Why Do Two Satellites Show Different Temperatures?

Two satellites looking at the same patch of ocean, even minutes apart, can legitimately report different numbers, not because one is wrong, but because resolution, pass timing, and quality filtering differ by sensor, and because two sensors can be measuring genuinely different things.

Different resolution

A VIIRS SST pixel averages roughly 2km of ocean; a MODIS pixel averages roughly 1km; an AMSR2 microwave pixel averages roughly 25km. If a sharp temperature front runs through that area, a coarser pixel blends both sides of it into one average: a real mathematical consequence of pixel size, not measurement error. See native resolution, explained.

Different pass time

Each satellite crosses a given region at its own time of day, and sea surface temperature genuinely changes over hours: daytime solar heating warms the very top layer, then wind and nightfall mix it back down. A morning VIIRS pass and an early-afternoon MODIS Aqua pass over the same water can disagree simply because the water itself changed between the two overflights. See why doesn't my boat's gauge match the chart for more on this diurnal effect.

Different quality filtering

Different sensor product lines flag their own cloud and quality-suspect pixels on different scales, and the scales aren't even oriented the same direction. ACSPO-processed SST (used for VIIRS and AMSR2) treats a quality level of 3 or higher as clear; MODIS treats a quality value under 3 as good. A pixel near the edge of “acceptable” can pass one product's filter and fail the other's, producing two different readings from data that both sensors technically captured.

Sometimes a genuinely different measurement

Infrared sensors read the sea surface skin temperature, in the top few micrometers. AMSR2's microwave reading is classified by GHRSST as a subskin temperature, the base of the thin conductive layer just beneath the skin, a different physical quantity by GHRSST's own definitions, even though the two normally track closely. This is a smaller effect than resolution or timing in practice, but it's a real one, not an error in either sensor.

None of this means either satellite is broken. See do all fishing apps use the same satellite data for how these small, legitimate differences compound across apps that each pick their own sensor mix. Compare sensors directly on the SST map. Back to Learn.