# Reading the recorded globe

Read the recording alongside the source timestamps. The cloud image, earthquake reports and rendered globe each describe a different moment.

[HTML](https://lukejbyrne.com/library/guides/jmyhuxvpox4/)
Content language: en
Author: Luke Byrne
Video published: null
Guide published: 2026-11-07

[Watch the video](https://www.youtube.com/watch?v=jmYHuxvpox4)
Video language: English

- [Download the replay observations (CSV)](https://lukejbyrne.com/library/downloads/gods-eye-replay-observations.csv)

## Check the replay at 2:10

Pause as the camera changes perspective. Compare the distance stated in the reply with the position shown on the recorded map. The two disagree. A successful camera change tells us nothing about the accuracy of the explanation.

1. Pause at 2:10 and note what changes on the recorded map.
2. At 3:16, write down the earthquake count and its time window.
3. At 4:14, compare the NOAA image timestamp with the time of the test. Use the CSV to check your notes.

## Read the weather image at 4:14

The NOAA infrared image in the recording is stamped 11:00 UTC. The test took place after 12:30 UTC on 4 November. The image was already at least 90 minutes old. Infrared brightness shows a temperature signal. It cannot confirm rainfall at a particular street.

## Read the earthquake report at 3:16

The USGS snapshot contained 18 events of magnitude 4 or greater in the previous 24 hours. The largest in that sample was magnitude 5.2 in Kyrgyzstan. Changing the time window or minimum magnitude changes the count.

## Compare the layers

Orbital positions are calculated from orbital data. The moving street traffic is simulated, and the visual presets change the rendering. When two layers overlap on screen, check the source and time for each before drawing a conclusion.

## The takeaway

For each observation, note the source, its time and what the image can show.

## Workflow resources

- [earthquake.usgs.gov](https://earthquake.usgs.gov/)
- [noaa.gov](https://www.noaa.gov/)

## Chapters

- [0:00 A real aircraft tracked over London](https://www.youtube.com/watch?v=jmYHuxvpox4&t=0s)
- [0:31 What God's Eye View is and where the data comes from](https://www.youtube.com/watch?v=jmYHuxvpox4&t=31s)
- [1:14 Setup: Node, npm ci, doctor and dev](https://www.youtube.com/watch?v=jmYHuxvpox4&t=74s)
- [1:48 The keyless map and London presets](https://www.youtube.com/watch?v=jmYHuxvpox4&t=108s)
- [2:10 Cockpit view](https://www.youtube.com/watch?v=jmYHuxvpox4&t=130s)
- [2:40 Satellites from CelesTrak orbit data](https://www.youtube.com/watch?v=jmYHuxvpox4&t=160s)
- [3:16 Earthquakes from USGS](https://www.youtube.com/watch?v=jmYHuxvpox4&t=196s)
- [3:52 Public traffic cameras](https://www.youtube.com/watch?v=jmYHuxvpox4&t=232s)
- [4:14 Weather images from NOAA](https://www.youtube.com/watch?v=jmYHuxvpox4&t=254s)
- [4:54 Visual styles and simulated traffic](https://www.youtube.com/watch?v=jmYHuxvpox4&t=294s)
- [5:18 Optional 3D city tiles and costs](https://www.youtube.com/watch?v=jmYHuxvpox4&t=318s)
- [5:42 Getting AI to control the map](https://www.youtube.com/watch?v=jmYHuxvpox4&t=342s)
- [6:37 Final thoughts](https://www.youtube.com/watch?v=jmYHuxvpox4&t=397s)

## Sources and limitations

These observations come from the recording made on 4 November 2026. Each source has its own update interval.

[Original video](https://www.youtube.com/watch?v=jmYHuxvpox4)