Pacific Ecological Performance by country

Marshall Islands

Eco rank
#5/ 16
score 57/100

How this data viz is built

The data, the methods and the technology behind this Pacific ecological data viz.

Each country page turns raw official statistics into an interactive profile — ranking, key indicators, detailed trends, comparisons and a living-reef view. Here is the whole chain, end to end — nothing is hidden or hand-tuned.

16 territories20 indicators1850–2026 coverage12 official dataflows0 back-end servers

Where the data comes from

Every value is an official statistic published by the Pacific Community (SPC) on the Pacific Data Hub .Stat, the reference source of the Pacific Dataviz Challenge. Series are pulled as machine-readable SDMX dataflows — no manual copy-paste, so the numbers are exactly the ones the statistical offices publish.

Climate changeEnvironmental taxesAgricultural productionPower generationTourism arrivalsMeteo monitoring networkPopulation (NMDI)SDG 3 / 6 / 11 / 15

From raw dataflows to the reef

1

Extract

SDMX dataflows downloaded from the Pacific Data Hub into flat CSV files.
2

Consolidate

Merged into one country × year table of 192 columns with a data dictionary.
3

Curate

20 indicators kept; 6 too-sparse territories set aside → 16 fully comparable. Each indicator tagged with family, polarity, unit & scale.
4

Load

CSV compiled to DuckDB / Parquet and queried with SQL inside your browser via WebAssembly.
5

Render

Query results feed the charts and the generative canvas reef.

The three families of indicators

Ecological performancescored

State & action a country can influence — biodiversity (Red List Index), land degradation, GHG per capita, renewables, safe water. These 5 are the only ones that make the score.

Climate pressureshown

Forces a country endures — sea level, sea & surface temperature, rainfall, disasters. Displayed for context, never scored: they are suffered, not chosen.

Contextshown

Economy, society & governance — tourism, electricity, population, fisheries management, health. Helps read a territory, outside the score.

How the score & ranking work

For each of the 5 ecological-performance indicators, we take a country’s latest available value and turn it into a percentile score from 0 to 100 — the share of territories it beats on that indicator (direction flipped for “lower-is-better” indicators). The overall score is the average of those five percentile scores, not an average of ranks.score = mean( percentile_score(latest value, 0–100) ) over the 5 indicatorsrank #/16 = the country’s position once every territory is sorted by that score; exact ties share a rank. A territory needs at least 4 of the 5 indicators to be ranked. The reef reuses the same idea year by year: every branch segment is coloured by its percentile that year.

The living reef, decoded

  • One branch per indicator, growing from its oldest year at the base to the latest at the tip — longer records grow taller.
  • Colour has three readings you can switch: vs the Pacific (rank), vs the country’s own record (its worst→best year), or year-on-year change.
  • Hover anywhere to read the exact year under the cursor: real value & unit, Pacific average, best in the region, rank and percentile.
  • Every reef is deterministic — grown from the country’s name through a seeded pseudo-random generator, so a territory always yields the same reef — hand-drawn on a bare HTML5 canvas, no charting library.

Built with

Evidence.dev · SQL + MarkdownSvelteKit · static prerenderDuckDB · WebAssemblyUniversal SQL · in-browserHTML5 Canvas · generative reefNode.js · data preparationVercel + GitHub CI/CD

Honest limits

  • Ranks are relative: “top of the Pacific” can still be far from a global sustainability target.
  • The score uses each indicator’s latest available year, which can differ between indicators and countries.
  • Only the 16 territories with complete-enough core data are ranked; six were set aside for fairness.
  • Climate-pressure indicators are shown but never scored — a country is not penalised for the warming it endures.

Data: official datasets of the Pacific Dataviz Challenge — Pacific Community (SPC), Pacific Data Hub .Stat. “Pacific average” = simple mean of the other territories with data that year. All processing is reproducible from the source dataflows.