Skip to content

Data & Methodology

How ReefLine turns pixels into insight

Everything in ReefLine is calculated from two NASA satellite records. This page explains the sources, each processing step, how much data is missing, and the limits you should keep in mind.
Selected reef
Lizard Island · Northern Great Barrier Reef, Australia
Analysis period
Full record
Latest satellite months
SST Mar 2026 · Chl-a Apr 2026
Reef analysis areas
11
Valid pixel values used
379,815
SST months, Jan 2003 – Mar 2026
279
Chlorophyll-a months, Jul 2002 – Apr 2026
286

Sources

Two satellite records

Both indicators come from the same instrument, processed by NASA and served through NOAA, which keeps them consistent with each other.

Sea-surface temperature

Satellite and sensor
MODIS on NASA's Aqua satellite
Product
Level-3 mapped monthly composite, daytime 11 µm sea-surface temperature, masked for quality
Processing
NASA Ocean Biology Processing Group reprocessing R2019.0, science quality (2003 to 2019), joined with a continuation file from January 2020
Resolution
1/24 degree grid, about 4.6 km at the equator
Units
Degrees Celsius
Access
NOAA CoastWatch ERDDAP, dataset erdMH1sstdmdayR20190SQ and its continuation

Chlorophyll-a

Satellite and sensor
MODIS on NASA's Aqua satellite
Product
Level-3 mapped monthly composite of the standard chlorophyll-a product (chlor_a)
Processing
NASA Ocean Biology Processing Group reprocessing R2022
Resolution
1/24 degree grid, about 4.6 km at the equator
Units
Milligrams per cubic metre (mg/m³)
Access
NOAA CoastWatch ERDDAP, MODIS-Aqua chlorophyll-a R2022 monthly composite

Processing

From satellite pixels to reef values

The same ten steps are applied to every reef and both indicators.
  1. 1

    Define an analysis area

    For each reef, ReefLine reads an 8 by 8 grid of satellite pixels centred on the reef, roughly 33 to 37 km across. The grid captures the water around the reef rather than the reef structure itself.

  2. 2

    Collect every monthly composite

    Monthly values for each pixel were downloaded for both indicators, from the first available month to the latest month published at the time of processing.

  3. 3

    Keep gaps as gaps

    Pixels flagged as missing because of cloud, glint, land or quality masks stay empty. Nothing is interpolated or filled in, and months without any published composite are shown as gaps.

  4. 4

    Average the clear pixels

    Temperature uses the arithmetic mean of the valid pixels. Chlorophyll-a varies over orders of magnitude, so it uses the geometric mean, which describes a typical value better.

  5. 5

    Record coverage

    The number of valid pixels is stored for every month, so every chart can show how much of the area the satellite actually saw.

  6. 6

    Join the temperature files

    The R2019.0 science-quality record is used through December 2019 and the continuation file from January 2020. Across their 22 overlapping months (January 2020 to October 2021) the two files are identical at every reef, so no adjustment is applied.

  7. 7

    Build a baseline

    The average for each calendar month over 2003–2012 becomes the reef's baseline. Chlorophyll-a baselines are averaged on a logarithmic scale.

  8. 8

    Calculate anomalies

    Each month is compared with the baseline for the same calendar month. Temperature anomalies are differences in °C, chlorophyll-a anomalies are percentage differences.

  9. 9

    Estimate trends cautiously

    Trends are straight lines fitted through monthly anomalies. Because warm and cool spells persist, the 95% range is widened using an effective sample size, so it is more conservative than a simple fit. Trends describe the past, they do not predict the future.

  10. 10

    Flag warm-season extremes

    Months are compared with the warmest calendar month of the baseline, an idea borrowed from NOAA Coral Reef Watch. ReefLine uses monthly data, so this is a simplified indicator rather than Degree Heating Weeks.

Temporal coverage

What is covered, and what is missing

No monthly composite was available for Dec 2024 or Dec 2025 (temperature) or for Jan 2024, Nov 2024, Dec 2024, Oct 2025, Nov 2025, Dec 2025 (chlorophyll-a). These months appear as gaps everywhere in ReefLine.
SST, R2019.0 science quality
SST, continuation file
Chlorophyll-a, R2022
Baseline 2003–2012
200420062008201020122014201620182020202220242026
SST months usedOverlap checked, not usedChlorophyll-a months usedMonth with no composite
Data coverage for each reef
ReefArea centreSST monthsChl-a monthsSST pixel coverageChl-a pixel coverageAlways-empty pixels
Lizard IslandAustralia14.67° S, 145.54° E277 / 279280 / 28699.9%99.0%0 of 64
Heron IslandAustralia23.46° S, 151.92° E277 / 279280 / 286100.0%100.0%0 of 64
Ningaloo, Coral BayAustralia23.13° S, 113.71° E277 / 279280 / 28684.0%83.7%5 of 64
Dampier StraitIndonesia0.54° S, 130.67° E277 / 279280 / 28699.3%85.9%0 of 64
Rock IslandsPalau7.17° N, 134.38° E277 / 279280 / 28699.9%93.8%0 of 64
North Malé AtollMaldives4.33° N, 73.50° E277 / 279280 / 286100.0%98.2%0 of 64
Giftun IslandsEgypt27.17° N, 34.00° E277 / 279280 / 28698.4%98.3%1 of 64
Lighthouse ReefBelize17.33° N, 87.54° W277 / 279280 / 286100.0%98.6%0 of 64
Looe KeyUnited States24.50° N, 81.42° W277 / 279280 / 286100.0%94.3%0 of 64
MooreaFrench Polynesia17.54° S, 149.83° W277 / 279280 / 28699.6%98.2%0 of 64
French Frigate ShoalsUnited States23.75° N, 166.21° W277 / 279280 / 286100.0%100.0%0 of 64

Limitations

What to keep in mind

Satellite records are powerful because they are long and consistent. They are also indirect, and these are the main ways that matters.

Clouds and gaps

Optical and infrared sensors cannot see through cloud, so some months rely on fewer pixels than others. Coverage is shown wherever values appear.

The surface, not the reef

Satellites measure the top layer of the ocean. Water at the depth where corals grow can be cooler or warmer, especially in deep or strongly mixed water.

Shallow and coastal water

Over shallow reef flats and near coasts, light reflected from the seafloor or carried by sediment can raise chlorophyll-a estimates. High values there need in-water confirmation.

Monthly averages smooth extremes

A short, intense heatwave can be diluted in a monthly mean. Daily products are better for detecting bleaching-level heat stress.

Pixels larger than reefs

A 4 km pixel mixes reef, lagoon and open water. Pixels that touch land can be masked or affected by land, and some never return a value.

Choices shape numbers

Anomalies depend on the 2003–2012 baseline, and trends depend on the period you choose. ReefLine shows both choices wherever numbers appear.

A delay before data arrive

Monthly composites are published after the month ends and are sometimes reprocessed. The newest month in ReefLine may be several months old.

One sensor, one era

Every value comes from MODIS on Aqua, which has operated since 2002 and is now well beyond its planned mission life. Future data may come from different sensors.

What ReefLine can and cannot tell you

It can

  • Show how sea-surface temperature and chlorophyll-a around a reef changed month by month since 2002
  • Compare a month or year with the same reef's long-term average
  • Estimate long-term trends, with their uncertainty
  • Compare the environmental records of different reefs
  • Highlight periods and places worth investigating more closely

It cannot

  • Measure coral cover, coral health or bleaching
  • Prove what caused a pattern, such as pollution or runoff
  • Predict future temperatures or how ecosystems will respond
  • Describe conditions below the surface or inside individual pixels
  • Replace in-water monitoring and local knowledge

Glossary

Terms used in ReefLine

Sea-surface temperature (SST)
The temperature of the ocean's top layer, as estimated from infrared radiation measured by the satellite.
Chlorophyll-a
A pigment in phytoplankton. Its concentration, estimated from ocean colour, is a proxy for phytoplankton biomass.
Monthly composite
A map that combines every valid satellite observation of a month into one value per pixel.
Pixel coverage
The share of the 64 pixels around a reef that returned a valid value in a given month.
Baseline
The average for each calendar month over 2003–2012, used as the reference for anomalies.
Anomaly
The difference between a month and the baseline for the same calendar month.
Geometric mean
An average taken on a logarithmic scale, suited to values that vary over orders of magnitude.
Warmest-month average
The highest of the twelve monthly baseline temperatures, similar in spirit to the maximum monthly mean used in coral reef monitoring.
95% range
The range of trend values consistent with the data, after allowing for persistence from month to month.