GOES-16, the Geostationary Lightning Mapper, and the Space-Based Weather Pipeline Behind the Satellite Overlay on Your EFB

GOES-16, orbiting 22,300 miles above the equator, has been continuously scanning the western hemisphere for developing convection since December 2017. Every weather product a pilot uses for flight planning flows downstream from this satellite and its successor, GOES-18. Understanding the data pipeline - from sensor to screen - changes how you interpret what you’re looking at.

Why Geostationary Orbit Is the Only Practical Tool for Watching Weather Develop

Weather satellites operate in two fundamentally different orbital regimes, and the distinction determines what kind of data they can produce.

Low-earth orbit satellites sit a few hundred to a few thousand kilometers above the surface. They move fast, covering the entire globe through successive passes over the course of a day. The tradeoff: they see any one location for only a few minutes per pass. That works for global mapping. It doesn’t work for watching a summer thunderstorm complex develop over several hours.

Geostationary orbit solves that problem. At approximately 35,700 kilometers above the equator, orbital mechanics produce an orbital period of exactly 24 hours. The satellite appears stationary above a fixed point on the ground - permanently. A meteorologist can watch a line of cells frame by frame, in near real time, and determine whether the system is organizing or losing energy. That continuity is the entire point.

The GOES Program Before GOES-16: Real Limitations

NOAA has operated geostationary weather satellites since 1975 under the GOES (Geostationary Operational Environmental Satellite) program. The earlier satellites were useful, but they imposed constraints that aviation weather products couldn’t fully overcome.

GOES-13, which covered the eastern United States before the current generation, scanned the full western hemisphere disk roughly every 30 minutes. A dedicated severe weather sector could scan every 5 minutes. Visible channel resolution was approximately 1 kilometer per pixel; infrared channels measuring cloud-top temperature ran at 4 kilometers per pixel.

For a slow-moving frontal system, that was workable. For a rapidly developing afternoon thunderstorm complex over the Midwest in summer, significant escalation could occur between scans. The data existed - it just didn’t refresh fast enough or resolve fine enough to track fast-moving convection with the precision forecasters needed.

GOES-16 and the Advanced Baseline Imager: The Numbers That Changed Operational Weather

The GOES-R series represented a deliberate break from incremental improvement. NOAA and NASA developed the program over more than a decade. GOES-16 launched in November 2016 and was declared fully operational in December 2017. GOES-18 - which replaced GOES-17 after a cooling system failure limited that satellite’s capabilities - went fully operational in January 2023. Together, GOES-16 covering the east and GOES-18 covering the west provide continuous dual coverage of the continental United States.

The primary imaging instrument is the Advanced Baseline Imager (ABI), operating across 16 spectral bands compared to 5 on the previous generation. Those channels simultaneously reveal water vapor distribution at three different altitude layers, cloud-top temperature, snow and ice discrimination, volcanic ash, dust, and cloud phase - whether a cloud consists of liquid water droplets or ice crystals. Cloud phase has direct icing forecast implications.

Scanning speed improved just as dramatically:

Full western hemisphere disk: every 15 minutes (previously 30)
Continental United States: every 5 minutes (previously 5 minutes only for severe sectors)
Mesoscale sector (roughly 1,000 km on a side, centered on active weather): every 60 seconds

Visible channel resolution improved from 1 kilometer to 0.5 kilometers per pixel. Individual convective towers are resolvable. Anvil spreading from a mature supercell is visible. The leading edge of a developing storm complex appears on satellite before it produces enough precipitation to register on radar.

The 60-second mesoscale scan rate is what changes real-time decision-making. A cumulus tower building over central Florida in afternoon heat: on the infrared channel, cloud-top temperatures are dropping as the storm climbs. Rapid cloud-top cooling is a reliable signature of an intensifying updraft. When tops are cooling fast, the storm is still ascending. That’s actionable information with a time component.

The Geostationary Lightning Mapper: The Instrument Most Pilots Have Never Heard Of

The imaging improvements on GOES-16 are significant. The instrument that changed aviation weather most is one most pilots have never heard of.

The Geostationary Lightning Mapper (GLM) is the first instrument of its kind ever flown on a NOAA operational weather satellite. It is an optical transient detector - it watches the atmosphere and specifically hunts for brief flashes of light with the spectral signature of lightning.

This is categorically different from ground-based lightning detection networks. The GLM sees lightning from space, continuously, across the entire observable disk, day and night, including through the tops of the storm clouds themselves.

The sensor operates in a narrow spectral band near 777 nanometers, close to the peak emission wavelength of lightning plasma. It filters out sunlight, city lights, and moonlight, detecting only that transient optical pulse. The detector runs at roughly 500 frames per second, capturing individual strokes within a flash, not just the flash as a whole.

The GLM detects approximately 70 to 90 percent of all lightning activity in its coverage area - intra-cloud, cloud-to-cloud, and cloud-to-ground. Ground-based networks are excellent within sensor range, but their detection efficiency drops with distance. Over the open ocean, remote Alaska, and the Gulf of Mexico, the GLM sees what ground networks cannot.

Why Lightning Flash Rate Is a Storm Intensity Indicator

Lightning is not just a hazard location marker. It is a data source about storm kinematics.

A storm producing high lightning flash rates has a strong electrical field. That electrical field requires vigorous charge separation, which comes from collisions between ice particles and graupel in an active updraft. Strong updraft means severe icing potential, significant turbulence, rapid vertical development, and hail. The lightning flash rate describes the physical intensity of the storm’s engine.

Decades of peer-reviewed research have established a specific relationship: a lightning jump - a rapid doubling or more of total lightning flash rate within a 15 to 20 minute window - precedes severe weather (large hail, damaging winds, tornado) at a rate significantly above chance. The lead time between the lightning jump and the appearance of a severe weather signature on radar is typically 10 to 15 minutes.

Ten to fifteen minutes of lead time, before the hazard is visible on radar, is real working time for a route deviation or divert calculation.

The GLM feeds lightning density data to NOAA’s Weather Prediction Center in near real time. Aviation Weather Center products incorporate GLM data. Consumer EFB platforms are increasingly informed by the same underlying detection - the full-resolution GLM flash-rate data hasn’t fully penetrated every platform yet, but the pipeline is maturing.

How GOES-16 Data Gets From the Satellite to Your Tablet

The engineering of this pipeline matters operationally because it determines the age of the data you’re looking at.

GOES-16 downlinks raw sensor data to NOAA’s primary ground station at Wallops Island, Virginia, with backup reception sites. Raw data is radiometrically calibrated and geo-rectified - corrected for the satellite’s exact position and orientation so imagery aligns accurately with the geography below. From scan to calibrated product available for downstream use: roughly 2 to 4 minutes.

That processed data flows into NOAA systems, then into the Aviation Digital Data Service, commercial weather providers, and EFB platforms. By the time a satellite overlay appears on a pilot’s tablet, the data is typically 3 to 8 minutes old, depending on the product, the app’s update cycle, and data connection speed.

Nexrad radar data has its own timeline. The composite mosaic product in most EFB apps is typically 5 to 10 minutes old. In rapidly evolving convective situations, a storm can intensify significantly within a single data refresh cycle. Neither source is instantaneous. Know the latency.

How to Use Satellite Imagery Alongside Radar, Not Instead of It

Radar and satellite answer different questions. Used together, they produce a picture meaningfully better than either alone.

Radar shows where precipitation is occurring right now. Satellite shows cloud structure - including the developing portions of a storm that haven’t yet produced enough precipitation to register as a radar return. A cumulus tower building toward flight levels is invisible on ADS-B weather until it starts precipitating. On GOES-16 visible imagery, it’s a white tower extending toward the flight levels in near real time. On infrared, cloud-top temperature drops as it builds. On GLM data, lightning generation confirms an active, energetic updraft.

On any preflight involving convective activity near a route:

Pull up the satellite layer alongside radar - not instead of it.
During the day, use visible imagery. At night, use infrared.
Identify developing cells visible on satellite that have no corresponding radar return. These are building and could become a problem.
Find large radar returns where infrared cloud tops are warming. A storm with warming tops may be losing updraft energy.
If GLM data is available, note which cells are producing lightning. Flash rate trends matter more than the presence of lightning alone.

The data tells a story beyond colored precipitation blobs.

Known Limitations: Parallax, Latency, and Integration

GOES-16 has documented limitations every pilot using the data should understand.

Parallax displacement is inherent to geostationary geometry. Because the satellite views the earth from a fixed point above the equator, it doesn’t have a vertical view of high-latitude locations. A thunderstorm over interior Alaska appears displaced from its actual surface position. The further from the sub-satellite point directly below the satellite, the more pronounced the effect. For the central and eastern continental United States, displacement is small. Further north and west, it increases. Some EFB platforms apply parallax corrections automatically; others don’t.

Data latency is real and variable. The 3-to-8-minute pipeline lag, combined with a developing storm’s rate of change, means the satellite picture is always slightly behind reality. In rapidly evolving convective situations, treat all data products with an understanding of their age.

Integration across data sources remains an interface design challenge as much as a technical one. Some EFB platforms combine radar reflectivity and cloud-top temperatures into a single composite view. Others still present layers separately. The best platforms are closing this gap.

What’s Next: GOES-19 and the ProbSevere Pipeline

GOES-U launched in June 2024 and will become GOES-19 upon reaching its operational slot - the final satellite in the GOES-R series. The following generation, the GOES-A series, will push spatial resolution and update rates further.

The downstream pipeline from space to cockpit is also maturing. NOAA’s experimental ProbSevere model ingests GOES-16 ABI imagery, GLM lightning data, and Nexrad reflectivity to generate real-time probability of severe weather for each individually tracked storm cell, updated every two minutes. In operational evaluations, it provides actionable lead time before severe weather signatures appear on standard products.

ProbSevere isn’t in consumer EFB apps yet, but it is in the toolchain forecasters use when briefing routes. The path from research-grade product to operational pilot tool is established. Timeline depends on data licensing and commercial pipeline development.

A consistent finding in accident reports involving convective encounters: satellite imagery available at the time would have shown a developing hazard not clearly visible in the radar picture. The gap between what the data showed and what the pilot saw is exactly what better tools and better pilot education work to close.

Key Takeaways

GOES-16 has scanned the western hemisphere continuously since December 2017, with a mesoscale scan rate of every 60 seconds and visible resolution of 0.5 km per pixel - a major improvement over prior-generation capabilities.
The Geostationary Lightning Mapper (GLM) is the first space-based optical lightning detector on a NOAA operational satellite, detecting 70–90% of all lightning including over open ocean and remote terrain with no ground sensor coverage.
A lightning jump - a rapid doubling of total lightning flash rate within 15–20 minutes - precedes severe weather on radar by approximately 10 to 15 minutes, providing meaningful decision time.
By the time satellite data appears on a pilot’s EFB, it is typically 3 to 8 minutes old; Nexrad composite mosaics are typically 5 to 10 minutes old. In evolving convection, neither source is real-time.
Satellite and radar answer different questions. Developing convective towers are visible on GOES-16 visible and infrared imagery well before they produce a radar return - using both data sources together provides a materially better hazard picture than either alone.