Which Mount Agung Hazard Zones Matter for an East Bali Home?

A volcanic hazard map becomes useful for a property decision only when its location and the kind of hazard are both clear. A pin dropped somewhere in “East Bali” does not meet that test. Neither does a distance quoted from the summit: terrain can channel a flow into a valley well away from the crater, while nearby plots on different ridges may have different mapped exposures. Even a house outside a prominent coloured area needs a separate assessment of the route used to reach it.

The starting document is PVMBG's Peta Kawasan Rawan Bencana Gunungapi Agung, catalogued on MAGMA Indonesia with a 2015 publication year. The sheet distinguishes KRB III in red, KRB II in pink and KRB I in yellow. These are volcanic hazard zones describing potential exposure under relevant eruption scenarios. They do not form a live exclusion order, a planning consent or a scale of property desirability. The original legend is essential: coloured areas associated with material moving down the mountain differ from hatched zones for material ejected or falling from the eruption column. A narrow yellow corridor following a watercourse should not be read as though it were a simple ring centred on the summit.

Physical behaviour explains the shapes. Pyroclastic density currents, labelled awan panas in Indonesian, are fast, hot flows. Lava, rockfalls and water-rich volcanic debris can also follow topographic routes, although they are not interchangeable processes. A lahar is a slurry of water and volcanic sediment, often associated with rain and stream channels. Ballistic fragments (lontaran) and ashfall (hujan abu) have different pathways. Reading only the red, pink or yellow fill can hide these distinctions. A 2019 Udayana University study identifies the 2015 PVMBG sheet it used as a 1:50,000 map; that scale is for regional interpretation, not proof that an individual boundary wall lies inside or outside a zone. The reduced image distributed online offers still less positional certainty.

There are other useful maps, but their labels must be read carefully. BNPB's InaRISK service exposes a raster layer called INDEKS_BAHAYA_LETUSAN_GUNUNGAPI_V3, describing a volcanic hazard index. Bahaya means hazard; risiko concerns risk and kerentanan vulnerability. A risk or hazard index derived for regional screening is not necessarily drawn with the same boundaries or assumptions as PVMBG's KRB map. A separate national BIG geoservice displays KRB I, II and III as vector polygons, but its existence alone does not prove that the dataset matches Agung's 2015 edition at parcel scale. The Bali InaRISK dashboard explicitly warns that some positions are indicative and should not be relied on for legal or civil-engineering purposes.

A third source answers a different question: what PVMBG reported about volcanic activity at a specified time. The available Agung volcanic activity report for 9 August 2026 recorded Level I (Normal), while still asking communities to watch for lahars along rivers originating at Agung. Level I describes the observations underlying that report; it does not cancel the mapped longer-term hazards or establish the alert level on 19 October 2026. Historic exclusion distances are different again. A Ministry of Energy announcement dated 4 January 2018 described an adjusted activity restriction during the earlier eruption episode. Treating that episode's radius as a permanent 2026 land-use boundary would confuse an operational instruction with a hazard-map class. A fresh dated PVMBG report is required whenever present-day activity or restrictions affect a decision.

Property portals speak in destination names. Hazard mapping needs geography that can be located consistently: an administrative village, the property's actual position and the road connection. The distinction is particularly important around Agung, where the volcano's radial appearance on a tourist map can obscure drainage divides and elongated flow corridors. A precise-looking real-estate pin is not automatically a surveyed plot boundary, and a digital map is only as reliable as its source geometry, date and registration.

Sidemen is not a single location. Karangasem's published administrative list gives ten villages in Kecamatan Sidemen, including the village named Sidemen, Sangkan Gunung, Tangkup, Telaga Tawang and Tri Eka Buana. A search result saying “villa in Sidemen” may refer to very different positions within that district. The Telaga Waja river provides a concrete example of why a valley matters. Local spatial-planning material identifies stretches of that river associated with Sangkan Gunung and Tangkup, while government water-management reporting places Telaga Waja within the wider Tukad Unda catchment. This establishes a named drainage system worth investigating where relevant; it does not establish a KRB classification for either entire village, or prove that a particular villa or bridge will be affected. Even within one valley, the two banks and their access roads should be identified separately.

Amed introduces a different location problem. The Karangasem regency's own description puts the Jemeluk–Amed visitor area in Desa Purwakerti, Kecamatan Abang. Listings commonly use “Amed” more loosely, however, and neighbouring Bunutan is another village in Abang, not a second name for Purwakerti. A seafront accommodation, a property set back on rising ground and a house served by a lane crossing a drainage line are not equivalent map points. The tourist label cannot be overlaid on a KRB sheet as though it were a cadastral polygon. Nor does proximity to the sea rule out ashfall, which follows the behaviour of an eruption plume and the weather.

Tulamben belongs to Kecamatan Kubu according to the same local-government village inventory. Its position in the northeastern part of Karangasem makes it a separate comparison from Jemeluk–Amed, even when holiday websites group the places together. Agung's official map shows hazard shapes that extend unevenly into different sectors of the volcano, with elongated zones as well as broader areas. What the available reduced sheet cannot establish is whether a particular Tulamben gate, lane or bridge intersects a mapped strip. Other villages in Kubu should not be assigned Tulamben's assumed exposure either. The useful comparison is between identified sites and access corridors, not between promotional names for coastal districts.

The label “Agung foothills” is broad enough to be misleading on its own. Pempatan is in Rendang, Sebudi is in Selat and Jungutan is in Bebandem; these are different approaches to the mountain and different administrative routes for contacting local authorities. Sebudi also contains the Pasar Agung area identified by the Selat district description, while Pempatan lies in another village network. Two houses may offer almost the same view of the volcano but sit on different sides of a ridge or use different valley crossings to reach the main road. A map reader must distinguish the flow-related zones from the ejection and fallout components before treating any coloured shape as evidence about a home.

For all four examples, a defensible result would identify a verified site, the relevant KRB edition, what the legend assigns to that position and the access route independently checked against rivers and local plans. We do not have cadastral coordinates for example homes or a validated parcel-level overlay of the 2015 sheet. The entries below therefore describe published geographical distinctions and the outstanding question for each location, without inventing a KRB class for an entire destination. A map can guide which documents and local details to request; it cannot create an address that the listing never supplied.

These are location-specific screening questions, not assigned KRB classes for properties. A verified plot boundary and appropriately registered mapping are still needed for parcel-level conclusions.

The road outside a property deserves its own position on the hazard map. A villa may be on higher ground while a bridge connecting it to the main road crosses a lower drainage corridor. Equally, two streets visible in an online basemap might share a single bridge or rejoin on the same exposed stretch. Neither arrangement can be assessed from the distance between the house and Agung's summit. The relevant geography includes the actual driveway, river crossings, steep or narrow links and the route through which the village connects with its surroundings.

There is a documented reason to consider river systems separately. PVMBG's dated 9 August 2026 report for Agung specifically warned of potential lahars along rivers sourced on the volcano. Indonesia's Directorate General of Water Resources has recorded lahar sediment issues in the Tukad Unda basin, including waterways named Telaga Waja, Yeh Sah, Kaon, Petandakan, Panti and Sabuh. Its 29 October 2021 report described flood and sediment-control works following earlier changes to the watercourse. These records do not establish that a named bridge is threatened today, or that an access road will close during a future eruption. They do show why the valley network is a relevant feature of property due diligence, especially when a journey crosses streams fed from Agung's slopes.

Road access has at least two distinct meanings. One is ordinary usability: can a resident reach the house, does the lane function in heavy rain, and are there bottlenecks that affect day-to-day access? The other is organised emergency movement: which connections are acknowledged in the applicable village plan and what authorities would direct residents to do? A KRB sheet answers neither question in full. Mapping services may display bridges without validating their current condition, and a route that works in one season may require new information in another. Conversely, a road omitted from a small-scale map should not simply be assumed absent. The appropriate level of confidence is determined by the evidence for the actual route, not the number of lines on a screen.

Bali's provincial BPBD reported a verification meeting on 26–27 November 2025 for early-warning procedures and evacuation mapping in KRB II villages. Draft maps and procedures were reviewed and an agreement record was signed; the final map layouts were still to be refined. That is substantive progress in local planning, but it is not proof that one uniform, publicly available route applies across Karangasem. A participatory uMap for Pidpid in Abang, credited to a BPBD-associated account and modified in February 2026, identifies local roads, river features and directions as part of a village-scale exercise. Its own scope matters: information relevant to Pidpid cannot be repurposed as the agreed route for a rental in Purwakerti, a home in Sidemen or a property in Kubu.

Before a purchase or long lease, the owner and the competent village or district authorities should be able to identify which road sections and river crossings serve the particular plot, what alternative connections genuinely exist and which current document governs local warnings and evacuation organisation. A signed local procedure, an illustrative public map and today's road condition are three different pieces of evidence. Even their combination cannot guarantee passage under an unknown future event; it can reveal whether a property decision is being made with identifiable information instead of a broker's general assurance. Specific questions for the owner and BPBD follow.

The KRB map is a spatial account of specified volcanic hazards, not a forecast calendar. It cannot say when Agung will next erupt, calculate an annual probability for a dwelling or determine which valleys will be affected in the next event. The operating alert level is also time-bounded. Level I (Normal) in the report dated 9 August 2026 describes observations and an assessment for that reporting period; it must not be displayed as the verified level in October 2026 without a newer official release. Property decisions affected by live warnings require the latest dated PVMBG report and the directions of responsible authorities.

Ash deserves separate consideration because it is not confined to a river or the edge of a coloured flow corridor. The KRB legend includes zones for falling or ejected material, yet the amount and direction of ash from any actual eruption are affected by its characteristics, plume development and atmospheric conditions. A house near the sea is not automatically shielded, while a printed circle cannot provide a reliable future ash depth for an individual rooftop. Heavy fallout, larger ballistic fragments and debris moving along a valley are distinct exposure mechanisms. A hazard map can identify scenarios worth discussing; it does not certify a roof structure or predict a building's performance.

At smaller scales, map precision can become deceptive. A listing platform may place a pin at the approximate centre of a property or even at the centre of a village. The original 2015 sheet was designed to communicate volcanic hazard geography over a broad area, and the publicly displayed reduced image loses line detail. Even when a polygon is available digitally, its source year, registration and underlying resolution must be understood before claiming that a cadastral boundary is crossed. If those conditions have not been met, a parcel's KRB position remains undetermined. The absence of a coloured feature at an approximate point does not prove there is zero volcanic exposure, and it certainly says nothing about hazards that the specific layer never attempted to map.

A building licence, legal land rights, structural condition and insurance coverage are governed by different evidence. InaRISK itself warns against relying on the positional information in parts of its Bali dashboard for legal or civil-engineering determinations. An insurer's exclusions or conditions must be read in the actual policy. An owner saying the property has permits tells you nothing definitive about the location of a lahar channel, just as a KRB classification says nothing definitive about compliance with land-use or building rules. Long-term renters also need to distinguish building management arrangements from official emergency arrangements.

A credible assessment can therefore end with several well-defined unknowns. It may establish the correct village and nearby mapped processes, document that a bridge needs examination, and identify the appropriate BPBD office for a current local procedure. Where the plot is not accurately located or the usable hazard geometry is missing, the map cannot settle the parcel question. That does not make the regional information worthless. It defines the boundary between a useful screening tool and the address-specific decision that still requires evidence.