Coral-reef thermal stress and bleaching-context monitoring
Satellite thermal sensors track sea-surface temperature anomalies daily, giving reef managers quantified bleaching risk before divers enter the water. This page explains the physics, the sensors, and the honest limits of remote prediction.
Sensors
- NOAA-20 and SNPP VIIRS: Primary SST source for NOAA Coral Reef Watch. The VIIRS thermal infrared channels (M15/M16, ~11 and 12 µm) resolve SST to roughly 750 m at nadir with a swath of 3,040 km, giving near-daily global coverage. Absolute accuracy is approximately ±0.3 K under clear skies.
- Terra and Aqua MODIS: Provides the long SST climate record (2000 to present) underpinning the Maximum Monthly Mean baselines used in Degree Heating Week calculations. 1 km thermal bands, twice-daily revisit per satellite. Cloud contamination is the dominant data-loss mechanism at tropical latitudes.
- Sentinel-2 MSI: 10 m multispectral imagery (bands 1–4, 8, 8A) used to map reef geomorphic zones: fore-reef, crest, back-reef, lagoonal patch. No thermal capability. Revisit is 5 days at the equator under cloud-free conditions; cloud cover frequently extends effective revisit to weeks over tropical reefs.
- Landsat 8 and 9 TIRS: Thermal Infrared Sensor provides 100 m (resampled to 30 m) thermal data. Useful for nearshore SST mapping at finer spatial scales than VIIRS, but 16-day revisit limits its utility for tracking fast-moving thermal events. Contributes to long-term temperature trend analysis.
What a degree above normal actually does to a coral
Coral bleaching is not a disease. It is a stress response. Reef-building corals host symbiotic dinoflagellates called zooxanthellae inside their tissue; these algae supply up to 90 percent of the coral's energy through photosynthesis. When water temperature exceeds the local Maximum Monthly Mean (MMM) by roughly 1°C or more for a sustained period, the photosynthetic machinery of the zooxanthellae begins producing reactive oxygen species that the coral cannot neutralise. The coral expels the algae. Without them, the tissue turns translucent and the white calcium-carbonate skeleton shows through. That is bleaching.
Mortality is not immediate. A coral can survive a bleaching event if temperatures return to normal within weeks. The problem is accumulation. NOAA Coral Reef Watch quantifies this using Degree Heating Weeks (DHW): the sum of all SST anomalies exceeding the MMM by at least 1°C over the preceding 12 weeks, expressed in °C-weeks. Published field studies associate widespread bleaching with DHW values above 4 °C-weeks and significant mortality with values above 8 °C-weeks. These thresholds are empirical averages; local adaptation means some reefs bleach at lower values, others resist higher ones.
How NOAA Coral Reef Watch turns satellite radiance into bleaching risk
The NOAA Coral Reef Watch (CRW) programme has published daily global SST products since 2000, built initially on MODIS and now centred on VIIRS. The processing chain retrieves skin SST from thermal infrared radiance, applies atmospheric correction using numerical weather model outputs, and computes the anomaly against a climatological MMM baseline derived from the 1985 to 2012 AVHRR Pathfinder record. The result is a daily 5 km gridded product covering all reef pixels globally, updated within roughly 24 hours of satellite overpass.
From the SST anomaly field, CRW derives three operational products: the Bleaching HotSpot (anomaly exceeding 1°C above MMM), the DHW accumulation, and a categorical Alert Level from Watch (DHW above 0) through Alert Level 1 (DHW above 4) to Alert Level 2 (DHW above 8). These are openly accessible and form the standard context layer against which any in-water survey result should be interpreted. The 5 km pixel is a genuine limitation: a pixel can contain both a healthy deep-water area and a thermally stressed shallow back-reef, averaging the signal and potentially masking localised stress.
What high-resolution imagery adds, and what it cannot replace
Sentinel-2 at 10 m and commercial very-high-resolution imagery (sub-metre from platforms such as WorldView) cannot measure temperature. What they can do is map the physical structure of a reef with enough fidelity to disaggregate the thermal signal spatially. A geomorphic zone map derived from Sentinel-2 band ratios and water-column correction (the Lyzenga or Sagawa methods are well-documented in the literature) identifies fore-reef slopes, reef crests, rubble fields and lagoonal patches. Overlaying a DHW field onto this map lets a manager prioritise survey effort: which geomorphic zones are inside the high-stress pixel, and which are likely sheltered by upwelling or tidal flushing.
Spectral indices derived from Sentinel-2, particularly ratios of the red-edge and blue bands, can detect gross changes in reef reflectance between image dates, which correlates loosely with bleaching extent at the community level. This is not a substitute for in-water point-count surveys. Turbidity, depth variation, sun glint and variable water-column thickness all introduce noise that limits detection confidence. The honest position is that satellite optical data narrows the search area and tracks recovery trajectories over months; it does not replace the diver.
Latency, cloud and the gap between the sensor and the reef manager
The NOAA CRW daily SST product is available within approximately 24 hours of the satellite overpass, which is fast enough to issue early warnings before a thermal event peaks. The practical bottleneck is cloud. Tropical reef regions sit inside the Inter-Tropical Convergence Zone for much of the year. A VIIRS or MODIS thermal retrieval requires clear sky; persistent cloud cover can create gaps of several days in the SST time series precisely when a thermal event is building. CRW addresses this partly through multi-sensor blending and gap-filling interpolation, but interpolated pixels carry higher uncertainty than direct retrievals.
Sentinel-2 optical imagery faces the same cloud problem for structural mapping. A 5-day revisit under clear skies becomes a 30-day or longer effective revisit during the wet season over many reef systems. Managers commissioning post-bleaching damage assessments should plan for a multi-week acquisition window rather than expecting a single cloud-free image on demand.
Putting the data to work: from alert to prioritised survey
A practical monitoring workflow has three stages. First, continuous thermal surveillance using the CRW DHW product identifies when and where Alert Level 1 conditions are approached. This requires no custom satellite tasking; the data is open and updated daily. Second, when DHW thresholds are crossed, high-resolution optical imagery is commissioned or pulled from archive to produce a geomorphic zone map of the affected reef system. This map segments the reef into units that can be assigned to survey teams. Third, field surveys return point-count bleaching data that is georeferenced against the zone map, producing a validated bleaching-extent layer that feeds back into the monitoring record.
Satellize runs this workflow for clients operating in Pacific and Indian Ocean jurisdictions, combining open VIIRS and Sentinel-2 data with commercial tasking on client licence. The approach is similar in structure to the analytics we deliver for the Kingdom of Tonga crop-estimation programme: open-constellation data provides the continuous baseline; targeted high-resolution acquisitions answer specific management questions. For reef monitoring, the deliverable is a DHW alert feed, a geomorphic zone GIS layer, and a post-event bleaching-probability map that a reef manager can take into the field.
What this method cannot tell you
Satellite thermal data predicts physiological stress, not bleaching outcome. Local factors including water clarity, depth, tidal exchange, fish herbivory and coral species composition all modulate whether a given DHW value produces bleaching, partial bleaching or apparent resistance. A reef that has experienced repeated bleaching may have shifted toward more thermally tolerant coral assemblages, meaning historical DHW thresholds overestimate current risk. Conversely, a reef already stressed by sedimentation or nutrient loading may bleach at lower DHW values than the published thresholds suggest.
The 5 km resolution of the standard CRW product is a real constraint for small or fragmented reef systems. A single atoll with a complex lagoonal structure may fall within one or two CRW pixels, hiding the spatial heterogeneity that matters most to a manager deciding where to intervene. Downscaling using Landsat TIRS or nearshore SST models can sharpen the picture, but adds processing complexity and does not eliminate the fundamental limit that satellite radiometers measure a skin temperature averaged over their instantaneous field of view, not the temperature at the depth where corals live.
Typical figures
| SST spatial resolution (VIIRS CRW product) | 5 km gridded (native VIIRS thermal pixel ~750 m at nadir) |
| SST spatial resolution (Landsat 8/9 TIRS) | 100 m native, resampled to 30 m |
| Optical reef-structure resolution (Sentinel-2) | 10 m (bands 2, 3, 4, 8); 20 m (red-edge bands) |
| SST revisit (VIIRS SNPP + NOAA-20 combined) | Near-daily global coverage; cloud gaps degrade effective revisit |
| Optical revisit (Sentinel-2, equatorial) | 5 days under clear sky; often 20–40 days effective in tropical wet season |
| SST accuracy (VIIRS, clear sky) | Approximately ±0.3 K absolute |
| DHW alert latency (NOAA CRW) | ~24 hours from overpass to published product |
| Thermal archive depth | MODIS from 2000; AVHRR Pathfinder from 1985 (underpins MMM baseline) |
| Bleaching-risk thresholds (published empirical) | DHW >4 °C-weeks: widespread bleaching; DHW >8 °C-weeks: significant mortality |
| Delivery formats | GeoTIFF, NetCDF (SST/DHW fields); GeoPackage or Shapefile (geomorphic zones); PDF alert report |
Analytics Satellize can run
| Daily DHW alert feed | NOAA Coral Reef Watch SST anomaly accumulation over rolling 12-week window, applied to VIIRS SST fields | Automated daily alert email or API feed with current DHW value and Alert Level per named reef site |
| Reef geomorphic zone map | Sentinel-2 multispectral water-column correction (Lyzenga or Sagawa band-ratio method) and supervised classification into geomorphic zones | GIS polygon layer (GeoPackage) with zone labels, area statistics and depth-of-detection confidence flags |
| Bleaching-probability overlay | DHW field resampled and intersected with geomorphic zone map; probability score assigned per zone based on published DHW-bleaching response curves | Raster layer and summary table showing probability class per zone, for use in field-survey prioritisation |
| Post-event reflectance-change map | Bi-temporal Sentinel-2 blue/red-edge band-ratio differencing between pre-stress and post-peak-DHW acquisitions, masked for sun glint and cloud | GeoTIFF change layer with magnitude classes and area statistics by geomorphic zone |
| Multi-year thermal stress trend report | Annual maximum DHW extraction from MODIS/VIIRS archive (2000 to present) per reef polygon, with linear trend and exceedance-frequency statistics | PDF report with time-series charts per site and ranked vulnerability table |
| Recovery trajectory monitoring | Quarterly Sentinel-2 reflectance index time series over fixed transect polygons, tracking return toward pre-bleaching spectral signature | Time-series chart and GIS layer updated each quarter, flagging zones showing no recovery signal after 12 months |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.