This is a pretty good photograph:
… Though it’s so ridiculously overused and cliché that it surprises me that graphic designers, etc., still seem to like using it.
Maybe we’ve been so thoroughly conditioned by it that other photographs of the Earth just don’t look like we think the Earth is supposed to look. If it doesn’t have a big comma-shaped cloud formation under South Africa, it must be an alien planet!
It’s known as The Blue Marble, and was taken on December 7, 1972, by the crew of the Apollo 17 lunar mission.
To take a fully-sunlit non-composite photo of the Earth, you have to be on an imaginary line segment connecting some part of the Earth with some part of the Sun. And for best results, you can’t be so far away from the Earth that your camera can’t take a high resolution photo of it, as in this famous photo from Voyager 1, from about 6 billion km away:
(The above photo was not taken from between the Earth and Sun, but you get the idea.)
And you can’t be too close, or you’ll only be able to see a small part of the Earth. Low Earth orbit (e.g. the International Space Station) is too close. Here’s an extreme example, from a few feet off the ground:
As far as I know, the setting of the Blue Marble was the first and only time humans have been in (or at least, close to) such a position.
But, I admit, I really don’t understand why good Earth photographs seem to have been so rare until relatively recently. Geostationary satellites have existed since 1964, and every one of them has a better vantage point than Apollo 17 did. Here’s a to-scale diagram:
“BMP” indicates the distance at which the Blue Marble Photograph was taken (about 29,000 km from the nearest point on the Earth’s surface). “GSO” is the distance of GeoStationary Orbit (35,786km above the equator).
The Blue Marble photo was taken from far enough away to see a decent percentage of one hemisphere of the Earth. Opinions might differ, but I think farther would have been better. It’s close enough that distortion from perspective is significant — some parts of the Earth’s visible surface are 20% more distant than others (38400 km vs. 29000 km).
At BMP distance, the Earth has an angular size of 20.8 degrees, as shown by the dotted lines. (At GSO, the angular size is 17.4 degrees.)
Fast forward to 2015, when the DISCOVR climate observation spacecraft began operating. It does not orbit the Earth, but hangs out near the Earth-Sun L1 Lagrangian point, about 1.5 million kilometer from Earth. That’s over 40 times the distance of geostationary satellites. So it needs good telescopic instruments, but the nice thing is that it always sees the fully sunlit side of the Earth.
One of its instruments is the EPIC camera, which takes visible-light pictures. Here’s the first one released to the public:
It’s recognizable by the cloud formation that spells the word SEX (lower left, upside down).
You can get thousands of other EPIC Earth images at epic.gsfc.nasa.gov. EPIC stopped working in June 2019, and I assumed that was the end of it, but it’s been working again since February 2020.
The above photo is sometimes claimed to be the first fully sunlit non-composite photograph of the Earth since the Blue Marble in 1972. And maybe that’s true, depending on exactly what standards you use. It would mean that, before 2015, not a single geostationary satellite had ever been used to take a decent photograph of the Earth (at a suitable time: midday near an equinox).
It’s definitely not the case that there are no images of the sunlit Earth at all from that time period. It suffices to note that the Dish Network satellite TV service used to have a channel that was just a live video of the Earth, from one of its geostationary satellites. I think the channel was discontinued in 2013. This is a sample image from 2009:
At times near the equinoxes, the Earth was fully sunlit. Granted, the TV imagery was not very high quality.
Shortly after EPIC began operation in 2015, the Japan Meteorological Agency began operating a geostationary weather satellite named Himawari 8. Its data can be used to make high quality visible-light images, such as this one:
As with all geostationary satellites, all its images are of the same side of the Earth. You can usually recognize Himawari 8 images by the prominent position of Australia.
So, finally, somebody figured out how to get a geosynchronous satellite to take decent pictures. It only took 51 years.