Cairo's Ramadan dimming looked exactly like a satellite detector failure
Researchers initially suspected the instrument was broken. After years of investigation, they discovered that cities like Cairo and Amman's night brightness fluctuated with the Islamic calendar, not the Gregorian one—the anomaly curve matched exactly with Eid al-Fitr, when Ramadan ends.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Investigating a suspected instrument failure revealed Ramadan's signal in the data
The team was calibrating VIIRS on NASA's new weather satellite, testing stability against desert and moon. Cairo and Amman's night brightness showed anomalous spikes. Researchers initially suspected the detector's high-gain channel was broken. Comparing two datasets, they found the anomaly curve matched the Islamic calendar exactly—the days when Ramadan ended and Eid arrived. The instrument was fine; it had simply captured human cultural rhythms for the first time.
— Miguel RománBright cities were erasing rural areas from the satellite's view
While chasing Christmas lights for calibration, the team realized they'd been fixated on high-brightness urban areas, blinding themselves to dim light from villages and remote communities—city lights ‘obscured’other places in the data. This became Black Marble's true turning point: capability to see not just prosperous cities, but the entire distribution of human activity, including areas without electricity and places vulnerable to power outages from disasters. The team described this as ‘giving us a new pair of glasses’.
— Miguel RománEight years after opening the data, external teams were producing most of the research
Black Marble started as a side project by a handful of people. It wasn't until NASA leadership pushed for complete data openness—allowing anyone to reproduce results from raw data—that the ecosystem expanded. That was eight years after the initial discovery. After opening the data, the team grew from five people into a larger one. Most related research now comes from external institutions; NASA itself became responsible for ‘maintaining data integrity’. When others encounter problems, they still come back to NASA.
— Miguel RománWith just one spectral band, the satellite couldn't detect true LED adoption
The satellite carries only one panchromatic band, mixing all visible and infrared light together. It cannot distinguish colors. The team deliberately traded the blue band during design for infrared capacity needed to monitor wildfires and natural gas flaring. The result: when Europe began replacing street lights with LEDs in recent years, the data appeared to show ‘dimming’. But LED's blue component falls in wavelengths the satellite cannot see. Actual brightness didn't drop.
— Miguel RománA 2016 blackout was an early warning for Hurricane Maria—a year before it hit
In September 2016, a large power outage struck the Aguirre plant in Puerto Rico, a year before Hurricane Maria. The team was still manually downloading data and hadn't moved to cloud infrastructure. Román reviewed Puerto Rico's night scene as usual one evening and found an entire region had ‘vanished’. This event was later confirmed as an early signal of Puerto Rico's fragile power grid. A year later, Hurricanes Irma and Maria made landfall in succession, causing what was then the largest power outage in U.S. history. Black Marble shifted from ‘curiosity-driven’to specifically tracking post-disaster recovery and resilience.
— Miguel RománData fusion achieved 30-meter resolution, but with important limits
Black Marble's standard product has only 500-meter resolution. FEMA and other users complained: ‘too coarse; I need precision down to individual buildings’. The team had an idea: use Landsat—a NASA/USGS satellite designed to photograph buildings and streets. The hypothesis: more concrete means likely more light. They fused Landsat's built-area layer with night light data to create an image that ‘looks correct’but is essentially simulated. Román emphasized such products are only for situational awareness, not replacing real physical measurement. But many users want ‘60 percent solutions that don't make them wait three years for a paper’.
— Miguel RománAfrica's night light signal is weaker than moonlight noise—the real frontier
10 percent of the global population still lacks electricity, concentrated mostly in Africa. There, artificial night light signals are weaker than background noise from moon reflection—the worst signal-to-noise scenario in the entire system. Román believes this difficulty is precisely why NASA must invest maximum effort in data credibility: the World Bank uses Black Marble to evaluate multi-billion-dollar electrification projects. If measurements are wrong, the entire decision chain breaks.
— Miguel RománIt's not a black box—it's physics you can inspect and verify
Asked why spending time explaining algorithms to journalists, Román emphasizes Black Marble's core value is explainability. It's not a neural network that works for inscrutable reasons, but a method grounded in physics—built on decades of accumulated radiative transfer modeling by predecessors. Every assumption is public, testable. He believes convincing people results aren't fabricated matters as much as discovery itself.
— Miguel RománIn their own words · checked verbatim
The instrument was fine, but it had given us a gift. They had given us a capability to monitor humanity from space and literally culture from space.
Miguel Román5:20
A lot of work goes into convincing yourself that you're not being fooled because the easiest person to fool is yourself.
Miguel Román13:40
It's not dimming because Europe is poor. It's not dimming because there's some power. It's dimming because all of Europe decided to switch two LED lights in three years.
Miguel Román15:35
Not everything that's important in life is the thing that shines the brightest.
Miguel Román28:50
if you care about electrifying Africa, where the nighttime signal is three times smaller than the moon-condaminated signal, that's 10% of the human population right there.
Miguel Román37:00
It's not a neural network that somehow works. This is actual physics.
Miguel Román46:30
Figures
| Black Marble HD resolution | 500 meters to 30 meters | 30:21 |
| Global electricity access gap | 10 percent of population without power | 37:18 |
| Barranquitas blackout duration | 60 days | 25:40 |
| From discovery to award | 15 years | 41:10 |
Glossary
- Black Marble
- NASA project using satellite night light data to monitor human activity and energy conditions.
- Golden Goose Award
- An award recognizing U.S. federal research grants that initially seemed far-fetched but generated major social value.
- VIIRS
- Visible-infrared imaging radiometer suite on U.S. weather satellites; used to detect night lights.
- nadir
- The angle at which a satellite looks straight down at Earth's surface, where light signals are least distorted.
- downscaling
- A statistical method using high-resolution ground or satellite data to calibrate low-resolution satellite measurements.
How to listen
Engineers and policy researchers interested in satellite remote sensing, disaster data, and energy electrification.
The opening 0:00-0:49 is a preview clip of later content; you can skip it and start with the calibration story.