Why Satellite and Buoy Water Temperatures Can Disagree

Sea surface does not identify one fixed measurement depth

Infrared satellite instruments sense an extremely thin surface layer, while contact thermometers measure the water surrounding their installation depth. The Group for High Resolution Sea Surface Temperature distinguishes skin, subskin, depth, and foundation temperatures because the upper ocean can have a changing vertical temperature structure. Foundation temperature is intended to exclude the daily warming cycle; it is not simply another name for the warmest water visible at noon. GHRSST explains these temperature definitions.

Buoy depth also varies with the platform. NOAA's National Data Buoy Center describes temperature sensors near one meter below the waterline on common smaller moored buoys and deeper installations on some larger hulls. That specification describes NDBC buoy configurations, not every coastal pier sensor. For a particular observation, use its own station metadata rather than assuming all instruments occupy the same depth. NOAA's buoy sensor-height and depth guide documents these differences.

The MUR map combines observations into an analysis

NASA's Multi-scale Ultra-high Resolution product combines satellite retrievals and in-water observations into a daily gridded temperature field. An analysis fills spatial gaps using an explicit estimation method; a filled ocean pixel does not prove that a cloud-free satellite directly measured that exact location that day. NASA's published pipeline describes the combination of infrared, microwave, and buoy inputs. The PO.DAAC MUR overview sets out that processing chain.

NASA describes MUR's analyzed variable as foundation sea-surface temperature and its grid spacing as 0.01 degrees, approximately one kilometer. Grid spacing is not a guarantee that every harbor entrance or beach pocket is resolved independently. Distinct products can use different spatial support even when they display the same number of decimal places. NASA's MUR science overview explains the foundation-temperature target and analysis scale.

What clicking the colored water actually returns

Water Temperature Live displays NASA GIBS imagery for the selected MUR analysis date. GIBS distributes visualizations as dated map tiles with predefined resolutions and projections. A visualization is a rendered representation of scientific data, rather than a direct download of its full numerical fields. The image date matters: loading a tile now does not turn its analysis into a measurement made now. NASA's GIBS access documentation describes those time and tile dimensions.

Our click tool reads the native image tile, matches the pixel's exact color to NASA's published palette, and shows the midpoint of that temperature interval. It also displays the interval and analysis date. An end-of-scale color gives a bound, and a transparent or unrecognized pixel gives no usable estimate. This design avoids extracting invented precision from colors, but the midpoint is still an approximation introduced by the visualization. It is separate from a station observation. See our map sampling methodology.

The table describes this site's current presentation; it does not assign one accuracy figure to every instrument or satellite pixel.

A useful comparison starts with matching records

First, confirm the units and identify the two timestamps. Then check whether the marker and sampled pixel actually occupy the same water setting. A harbor instrument and an offshore analysis cell are different comparison targets even if both appear beside one city name. Next check the instrument depth, where available, and whether the satellite product estimates a daily background rather than the particular hour sampled by the station.

Only after those checks should a persistent difference be investigated as a potential data problem. Keep the exact coordinates, station identifier, analysis date, and observation time when reporting it. A repeatable offset is more informative than an isolated screen comparison. Conversely, close agreement between two displayed numbers does not establish that an entire coastline has the same temperature: agreement at one comparison point cannot supply observations elsewhere.