Reading path: Coastal field methods →
A satellite measures light before it estimates biology
An ocean-color sensor records visible light leaving the sea and reaching the satellite. Processing first estimates the water-leaving signal after accounting for the atmosphere; algorithms then use that signal to derive products such as chlorophyll-a concentration or light attenuation. The map is therefore a chain of measurements and calculations, not a photograph of phytoplankton. NOAA describes chlorophyll as a proxy for surface phytoplankton, a useful ecological indicator whose meaning depends on the product and its assumptions. [1] [2]
That distinction matters when reading a color scale. A bright patch on a chlorophyll layer represents a higher estimated value under that algorithm, not a direct count of cells, proof of a harmful species, or a measurement of the full water column. The satellite observes the ocean’s optical surface layer. It does not identify organisms by sight, and one map cannot by itself establish what caused a pattern or what risk it poses.
Coastal water mixes several optical signals
Open-ocean algorithms often rely on relationships between visible wavelengths that work best where phytoplankton dominate the optical signal. Near shore, suspended mineral particles, colored dissolved organic matter carried by rivers, shallow bottoms and diverse biological material can all alter the light reaching a sensor. NOAA notes that these constituents complicate coastal interpretation, and its VIIRS documentation specifically flags persistent issues in colored-dissolved-organic-matter-rich waters. A high estimated chlorophyll value can therefore reflect algorithm sensitivity to more than pigment alone. [1] [2]
The direction of error is not universal. Sediment-rich water after runoff may become brighter or more turbid, while dissolved material can absorb particular wavelengths. Bottom reflectance may matter in clear, shallow water. Atmospheric correction also has to separate a relatively small water signal from light scattered by air and aerosols. These mechanisms can overlap, so a simple rule such as “green means algae” is not dependable in an estuary or surf zone.
Start with the product definition and quality flags
Before interpreting a map, identify whether it shows chlorophyll-a, remote-sensing reflectance, suspended matter, or a clarity measure such as Kd(490). Those are related but not interchangeable quantities. A chlorophyll layer estimates pigment concentration; Kd(490) describes attenuation of blue-green light and is often used as an indicator of clarity. Check the sensor, processing level, spatial resolution, units, compositing period and quality flags in the dataset documentation. A daily composite, an eight-day mean and a gap-filled analysis answer different questions. [1]
Then inspect what is missing. Clouds block ocean-color observations, and algorithms may flag pixels affected by land, glint or other conditions. A blank area is not a zero reading. Gap-filled products can be useful for regional context, but interpolation adds modeled estimates where observations were absent; distinguish those values from direct valid satellite retrievals. NOAA’s product descriptions identify when gap-filling methods are used, which makes the processing history part of the evidence rather than a footnote. [1]
A worked reading of a colored coastal patch
Imagine a map showing higher chlorophyll estimates in a plume extending from an estuary. The defensible first statement is narrow: the selected product estimates more chlorophyll-like optical signal in those valid surface pixels for that compositing window. Next compare the plume’s outline with recent rainfall or river discharge, suspended-matter products, reflectance bands and nearby cloud or quality masks. If the pattern aligns with sediment-rich water, that is a competing explanation to test, not proof that the chlorophyll estimate is wrong.
A stronger ecological interpretation needs independent evidence: water samples analyzed for pigments or cell composition, field measurements of turbidity, and repeated observations that show whether the signal persists or moves. Match the time and place carefully. A bottle sample taken hours later or several kilometers away may not represent the satellite pixel, especially in a tide-driven plume. NOAA’s validation description uses satellite-to-field matchups and reports coastal limitations; this illustrates why validation is conditional on water type and matchup design. [2]
Use time series as comparisons, not automatic explanations
A time series can make a map more informative by comparing the same location with its own seasonal baseline. Prefer like-for-like products and note sensor changes, processing updates, missing scenes and the length of any composite. An anomaly means a difference from a stated reference period; it does not identify the cause. A high value after a storm might reflect runoff, resuspension, biological growth, or several processes together. The map narrows where to investigate; it does not settle the mechanism.
For a field notebook, record the product name and version, sensor, observation dates, pixel scale, quality mask, units and the exact location sampled. Add tide and weather context, then state what the satellite observation supports and what remains uncertain. That small discipline keeps a visually compelling image from turning into an unsupported claim about a bloom, productivity or animal food supply.
Sources & further reading
Have a correction or a source to suggest? Send a note →