NASA wants to see your pictures of the sky.

Starting on Sept. 15, the space agency’s PACE satellite (which stands for the Plankton, Aerosol, Cloud, ocean Ecosystem satellite) will circle the Earth every 98 minutes. While it’s overhead, the satellite will be collecting important environmental data, measuring how clouds and particles in the atmosphere absorb and reflect sunlight.

Here’s where you (and everyone else on land) comes in: Scientists want to compare what they’re seeing in the sky with what you’re seeing on the ground.

The pictures will help researchers to better understand the far-away images they’re capturing from space, said Dr. Andrew Sayer, the PACE atmosphere lead. Because each pixel in the PACE satellite images represent a distance of about 0.75 miles across, ground observations will give more detail and help fill in the gaps.

“We want to be able to identify whether or not each pixel in a satellite image is a cloud or not, because if it’s cloudy we want to determine the properties of that cloud, and if it’s not cloudy we want to determine the properties of the atmosphere and surface underneath the cloud. … So the ground observers will help us to assess that.”

On first read, this sounds ridiculous with the coarse resolution. I asked Grok to summarize what's really going on.

NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) satellite, launched in February 2024, is a global science mission, not a high-resolution mapping satellite. Its primary instrument, the Ocean Color Instrument (OCI), samples the Earth at roughly 1 km per pixel at nadir—on the order of 0.6–0.75 miles. That sounds crude next to commercial imagers such as WorldView Legion or Pléiades Neo (~30 cm) or SuperView Neo-1 (~25 cm). The difference is intentional.

Ocean color and aerosol–cloud retrievals require many photons and many wavelengths. The open ocean is dark; hyperspectral measurements from the ultraviolet through the shortwave infrared need a large ground footprint to achieve usable signal-to-noise. They also need a wide swath (OCI’s is about 2,600 km) so the planet can be covered every one to two days. Plankton patches, dust plumes, and cloud fields are typically kilometers to hundreds of kilometers across. A 1 km pixel is matched to those scales and continues the climate record begun by SeaWiFS, MODIS, and VIIRS. PACE’s advance is spectral, not spatial. A 30 cm resolution design would have been a different, narrower, more expensive mission that could not do this job.

The GLOBE Observer “PACE Clouds Challenge” is not an attempt to sharpen those 1 km pixels with phone cameras, nor to replace shipboard ocean-color validation. Professional ground truth for water-leaving radiance still comes from research vessels, buoys, AERONET-OC sites, and campaigns such as PACE-PAX. Citizen photographs address a different first-order problem: cloud detection.

Before OCI can retrieve ocean color or aerosols, each ~1.2 km pixel must be classified as cloudy or clear. Broken cloud and cloud edges are the difficult cases. One satellite pixel can mix cloud and clear sky; the mixture is ambiguous from orbit. A ground observer can report overcast, scattered cumulus, or thin cirrus at the same minute. Atmosphere science lead Andy Sayer has stated that this is why timed ground photos help assess the cloud mask.

Most volunteers will be on land, not at sea. That is acceptable. Cloud masking is required over land and ocean; PACE’s atmosphere instruments operate globally; and cloud systems are large enough that coastal and inland photos still sample the meteorology that affects nearby water. The phones are not measuring mid-Atlantic chlorophyll.

Participation is constrained. The GLOBE Observer app issues a flyover alert (typically 15 minutes ahead). The useful window is short—about five minutes of the afternoon overpass for a “PACE match.” The protocol asks for six daylight photos (zenith and the cardinal directions, plus down) plus a simple sky description. Most people will try this once or twice. Sustained data come from short public challenges, school classes, and a small core of regular observers. GLOBE Clouds has already accumulated more than a million observations and hundreds of thousands of photo sets over years; a 2024 eclipse campaign produced tens of thousands of observations in a month. Early in the current drive, NASA reported 775 PACE-matched submissions and asked for more. That volume is useful as an opportunistic check on the algorithm, not as a substitute for calibrated instruments.

In short: PACE’s coarse resolution is a design choice for global, hyperspectral ocean and atmosphere science. The photo campaign is a low-cost way to score the cloud mask at edges and in broken sky. It is complementary, limited, and dependent on a burst of volunteers plus classrooms—not on the public standing outside at 1:40 p.m. all winter.

The PACE satellite ties to my family history. Dad managed the first NIMBUS and ERTS projects at GE many decades ago. The early Nimbus series was an R&D weather laboratory. The Advanced Vidicon Camera System (AVCS) on Nimbus-1 produced sunlit cloud pictures at about 0.5 mile (~800 m) at nadir; some later AVCS/IDCS products are described as roughly 1–2 km. Infrared radiometers (HRIR, THIR) were coarser still, often several kilometers. Those instruments were built to see cloud systems and weather patterns, not streets or fields.

The direct ancestor of PACE is Nimbus-7’s Coastal Zone Color Scanner (1978–1986)—the first dedicated ocean-color sensor. CZCS had 825 m pixels at nadir, a ~1,566 km swath, and six bands (four ocean-color channels plus NIR and a thermal band). It could not run full-time; typical collection was on the order of two hours per day. That was enough to prove phytoplankton could be mapped from space, not enough for a continuous climate record. Dad had moved on to ERTS from the NIMBUS project and didn't work on NIMBUS-7.

ERTS-1 / Landsat-1 (1972)
ERTS was a different mission: civilian land-resource mapping. The Multispectral Scanner that actually defined the program delivered about 80 m resolution (sampled near 57 × 79 m) in four broad bands over a 185 km swath, with an 18-day revisit. The Return Beam Vidicon was also rated near 80 m. That is already about 12 times finer linearly than PACE’s 1 km pixel—and about 150 times finer in area. Landsat could show highways and large fields; it could not paint the global ocean every other day.

I remember Dad explaining "false color" imagery to me when I was in high school and college. It was tough to wrap my head around until I took some physics courses.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.