Tanimbar Islands — building a social baseline

This case brings together socioeconomic compilation, household and community instruments, representative sampling and spatial migration planning in Tanimbar, Indonesia. The assessment connects the land and sea used by a project, the places through which people and goods move, and communities whose livelihoods and services may change. Island connections make field access part of the design.

This case revisits the site's historical Tanimbar exploration and a 2015 proposed supplementary baseline for the Abadi project. The draft describes an earlier offshore floating-LNG concept with a proposed logistics base near Olilit. Its site list and map editions record planning choices; they do not establish completed interviews, observed migration or realised impacts.

Project design has since changed. INPEX describes an onshore LNG concept, a revised development plan approved in December 2023, and engineering design beginning in August 2025. Its 20 February 2026 announcement reports environmental approval under Indonesia's AMDAL system. A contemporary assessment should use the relevant approved design and disclosure documents alongside these dated operator statements (project history, environmental approval).

The map rebuilds the historical layers in native Observable Plot. Village geography is attributed to BPS 2010; the archive objects were retained in 2019. The land outline provides separate 2020 regency context. Roads and infrastructure show historical mapped locations with unverified operating status. Hover a point for its name and identifier; use a closer map view to inspect the south, Selaru or north. The searchable table below records the proposed-site matches.

Build the socioeconomic compilation

The earlier notebook assembled BPS's 2013 Dalam Angka publications and extracted tables, including Tanimbar Selatan. Retain the publication, table and observation year for each series: a 2013 publication can report earlier observations. Pair these historical materials with the selected contemporary BPS editions, without treating different definitions or boundaries as a continuous series.

Keep source values beside cleaned fields and record each correction. Link tables to versioned geography through an explicit crosswalk, then investigate discrepancies with the publication or supplier. The dataset curation guide develops the source inventory and reproducible joins.

The historical rapid-appraisal plan reported uneven coverage between subdistrict publications and limited village-level and sex-disaggregated measures. It proposed a dedicated researcher to seek underlying records from BPS and sector agencies, including health, education, fisheries, agriculture, trade and land administration. This was a data-acquisition strategy; the proposal alone does not establish that the records were obtained.

Its interview priorities connected those records to lived conditions: Puskesmas staff, traditional practitioners and patients for health access; women and people relying on subsistence activities for service affordability; traders and producers for markets; and homeowners and suppliers for construction costs. Use these inquiries to explain differences in travel, price, season and service delivery, then connect the findings to clearly defined survey measures. Maintain a record of requests, received materials and unresolved gaps so that the baseline can be updated as evidence improves.

From facilities to connected communities

The historical proposal expanded attention beyond southern Yamdena and the logistics-base alternatives to northern settlements, Selaru and Larat. Its rationale combined potential exposure with settlement size, transport access, services, livelihoods and field feasibility. These remain useful dimensions to investigate, provided each is tied to a defined mechanism and a dated source.

Administrative polygons help organise these records. Fishing grounds, customary land relationships, supply chains and clinic journeys require their own evidence. Map residence, livelihood activity and service destination separately where their locations differ. Consultation should establish how people describe and use these relationships.

Connect instruments and sampling

The retained project instruments cover household composition, livelihoods, living conditions, services, resource access, mobility and project perceptions. The Indonesia baseline discussion links the individual modules and explains their relationship to BPS measures. They are historical question sets to adapt and test, rather than a completed contemporary survey.

For this case, connect household rosters to livelihood and resource-use records; connect mobility histories to residence and household membership; and connect access questions to actual destinations, travel mode, cost and season. Develop the variable dictionary, interview instrument and analysis together. Questionnaire and fieldwork guidance covers translations, consent, local units, piloting and quality control.

The TO7 draft lists 28 proposed villages. Search the transcription to inspect its geographic coverage. Names and codes are retained as written; Saumlaki and Sifnana share a code in the draft. Reconcile that identifier against the relevant BPS edition before joining statistics or geometries.

The point archive and draft agree on both name and code for 21 proposed sites. Seven need reconciliation: Bomaki and Matakus have no matching point; Mayano Raya/Meyano Raya and Watidal/Waitidal differ in spelling; Rumngeur, Sifnana and Tutukembong have different codes. The map highlights the 21 agreed matches and preserves the unresolved records in the table. It does not move or rename points to force a match.

The draft also alternates between 71 and 72 villages and contains unfinished annexes and sampling calculations. Its purposive coverage rationale is useful for reconstructing the inquiry. A defensible probability survey needs a reconciled frame, explicit population and estimand, documented selection probabilities and an analysis that follows the design. Focus groups and key-informant interviews contribute explanation and local knowledge through different forms of evidence.

The archived map uses separate proposed fieldwork phases and includes names that differ from this draft list. Preserve each edition independently. A location appearing in a proposed-sites map confirms its inclusion in that map version; survey completion requires fieldwork records.

Investigate migration through change and exposure

The historical needs assessment identified a central analytical problem: population change had not been separated into natural increase and migration. A modern inquiry should distinguish births and deaths, arrivals and departures, temporary presence, and changes in household membership. Census and administrative totals need compatible boundaries and definitions before comparison.

Ask how project activities could alter those movements. Recruitment, accommodation, procurement and information about employment create different pathways. Compare plausible alternatives—for example, contractor accommodation and transport arrangements—with independently justified background change. Record scenario inputs as assumptions until observations support them.

For receiving settlements, examine both capacity and distribution. An available clinic building does not establish staffed care; regional water capacity does not establish reliable supply to a particular settlement. The native spatial worked example demonstrates local deficits and seasonal access using explicitly synthetic data. Applying it here requires verified local counts, service delivery and route costs.

Track outcomes for established residents and newcomers with respectful, consistent definitions. Disaggregate where the design supports it, and protect information that could identify households or sensitive resource use. The migration management plan connects these findings to responsibilities, funding and monitoring.

Rebuild maps from evidence

The original notebook combined administrative areas, village points, roads, ports, airports and village building outlines. Each layer needs a source, observation or compilation date, coordinate system and permitted use. Retain the original object and checksum alongside any cleaned extract. The optional close-up below restores seven archived structure layers as native vectors, loading one village at a time. Their observation dates and completeness are unverified; occupancy, household size and tenure require other records.

Six earlier rapid-appraisal illustrations show how district fills, village boundaries, settlement labels and points were combined over Google Earth imagery. Their visible credits include Landsat and, in the southern close-up, CNES/Spot Image. Retain these illustrations as dated map editions with their credits. Rebuild analytical layers from the geographic sources and link each symbol to an identified record; label size and dot colour in an illustration need a documented meaning before they can support an assessment.

For repeat imagery, retain acquisition dates, product identifiers, processing level, cloud and shadow screening, ground resolution and display rights. Sentinel-2 offers free imagery with bands at 10, 20 and 60 metres for landscape and broad settlement-change analysis. Detailed building interpretation may need licensed finer imagery or field verification. Increasing a PNG's export dimensions leaves the observation resolution unchanged.

Compare dates and methods before interpreting apparent change. Cloud, season, vegetation, tidal conditions and processing can alter an image. For an assessment, document how a mapped change is verified and related to the project and to affected people's use of the place. The map publication guide explains the separation between public interpretation and permissioned source material.

Connect current statistics and field evidence

BPS's Tanimbar Podes 2025 catalogue and Podes 2024 catalogue identify relevant village-statistics editions. Their catalogue records were checked for this discussion; their tables have not been transcribed or audited here. Keep a crosswalk between historical and current names, codes and boundaries, with the evidence for each match.

Use the Indonesia baseline discussion, dataset curation guide and spatial baseline methods to connect that inventory to a survey and assessment. The earlier Tanimbar notebook remains part of the historical record; its remote imports and data vintages require review before reuse.

Sources and retained editions

The public map uses the site's versioned Maluku extract, derived from geoBoundaries Indonesia ADM2 with BPS/WFP/OCHA attribution and CC BY 3.0 IGO licensing. It represents 2020 regency geography. The historical display extract retains village positions, shared internal edges, seven road features, five port points and two airport points, with original-file checksums and the join review. Of its 71 village points, 70 carry BPS 2010 attribution; Eliase has no recorded source, code or district. Population attributes are omitted. The 28-village transcription records its separate source checksum and identifier limitations.

Historical source editions comprise TO7: Supplementary Baseline Survey, its component-two needs assessment, the proposed-fieldsites map, and the Maluku rapid-appraisal discussion with six cartographic illustrations. They inform the retrospective method discussion above. They are retained separately from contemporary operator disclosures and from the administrative context layer. Protected originals, where authorised for the reader, are offered through the managed archive table.