Living with a Changing Climate Across the Pacific
01
A small share can contain a wide range
Taken together, the 17 Pacific entities covered by Our World in Data produced about 17.3 million tonnes of territorial carbon dioxide in 2023. That was 0.0454 percent of the world total.
The number is small at global scale. It does not make every Pacific emitter small. The official Pacific greenhouse gas indicator spans 0.1 to 86.7 tonnes per person in 2024. Palau and New Caledonia sit far from the regional cluster.
02
Every complete ocean record moved warmer
All 21 complete sea surface temperature records were warmer on average in 2014 through 2023 than in 1993 through 2002. The shift ranges from 0.21°C in French Polynesia to 0.49°C in Nauru.
The direction is shared. The size and timing are not. Each row below compares the same two ten year windows without turning the countries into one regional average.
03
The tracks were already here
IBTrACS contains 508 South Pacific storm tracks from 1980 through 2023. A track map can show where storms travelled and which ones reached high recorded intensity. It cannot prove that warming caused an individual storm.
All recorded tracks
Five hundred and eight paths cross the basin
The archive begins well before the period shown here. The 1980 start keeps the view inside the modern satellite era.
Recorded intensity
One hundred and three tracks reached Category 3 or above
The categories use the IBTrACS USA Saffir Simpson field. Agency wind standards differ, so this is an orientation layer rather than a trend test.
Four event reports
Pam, Winston, Gita and Harold come forward
These tracks connect the basin map to the event assessments later in the story. They do not turn annual ocean warming into event attribution.
04
Heat becomes stress when it stays
Degree Heating Weeks accumulate coral heat stress over twelve weeks. NOAA marks 4°C weeks as a level where significant bleaching can occur. At 8°C weeks severe bleaching and mortality become likely.
The calendar shows the maximum annual value at eight Pacific regional virtual stations. It does not measure observed bleaching on the reef.
05
Annual rainfall alone does not explain harvests
Across 450 paired country years in 15 places, the correlation between annual rainfall anomaly and annual crop yield change is 0.04. That is close to no linear association at this scale.
This does not show that rainfall is unimportant. It shows that one national annual rainfall number is a poor explanation for harvest change on its own.
06
The sea level record moves into higher bands
The official values are stored in 0.1 metre steps. The visual therefore uses five broad bands rather than drawing a falsely precise line.
Most country records occupy the 0.1 metre band in recent years. These national ocean averages do not show a shoreline moving inland. Waves, sediment, engineering and land motion also change coastlines.
07
One storm can touch most of a country
Event assessments describe named cyclones. The official Pacific tables report annual totals compiled through disaster monitoring systems. The scopes differ and the numbers do not always agree.
That disagreement matters. The official annual affected series records zero for Vanuatu in 2015 while the Pam assessment documents 65,000 people displaced. Zero in the table is a reporting problem, not evidence that nobody was affected.
08
Electricity systems are changing at different speeds
In 2023, renewable sources supplied 94.4 percent of recorded electricity generation in Tokelau and 52.1 percent in Fiji. The other 16 places in the disaggregated power dataset remained below half.
Electricity is only part of energy use. Transport fuels and other final energy demand sit outside this chart. The separate SDG measure of renewable final energy is shown in the details where it is available.
Explore
Follow one place across the story
Choose a country or territory to connect its ocean, rain, sea level, disaster and power records. Missing observations stay missing. They are never drawn as zero.
What the evidence can carry
The imbalance is visible. The causal chain still needs care.
Pacific contributions to global territorial CO₂ are small in total. Ocean warming appears across every complete record. Cyclone tracks and reef heat stress show where exposure takes shape. Event reports show how large the consequences can become. Power data show that change is underway.
No single chart proves that a specific storm, harvest loss or coastline change came from climate change. The story becomes stronger when it says exactly what each dataset can support.