laws. In the latest regulation, we have summarized four essential phases of DRR on volcanic hazard that should be analyzed and correlated with the impacts of eruptions, i.e., (1) monitoring, (2) mapping, (3) emergency, and (4) recovery (Figure 7).

4.1. The Monitoring Phase

Indonesia has an observation program that runs online and provides real-time data. Hundreds of volcanoes in Indonesia are monitored using an application-based platform called MAGMA (Multiplatform Application for Geo-hazard Mitigation and Assessment in Indonesia), which is accessible at https://magma.esdm.go.id/ (accessed on 27 December 2020). As for the volcanic activities, MAGMA has developed several real-time monitoring using CCTV (25 volcanoes) and live seismograph systems (11 volcanoes); all of them can be accessed online. Similarly, the Etna Volcano Observatory in Italy shows real-time data on an online platform (accessible at http://www.ct.ingv.it/ (accessed on 27 December 2020)) from several resources, such as video (CCTV) and tremor signals. Etna’s observatory also broadcasts real-time warning simulation of potential volcanic ash dispersion, especially for air traffic. MAGMA does not have this type of simulation. The Philippines also develop real-time data monitoring for all of its volcanoes, which accessible at http://www.phivolcs.dost.gov.ph/ (accessed on 27 December 2020), but the data are not presented online-spatially. These three examples show a similar method and data with distinct advantages for each covered region in this monitoring phase. Through CCTV and seismographs, the program can observe the global impacts of volcanic activities, particularly those related to local impacts, such as minor changes in volcanic edifices and ashes dispersal for air traffic warning. As mentioned before, the monitoring program is crucial: e.g., in the period of dome growth, which allows to forecast the nature of eruption or even the VEI scale.

Nevertheless, online volcano observatory in Indonesia still requires improvements. For example, the real-time ash modeling developed by the Mount Etna Observatory will be rather challenging to apply in MAGMA because, instead of focusing on one volcano, MAGMA has nationwide coverage. Consequently, the integration of map data acquired from VAAC (Volcanic Ash Advisory Center) into MAGMA to display real-time information on ashfall spread is relatively more straightforward. Visual data (images and videos from CCTV) published through online resources are very beneficial for enhancing disaster preparedness. In terms of air traffic crisis management, satellite imagery is helpful due to its ability to track hazardous ashes for several hours after the occurrence of an eruption [93]. Since climate changes have listed yet on the general DRR program related to global impact of eruption, the ability of satellite image for tracking down the spreading gasses and aerosols (up to days and even weeks) might be helpful for management improvement especially for climate changes modeling.

The real-time monitoring of volcanic activity level is essential for broad level preparedness and to support the policy decision in the other DRR phase, i.e., emergency phase. However, in the decision-making of volcanic alert level, there might be false alarms. People’s acceptance of the false alarm is crucial to the successful mitigation processes [94]. In the recent case of Merapi eruption (2020–2021), the danger zone is set within a radius of 5 to 10 km. Pyroclastic flows are also forecasted in the S, SW and NW directions. When the eruption occurs, pyroclastic flows arise in the SW direction. Most of the displaced people on the southern-eastern flank of Merapi then gradually returned home one day after the event.

4.2. The Mapping Phase

Mapping is an essential program in disaster management, especially for disaster preparedness and impact prediction. The most widely mapped aspect in disaster management is hazard. A hazard map is critical as basic information in disaster policy. Indonesia had developed a manual for multi-hazards mapping by Geospatial Agency of Indonesia in 2020, but not all nine types of hazards (i.e., earthquake, tsunami, volcano, flood, landslide, typhoon, drought, forest/land fire, high tide) in this country have a fixed method of mapping. The Center for Volcanic and Geological Hazard Mitigation (CVGHM) has developed maps for volcanic hazard. An example of such maps is shown in Figure 8. Recently, 62 volcanic hazard maps, containing the distribution of disaster-prone areas or Kawasan Rawan Bencana (KRB), have been produced for four hazardous zones: Java, Sumatera, Bali-Nusa Tenggara, and Celebes-Molucca. In the KRB map, hazards are divided into three major classes or zones: KRB 1, KRB 2, and KRB 3. Every zone represents a certain radius from the source of hazard and contains information on fall deposits. Based on the hazards, KRB 1 consists of areas threatened by lahars, ashfall, and acid rain. The KRB 2 covers larger areas that are potentially exposed to lava flows, surge, ballistic projectile, and ashfall. Meanwhile, KRB 3 includes areas with a further radius from the crater, including pyroclastic flows, lava flows, gases, and acid water. The Geospatial Agency of Indonesia has also developed a method for volcanic hazards mapping; however, it focuses only on the post-eruption phase. It deals with the rapid assessment after an eruption using an unmanned aerial vehicle (UAV), mobile mapping, and participatory mapping.

PHIVOLCS (Philippine Institute of Volcanology and Seismology) developed volcanic hazard maps that are very different from those found in Indonesia. A volcanic hazard map in the Philippines has been produced based on hazard type at a specific volcano. For example, the hazard map for Bulusan volcano, which consists of two types of hazards, i.e., lava flow and pyroclastic flow, differs from Hibok-hibok volcano, which is threatened by tephra fall, lahar, lava flow, and pyroclastic flow. PHIVOLCS also has developed an eruption-induced tsunami hazard map. The integration of volcano hazard map and subsequent tsunami occurrences in Philippine provides a brief information percentage of the prone area exposed and their status, e.g., small, partially, and largely.

The Indonesian KRB map has been developed to provide detailed spatial data, but it is still lacks information on the possible local impact in every zone. The variety of impacts on the volcanic structure and/or the hydraulic network can be implemented in the mapping process, notably from zones (KRB) 2 and 3. Furthermore, the impacts of potential sector failure or collapse and debris avalanche are not displayed either on the hazards maps. They could better understand the potential hazards, rather than just those last experienced. Unfortunately, this information remains unpublished except for scientific publications. In this case, the PHIVOLCS mapping method is more advanced and covers the variety of possible local impacts, for example the one on water bodies: subsequent tsunami hazard. On the contrary, impact of volcanic activities on water bodies in Indonesia is undeveloped, especially in two concurrent events like an eruption and its subsequent tsunami. In Indonesia, many volcanoes pose such threats [54], but their hazard maps still do not integrate the probability of subsequent tsunami occurrences. At the same time, the tsunami hazard map has not yet accommodated eruption-triggered tsunamis either. Most of the tsunami maps are made by the Meteorological, Climatological, and Geophysical Agency or BMKG based on the scenario of tectonic plate movements.

4.3. The Emergency Phase

The emergency phase in this section covers two programs, i.e., evacuation and relocation. Management strategy in the emergency phase is only applied to the affected villagers. In Indonesia, extensive evacuation process runs when a volcano is about to erupt in the next 24 h and the volcano has reached level IV. Vulnerable groups (e.g., elderly) might be evacuated when the volcano has reached level III. Volcanic alert levels in Indonesia consist of four-level i.e., from I to IV, which share similarities to the systems of alert levels in many countries like Italy [95].

The evacuation process followed by relocation in the DRR phase commonly generates a secondary problem related to livelihood. There are two types of relocation: temporary and permanent. In the case of Merapi, villages on the lower slope were buried by 2 m-thick lahar deposits, forcing the villagers to move and resettle in different areas: permanent relocation [29,96]. Displaced populations often complain about the difficulties in performing their dairy and agriculture-based livelihoods during evacuation-relocation period because they need to commute to their former village where their farmlands are located [11]. To prevent those difficulties during the crisis, the supporting activities and work programs should be prepared in relocation zone, e.g., building animal housing completed with water and food supply [97,98]. However, several requirements for successful resettlement are essential, including access to livelihood, social network, and building specification [61]. This process requires concerted preparation, given that the government has to prepare the source of livelihood for the relocated people. This program is possible if the government has already prepared hazards (especially local impacts), vulnerability assessment, and risks. It is crucial both for temporary and permanent displaced populations.

Vulnerability assessment is thereby powerful in describing the impacts of human societies on dangers in a disaster-prone zone, which is useful for supporting the evacuation-relocation program. In the manual for vulnerability mapping in Indonesia, the social parameter does not cover livelihood analysis. However, the environment parameters are only measured by the extent of the vegetated area, both the naturally growing plants and cultivated crops. For further improvement, environmental vulnerability also needs to assess vegetation species, vegetation density, and biomass analysis.

A case study of five Washington volcanoes (Glacier Peak, Mount Baker; Mount St. Helens, Mount Adams, and Mount Rainier) shows that detailed population-exposure analysis in vulnerability assessment is beneficial during emergency phase for addressing potential losses and people awareness [99]. In Indonesia, data related to population exposure in social vulnerability assessment are poorly developed. Regardless, the vulnerability mapping method is deemed enough for accommodating the possible impact on a map scale of 1:25,000. An improvement of vulnerability assessment can be made by developing physical, social, economic, and environmental parameters at detailed scale [10,100]. Details on the impacts on societies and the environment will undoubtedly be constructive in the evacuation-relocation process, especially after being integrated with the improved vulnerability map.

New concepts have been developed to overcome the problems in the evacuation-relocation programs. The concept of sister village, sister city, or sister island becomes good choices that are currently being developed to reduce the post-disaster livelihood crisis [101]. The sister village concept was implemented in the aftermath of the 2010 Merapi eruption as a mitigation measure against the next eruption [102]. Sister island in Indonesia has been developed technically in Ternate island threatened by Gamalama volcano by involving neighboring islands: Tidore, Hiri, and Halmahera, both for temporary and permanents displaced population [101,103]. The mapping phase is helpful for selecting the proper sister village and sister island. Other examples of a massive relocation that adopts the sister island concept is found in the Vanuatu archipelago. Ambae Island residents were relocated to neighboring islands, i.e., Santo, Maewo, and Pentecost [104]. The successful sister island program in Vanuatu was influenced by concerns raised in livelihood analysis. Therefore, societal impacts analysis is important to support the evacuation-relocation program in the form sister village or sister island.

4.4. The Recovery Phase

The recovery phase is carried out in the final phase of disaster management, especially after the other DRR phase sections have been implemented. The global impacts of volcanic eruptions have no direct influence on this phase, but the local impacts do. That is to say, there is no practical international framework that answer the needs generated by the global impacts if not the insurance system, which still leaves a large portion of the world population helpless. Local recovery programs are varied and range from social to structural recovery. Considering the recovery of hydraulic structures at Merapi volcano, the 23 sabo-dams of the Krasak river have shown that damage levels range from 5 to 23%. Thus, the reconstruction of sabo-dams with damage level > 20% was immediately implemented [105]. However, reconstruction of other river channels structure and dredging was also necessary for dikes or water dams. Besides, this management also has to reconstruct transportation infrastructures, primarily the ones that are damaged due to various hazards such as tephra fall, PDC, lava flow, and lahar. This type of hazard may have resulted in minor or major morphological changes on most volcanoes, which are vulnerable to the burial of roads and railroads, damage to airport runways and destruction of bridges. [106]. The foremost attention is supposedly given to house recovery, particularly because rebuilding houses in the restricted zones or areas at risk of volcanic morphological change is highly inadvisable. Successful recovery also depends on how strong livelihood is affected. Sinabung eruption did not cause any changes to the community social structure [107]. They returned to work as farmer during recovery period with supporting program from government. This is different from the impact of the Soufriere Hills Volcano eruption (Montserrat), which caused changes in the social structure from homogenous to heterogeneous and required long-term recovery [108]. However, both took a long time to get back to normal. In addition, the reforestation program is also crucial to restore environmental quality, therefore, plant species identification and biomass assessment in the environmental-impacts analysis provide beneficial information in the environmental recovery.

5. Conclusions

Volcanic eruptions with VEI ≤ 4 may have measurable impacts at the local scale. These local impacts are much more substantial during eruptions with VEI > 5 and may also have global consequences due to gases and ash is released. Therefore, volcanoes in remote areas have the potential to wreak havoc in distant parts of the world. This paper provides a review and categorization of local and global impacts of volcanic eruptions and describes how they relate to hazards and DRR. At the local scale, information on local impacts is beneficial for DRR phases, i.e., monitoring, mapping, emergency, and recovery. The pre-disaster phase, such as monitoring and mapping, has to consider any observable impacts (drainage system, structure, water bodies and society) that can be mapped to enable accurate loss estimation and proper management planning. To manage the emergency phase, it has to concern the assessment of potential impacts on societies and environment (vulnerability) to support the successful evacuation-relocation. In the recovery or post-disaster phase, detailed information on livelihood resources is crucial to speed-up the community resilience. However, up to now, the global impacts only fit one of DRR phases, i.e., monitoring. It was concerned on air traffic issues perturbed by volcanic ash, running by VAAC. For instance, the climatic impacts from volcano eruption-released gas remain unconcerned due to various limitations, e.g., uncertainty of ejected chemical volume and the modelling method. This contribution may allow us to develop data-informed decision making in volcanic hazards management.