The quality of the image is still relatively poor as the main focus of the mapping camera was the lunar surface, and thus only broad patterns can be made out rather than precise detail. This problem is compounded by the lack of visible spectrum data from the satellite, but it it still possible to show that the area over Argentina is relatively clear (green arrow), and is bounded north and south by cloud (the red arrow identifies the northern cloud mass). The IR image and the visible spectrum images from the 14th both indicate an area of coastal cloud off South America's Caribbean coast (yellow arrow) that is absent by the 16th, and that forms a separate lobe of cloud to the Amazon cover.
The north Atlantic & Caribbean areas are difficult to discuss given the presence of a data anomaly over most of it in the visible spectrum images, but the IR photograph from the 15th does suggest light cloud cover over the former compared with thicker cloud over the latter. The blue arrow shows similar areas of circular cloud masses over the Caribbean. The IR camera does not pick up the much colder air masses over the Antarctic.
The partial view of the southern hemisphere in 3D corresponds well with what’s visible in the MMC image.
Despite the relatively poor quality view of Earth we have here, there is still a Landsat path with which we can compare it. That path was commenced at 13:06, 7 hours before the Apollo image, while a pass over Newfoundland was taken at 14:34. The latter pass is is in an area almost entirely covered by darkness at the time of the Apollo image so is not covered here. The details are given below in figure 4.9.64b.
Figure 4.9.65: AS17-
The crescent Earth is an early indicator that some time has passed since the last photographed examined here. The timeline tells us that the LM & CSM were re-
At the risk of carrying out exactly the sort of pre-
“We've been taking [Earth's] picture just as we came up”
Can we therefore find weather patterns on the crescent Earth that we would expect to find at that time and date?
Before going on to answer this question, it is first necessary to point out that only the IR spectrum image was available covering this time. No data were available for the 15th in the visible spectrum, which is the mosaic that would have been required to cover the early hours of the 16th. The time of the orbital pass at the terminator would be around 06:00 on the 16th.The most obvious feature to be seen is the large arc of cloud just on the terminator, and this can easily be made out on the IR mosaic. Once this is correctly picked out, all other weather features are easy to match up, and the we therefore have an an Apollo photograph taken 2 hours before even the very first part of the infra-
One of the best clues to look for here is the return of Storm Violet into the frame -
Figure 4.9.68a: GAP scan of AS17-
Figure 4.9.68b: AS17-
Figure 4.9.69: AS17-
If the colour version of the image is examined carefully, it is possibly to make out the coast of South Africa, and the fog shrouding that coast is marked by the red arrow. Land continues north of that fog almost all the way up the image, so we can be confident that our prediction for what should be visible is correct, and the images do show Africa.
As far as the NOAA images are concerned, the visible spectrum image (now available again) has the usual problem of half of Africa being commenced at the start of the day in question, and the other half 24 hours later. The IR spectrum however has no such problems. The times for these mosaics at the terminator are roughly 19:00 GMT for the IR image, and 07:00 for the visible spectrum. For the visible spectrum, most of Africa was started on the 16th, but the terminator area would actually have been imaged on the 17th. The Stellarium image confirms that Africa should be visible at the time of the photographs, but were they both taken at the same time?
Neither image shows the moon, so there is no reference point that we can use to say “this one was taken before that one”.
We can also conclude that the clouds shown in the central part of the image are the same ITCZ clouds, and the long strips of cloud in the top half of the image are over the Sahara in the mosaics.
If the 'V' shaped cloud picked out by the blue arrow is looked at more closely, it appears that the colour version shows it to be further from the terminator than the black and white one. This is matched by other surface and atmospheric features on both images, for example the long band of cloud shown by the magenta arrow ends shortly before the terminator in the colour image, but is coincident with it in the black and white. Likewise the distance between the coastal fog (red arrow) and the terminator is much less in the black and white compared with the colour. This would suggest that colour image was taken first. It also demonstrates that the Earth is moving as time passes, and not some static representation of Earth.
We also have a good pass over Southern Africa by Landsat, and the details are shown in figure 4.9.70.
Figure 4.9.67a: AS17-
Now we come to the tricky part -
Initial studies focused on South America as the land mass on the eastern limb. This is because the mission photo index records it as being taken between two lunar surface features -
What we need to do is trace through the records to see when these images were taken. We have this exchange at 08:22:49 in the transcript during rev 66:
"I took a picture of that one in Smythii -
That image is AS17-
A couple of hours later, recorded as 09:01:46 in the transcript, we have Schmitt stating that:
“Bob, that's the most beautiful crescent Earth I've ever seen.”
“That bright spot. See it -
If you look at the photograph you can see that there is indeed a bright solar reflection in the centre of the Earth crescent, and it seems likely that Schmitt was impressed enough at the time to photograph it. Shortly after this they then enter a rest period and there is no communication recorded with the ground. By way of confirmation that he photographed Earth, we have this remark from the Command Module transcript as he continues to discuss Earth’s crescent:
"Oh, darn. There's the picture. Oh gosh. Never even thought about it. Oh well. Remember it, gang."
So where has the confusion come from? It would appear that it comes from the crew relating back details of photographs taken during the preceding hours when during their rest.
For example at 09:12:49 and 09:13:04 we have Evans saying:
“Frames 163 and 164 and 165 were taken of the mud craters and Smythii...that's mag November November.”
“Frames 166 and 167 were taken of a crater...to the west of Sulpicius Gallus.”
We also know that the Schmitt wanted specifically to take the Sulpicius Gallus area much earlier, as recorded at 08:20:55
“I'm going to try to also shoot, if you'll let me, two pictures on KK of that depression -
Again, if you read the transcripts at around this time you will find many references to using KK instead of NN (magazine 148), and there was obviously some confusion between the two.
The obvious conclusion is that the image of Earth was not taken during rev 73 but several orbits earlier, and the way the information has been given by the crew has led to them being mis-
Having decided when we are seeing the image, we can look at what’s in it. Again we are restricted to an infra-
We also have a Landsat pass over Australia on this date, and we can see this in figure 4.9.67c.
Figure 4.9.71: GAP scan of AS17-
The NOAA satellite would have imaged the area around the terminator at around 13:30 GMT.
The EVA commenced at about 20:30 on the 17th, ending just over an hour later. Evans makes several references to the crescent Earth and mentions wanting to take photographs of it, but doesn't mark specifically when photographs were taken Although the photograph used (featuring one of the cameras that recorded the event) appears relatively early in the EVA sequence, a mid-
The most obvious weather pattern, and one that can definitely be identified, is the large 'tick' shaped mass in the northern hemisphere, partially obscured in the NOAA mosaic by data errors. This band, the cloud mass to the north of it, and the gap between them, are easy to pick out. The Amazonian cloud masses and south Atlantic weather bands are also simple to locate once the correct orientation of the land masses at the time of the EVA is determined.
A few photographs taken during the EVA show Evans with the Earth in the background. Unfortunately these are all, by and large, showing a very small and out of focus Earth and there is very little detail to seen in them. In the interests of completeness (bit little hope of conclusiveness), the image with the best view of Earth, AS17-
Figure 4.9.73: GAP scan of AS17-
The view of Earth presented in figure 4.9.73 is so blurred there is little point in doing anything more than identify areas of light and shade. We know roughly when the image was taken, we know roughly what should be there but it would be unwise to state definitively what is there.
What has been picked out is the high latitude cloud mass (green arrow) to the north of the 'tick' shaped band (blue arrow), and there is a suggestion of darker blue between those two masses. The location of the Amazonian clouds is pointed out (magenta arrow), and again there is a suggestion of a gap between that and the blue arrowed cloud to the north that may well be the Caribbean Sea. Whether these are reasonable conclusions to draw is left to the reader to decide.
At the end of this magazine come the final pictures of Earth. There are half a dozen of them, seemingly in sets of 4 and 2, and they follow a number of photographs of the moon. Close inspection suggests that the first 4 were taken about an hour before the last 2, and this will hopefully be demonstrated by the following. Interpretation is difficult for these remaining shots in that it’s difficult to make out any land features with any certainty, there are no references at all in the transcripts, the satellite record is somewhat patchy, and while we are still using the view from the moon as a reference point the crew are now nearer Earth than their lunar embarkation point, so the width of the crescent is not necessarily as reliable as it was when it comes to determining the date of the photograph.
The first of two images, one from each set, is shown below and is AS17-
Figure 4.9.75a: GAP scan of AS17-
The key to identifying when this picture was taken lies not so much in the cloud patterns as in what can be seen on the surface. The glare from the sun and the subsolar point shows that there is a considerable amount of ocean on view, but we do also have evidence of land. We’ll deal with that shortly, but first let’s try and work out where it could be. By looking at various astronomical simulators, we can work out that at this position on the Earth’s crescent at this time of year there are only three locations that show land masses -
Figure 4.9.77: AS17-
Figure 4.9.78: AS17-
As before, the thickest part of the Earth's crescent shows the coast of central America, but in this instance it has rotated further round and much more of it is visible in the photograph. An estimated time lapse of around an hour has been used, and this does seem to correspond with the change in the amount of central America visible in the two Apollo images.
In broad terms the weather systems are still the same, but they do not necessarily appear in the obvious place, thanks to the orientation of the Earth and its rotation.
As before we have the large cloud mass over the USA (blue arrow), and the long 'tick' shape (green arrow) ends in the gulf near the central American coast. The red arrow picks out the cloud over northern Amazonia.
The magenta arrow picks out what appears to be coastal fog, which would explain why it does not appear on the IR image, either because it was ephemeral in nature or because the cool air mass that would produce this fog is less visible to IR cameras. The yellow arrow is really only there to pick out features within those coastal fog banks. The cyan arrow points to the curve of the top end of a cyclonic south Atlantic system. The times for the NOAA mosaics would be an hour further on than those for the previous image, assuming that the time gap between the Apollo images is correct.
A final image from Apollo 17 can be found in the 16mm footage. After the example given in figure 4.9.74, there are a couple of slow pans taken of a crescent Earth occurring in the midst of long distance shots of the Moon. Examination of the lunar images shows that, like the previous 16mm image used, the footage is reversed, and as is demonstrated by the brief analysis given elsewhere on this site (Sideways on) the view of the Moon is not quite what would be seen from Earth, thanks to the position of the astronauts in cislunar space.
By taking several screenshots of the pan, a full version can be assembled, and this is shown below in figure 4.9.79.
During orbit number 62, the first 55 frames of the Metric Imaging camera use an oblique angle to take images of the lunar surface, capturing the lunar horizon as they do so. As this orbit reaches Mare Smythii, the camera captures a sequence of images showing Earth rising. The ASU have animated that sequence here. Unlike other Earthrise sequences, this capture is entirely coincidental: the photography is automated, and it just so happens that Earth is in the right place, as opposed to other sequences where the astronauts have set out to photograph this orbital feature.
The image chosen from this sequence is AS17-
Figure 4.9.62: Apollo 17 Metric Mapping Camera image AS7-
The distortion introduced by the oblique angle can be corrected using image editing software, so that the position of landmasses can be better approximated in comparison with Stellarium. It must be stressed that the only alteration carried out here is to alter the brightness and contrast of the image, and to make the crescent shape of the Earth less distorted.
The first step to comparing with satellite images is to determine when this image was taken. Orbit 62 was commenced at 206 hours and 14minutes on the 15th. The mission transcripts record a conversation before the commencement of this orbit marking the opening of the Mapping Camera before LOS at the end of orbit 61. There is also this conversation as the crew approach Mare Smythii (visible on the lunar horizon in figure 4.9.61) on orbit 62 at 206 hours 41 minutes:
CDR: There's ... Oh, fantastic. That's the first one I've seen. Out 5. Beautiful, beautiful. Look at that ... Look at that -
CDR: Yes. Let's -
CMP: I don't know, I've looked …
CDR: Yes .... a little late. We needed it just when it came up. Yes, let's make sure we do that. That is beautiful. It's just unbelievable
This would give a time of the Earthrise on orbit 62 of around 20:40 on December the 15, during which time they are recorded as undertaking lunar orbital science.
Stellarium suggests that what should be in view at that time is South America, with some parts of the southern USA also visible, cloud cover permitting. The crescent Earth's size is a good match for that shown by Stellarium, and although the original has been changed, the dimensions of the widest point are still consistent (see figure 4.9.63).
Figure 4.9.63: View of Earth from AS17-
Figure 4.9.64a shows the Earth from the mapping camera image in comparison with a number of satellite images. Unfortunately only IR data are available for the 15th, and that satellite image is shown on the left. Visible spectrum data from the 14th and 16th are shown to the right. The times along the terminator are around 11:30 and 23:30 for visible and infra-
Dating this image precisely is difficult. There are no specific references to filming in the transcript or mission reports. The width of the crescent suggests it is from roughly the same time as the preceding few images, and it must be from a later time than that suggested for figure 4.9.74, but there are no recognisable features showing that are visible in figure 4.9.76. As there is a degree of overlap between figure 4.9.75 & 4.9.77, it is reasonable to suggest that it was filmed slightly later than those two, but not so much later as to diminish the crescent’s size.
There are two features that suggest a possible time for the footage, and this is indicated in figure 4.9.80 below using a level adjusted version of 4.9.79.
Figure 4.9.80: Level adjusted photomontage from 16mm footage compared with NOAA image from 17/12/72 and Fourmilab estimate of terminator at 00:30 on 18/12/72.
The analysis presented in the figure above relies mostly on the assumption that the green and magenta arrows are correctly identifying cloud patterns off the western Americas. The blue and red arrows are useful guides to cloud free areas that are consistent with that suggestion, and other features also seem favourable to the idea that the film was taken very earlier on the 18th.
As with other images of a similar nature throughout this analysis, the reader is welcome to come up with their own suggestion.
No more full images of Earth exist for Apollo 17, and the next stop for the crew is the final re-
As with other missions,. That re-
Figure 4.9.64a: Close up of image AS17-
Figure 4.9.66a: AS17-
Figure 4.9.74: AS17-
Figure 4.9.79: Photomontage of stills from 16mm footage.
The first part of the question we need to ask here is ‘do we have the locations correct? For the South American pass we can make out the weather patterns more clearly as well as the angle of the western coast along Peru and Chile, so it seems reasonable that we have. The clouds here are relatively easy to interpret -
The crew continue to orbit the Moon for some time after this image, and it is another 4 orbits before the Earthrise image they want to capture is taken, using magazine 152. This magazine starts with that Earthrise sequence which covers images AS17-
As before the angles are oblique, and several hours have elapsed between the Landsat images and Apollo (South Africa was photographed at 08:05), but again we have a broad consistency between the two. We have clear air over most of the areas covered with some coastal cloud.
That’s the last of the images from lunar orbit -
4.9.1d – The last leg
During TEC, Ron Evans, the command module pilot, carried out an EVA to retrieve equipment from the CSM SIM bay. During that EVA, several photographs were taken that show the Earth, including the one below in figure 4.9.71, analysed in figure 4.9.72, AS17-
Figure 4.9.64b: Landsat paths plotted on Google Earth (top right), AS17-
Figure 4.9.70: Landsat passes recorded over southern and central Africa as shown on Google Earth (bottom left) and AS17-
It would hardly be reasonable to draw any solid conclusions from these images, but there we are -
Moving on from the confusion over image timings after the Earthrise sequence in magazine 152 shown earlier, there is an Earthset, comprising AS17-
In addition to photographs from magazine 152, we also have a contribution from 139, this time in black and white. Magazine 139 was taken to the lunar surface, and the first half of the photographs on it are those from an EVA. AS17-
Again, we can use the Photo Index and mission transcript to act as a guide as to what part of the Earth's surface is appropriate to start looking, and we find that the photos are recorded as being taken on orbit 71.
The mission transcript has this to say towards the end of orbit 71,
“...we're going to get your picture as you set this time."
LOS on orbit 71 is recorded as 225:45 MET, or 15:25 on the 16th.
Both these images are shown overleaf in figure 4.9.68 a and b, and they are analysed together in figure 4.9.69.
Figure 4.9.67b: AS17-
Figure 4.9.67c: Landsat image paths shown on AS17-
Figure 4.9.76: AS17-
The Earthview depiction shows what should be visible at 22:00 on the 17th. If the photograph was taken on the 18th, the crescent would be much thinner. Zooming in close to just north of the thickest part of the crescent reveals, faint but still detectable, the thin lines of the central American coast, and from there the bulge of south America becomes clear. Having identified those landmarks, placing the clouds becomes relatively straightforward.
The westward end of the 'tick' shaped cloud discussed previously is positioned over the Gulf and central America, and the clouds of northern Amazonia are also visible The glare on the western horizon shows the subsolar point is over ocean and not land. The video still appears to show a degree of additional rotation, and is therefore taken slightly later than the photograph.
As with the previous images taken around the EVA, NOAA would have imaged the terminator region on this image at around 13:00 on the 17th. The IR mosaic equivalent was imaged at around 01:00, and is therefore nearer to the time of the image than the visible spectrum one and improves the match between the two.
The second image is AS17-
Figure 4.9.75b: Areas of the globe with a subsolar point over ocean and land in the southern part of the crescent depicted using Worldwide Telescope.
Australia seemed like a promising candidate because of the apparent line of a coast to the south and north but it is difficult to tell whether this is something that has been introduced more by manipulating the image rather than something that is actually there. Africa has far too much land around the subsolar point, and not enough clouds to the north when comparing it with the satellite image. Africa, like Australia, would also be several hours further on than South America, and while intuitively one would imagine that the Earth crescent would be getting larger, we have these observations from the Cernan as they get nearer Earth:
"...the crescent is getting smaller and smaller although the moon is getting -
There’s another observation much later by Schmitt, made at 286:18
"For the first time in seemingly several days we see the Earth"
Which does seem to suggest that they hadn’t observed the Earth much after the series of photographs we are examining here, and that those photographs were taken not long after the EVA.
To bolster the suggestion that the area we are examining is focused on the Americas, we have this area visible in the Apollo image (figure 4.9.75c).
Figure 4.9.75c: Comparison of Central America shown by WWT and in AS17-
There isn’t anywhere else this could be, so that’s the area on which we need to focus. We can now try and look at the weather satellite data in figure 4.9.76.
Figure 4.9.66b: DMSP image of Tropical Storm Violet compared with a section of AS17-
If you care to check the location of the storm in the Apollo image and compare it with the coordinates given, you’ll find they match, just as the storm itself does perfectly.
The next image presents us with something of a dilemma, for reasons that will become apparent. It is a return to magazine 148, the final Earth image in the magazine, and is shown in figure 4.9.67a. It is analysed in figure 4.9.67b.
Figure 4.9.72: AS17-
Figure 4.9.81: Re-
We know that re-
Figure 4.9.82: Selected screenshots from the re-
The view is roughly towards the north, looking backwards as the CM heads for the Pacific, and we can look at what the satellite imagery shows for that area (figure 4.9.83).
Figure 4.9.83: NOAA-
As with other re-
This broadcast of the splashdown shows the broken cloud visible immediately below them when the drogues are released (figure 4.9.84)
Figure 4.9.85: the scene at the landing zone.
Again, not proof as such, but the satellite imagery (such as it is), the footage from the CM itself, and that taken on the ground, are entirely consistent with each other.
So there we are, Apollo 17 safely back to Earth. Time now for a quick look at the synoptic data.