Weather Satellites: Observing Earth's Atmosphere from Space

Geostationary Satellites

Geostationary satellites orbit at 35,786 kilometers above the equator, where their orbital period matches Earth's rotation, keeping them positioned over the same point on the surface at all times. This fixed position allows them to capture images of the same hemisphere every 5 to 15 minutes, creating time-lapse sequences that reveal cloud movement, thunderstorm development, hurricane evolution, and large-scale circulation patterns in near real-time.

The United States operates the GOES (Geostationary Operational Environmental Satellite) series, with GOES-East centered over 75 degrees west longitude and GOES-West over 137 degrees west. The advanced GOES-R series, beginning with GOES-16 launched in 2016, carries the Advanced Baseline Imager (ABI) that captures images in 16 spectral channels every 5 minutes for the full disk and every 30 seconds for targeted severe weather areas. This rapid scanning enables forecasters to watch individual thunderstorm cells develop, track lightning activity, and detect the early signatures of tornado-producing storms.

Europe operates the Meteosat series over the prime meridian and Indian Ocean. Japan operates the Himawari series over the western Pacific. China, India, South Korea, and Russia also maintain geostationary weather satellites, ensuring continuous coverage of the entire tropics and mid-latitudes. Together, these satellites form a ring of observation around the equator that leaves no significant gap in tropical and subtropical coverage.

The primary limitation of geostationary satellites is their distance from Earth, which limits spatial resolution compared to lower-orbiting satellites. Additionally, their equatorial orbit provides increasingly oblique viewing angles at higher latitudes, reducing the quality of observations above about 60 degrees latitude. Polar regions require a different observation strategy.

Polar-Orbiting Satellites

Polar-orbiting satellites fly at much lower altitudes of 800 to 900 kilometers, circling the Earth roughly every 100 minutes in north-south orbits that cross both poles. Because the Earth rotates beneath the satellite, each successive orbit covers a different strip of the surface. A single satellite covers the entire globe in approximately 12 hours, making two complete passes per day (one ascending northward, one descending southward).

The lower altitude provides much higher spatial resolution than geostationary satellites. Polar orbiters carry advanced infrared and microwave sounders that measure vertical profiles of temperature and moisture through the atmosphere, data that is critically important for initializing numerical weather prediction models. The NOAA series (currently NOAA-20 and NOAA-21) and the European MetOp series carry instruments that measure temperature profiles at 1-kilometer vertical resolution and moisture profiles that reveal the three-dimensional distribution of water vapor.

Microwave sensors on polar orbiters can observe through clouds, unlike visible and infrared instruments that are blocked by thick cloud cover. This capability is particularly valuable for measuring sea surface temperatures beneath persistent cloud decks, observing precipitation within storms, and monitoring sea ice extent in the polar regions. Microwave sounders have become the single most impactful observation type for improving numerical weather prediction accuracy, particularly over oceans where conventional observations are sparse.

What Satellites Measure

Visible imagery captures sunlight reflected by clouds and the surface, similar to a photograph taken from space. Thick, white clouds reflect strongly, appearing bright, while thin clouds and clear skies appear darker. Visible imagery is only available during daylight hours but provides the highest spatial detail of cloud structure, texture, and extent. Forecasters use visible imagery to identify cloud types, track fog formation and dissipation, and assess thunderstorm development.

Infrared (IR) imagery measures the thermal radiation emitted by clouds and the surface. Since objects emit more radiation at higher temperatures, the satellite can distinguish between warm low clouds (which appear gray) and cold high cloud tops (which appear bright white). IR imagery works both day and night, making it essential for continuous monitoring. Cloud-top temperature from IR data is used to estimate cloud height, identify deep convection associated with severe thunderstorms, and track tropical cyclone intensity by monitoring eye and eyewall temperatures.

Water vapor imagery uses specific infrared channels that are absorbed and re-emitted by water vapor in the upper and middle troposphere. The resulting images show the distribution of moisture at these levels, revealing features like dry intrusions, jet stream position, and atmospheric rivers (narrow bands of concentrated moisture transport that carry as much water as the Amazon River). Water vapor imagery is one of the most valuable tools for understanding the atmospheric dynamics that drive weather system development.

Specialized instruments measure many additional variables. Lightning mappers on geostationary satellites detect and map lightning activity continuously across their field of view. Scatterometers on polar orbiters measure ocean surface wind speed and direction by analyzing the roughness of the sea surface. Radar altimeters measure sea surface height, which reveals ocean current patterns. GPS radio occultation sensors measure how satellite signals bend as they pass through the atmosphere, providing temperature and moisture profiles with high vertical resolution.

Impact on Forecasting

The impact of satellite observations on weather forecasting has been transformative, particularly for the Southern Hemisphere and oceanic regions. Before the satellite era, the Southern Hemisphere had so few surface observations that forecast models could barely outperform simple persistence (assuming tomorrow's weather will be the same as today's). With satellite data, Southern Hemisphere forecast accuracy now approaches that of the Northern Hemisphere.

Tropical cyclone forecasting has been revolutionized by satellites. Before geostationary satellite coverage, tropical cyclones over the open ocean could go undetected until they approached inhabited coastlines, sometimes with devastating consequences. Today, every tropical disturbance worldwide is detected and tracked from formation to dissipation. The Dvorak technique, developed in the 1970s, uses satellite cloud patterns to estimate tropical cyclone intensity, and remains in operational use alongside newer automated methods that analyze satellite imagery with machine learning algorithms.

Satellite data assimilation into numerical weather prediction models has been one of the largest contributors to forecast improvement over the past three decades. Denial experiments, where satellite data is withheld from model runs and the resulting degradation in forecast accuracy is measured, consistently show that removing satellite observations reduces forecast skill by 1 to 2 days. This means that without satellites, today's 5-day forecast would be no better than a 3-day forecast.

The Expanding Role of Small Satellites

The traditional approach to weather satellites involved large, expensive spacecraft carrying multiple instruments, with each satellite costing hundreds of millions of dollars and requiring a decade or more from design to launch. The emergence of small satellite technology is changing this model. Constellations of dozens or even hundreds of small satellites, each carrying a single focused instrument, can provide more frequent observations at lower individual cost and faster replacement timelines when instruments fail or technology improves.

GPS radio occultation is one area where small satellite constellations have already demonstrated value. By measuring how GPS signals bend as they pass through the atmosphere, these satellites derive temperature and moisture profiles with exceptional vertical resolution. Commercial constellations now provide thousands of radio occultation profiles per day, supplementing the roughly 1,800 radiosonde profiles launched twice daily worldwide. These data have proven particularly valuable over oceans and in the tropics, where radiosonde coverage is sparse and the additional profiles measurably improve forecast model accuracy.

Machine learning is increasingly applied to satellite data processing, enabling automated detection of features like tropical cyclone eyes, convective initiation signatures, and volcanic ash plumes that previously required human analysts to identify. These algorithms process the enormous data volumes generated by modern satellite systems at speeds no team of human analysts could match, reducing the time between observation and actionable information reaching forecasters and emergency managers. These advances in both satellite hardware and data processing are accelerating the pace at which satellite-derived observations improve weather prediction worldwide.