Terrain modelling
The “Apollo Detectives” youtube channel (I’m not linking to it here, they don’t deserve the traffic) recently did a video examining this paper.
This is the abstract:
“The Apollo program was NASA’s (National Aeronautics and Space Administration) human spaceflight program, accomplishing landing of the first humans on the Moon from 1969 to 1972. Ever since there have been scientific and public questions about its legitimacy and claims that the associated Moon landings were staged by NASA and/or other organizations. In this paper we examine a number of the Apollo mission images through (i) a comparison with simulated views of Google Earth (Moon) and (ii) a photographic analysis of some of their features using Photoshop®. The functionality of latitude, longitude, elevation and elevation profile of Google Earth is addressed by means of a comparison with other available programs possessing the same features like NASA’s Moon Trek and Alcyone Lunar Calculator. The topographic analysis through Google Moon simulations indicate that the landscapes in Apollo mission images used were inaccurate presentations of reality and there are incorrect elevations and serious land feature omissions. Moreover, the Photoshop® analysis shows conclusively that images were staged, manipulated or altered.”
They make some bold claims, so it’s only right that those claims are examined. We’ll examine that with a little more attention to detail than the authors, or the peers that supposedly reviewed this paper, gave it.
The basic premise of the paper is that some views in Google Moon don’t look quite like those in Apollo panoramas. They start their work by justifiying the use of Google Moon as an appropriate tool.
Although available as a web interface before then, the Google Earth interface to the moon’s globe was added in 2009 to coincide with the 40th anniversary of the Apollo landings.
This page, and this paper, state that the topographic date used to produce 3D terrain views was sourced from Japan’s ‘Selene’ probe, as the Lunar Reconnaissance Orbiter wasn’t then in operation. Higher resolution photography from the LRO was added to the Apollo sites later, but there’s no information as to whether the surface model data was changed. Japan’s data claim an accuracy of between 3 and 5m vertically and between 5 and 8m on the ground (see here). That global map can be found here.
The paper starts off with a potted history of the Apollo project, but sets its stall out early by claiming that the photographic, live TV and 16mm film evidence is “largely questioned”.
No. It is not. Some people question it, and largely ignore the answers. The overwhelming majority of people are more than happy with the evidence as presented.
They then outline Google Earth. They express disappointment in the quality and accuracy of most of the lunar surface and length of the lunar day is incorrect, and make this point:
“The aforementioned facts along with the fact that by now a very high-
This is absolute nonsense. The fact is that accurate models of the lunar surface are freely available to the public, and from a variety of sources. As a member of the public who has availed himself of them I know this for a fact. It’s surprising that a supposed expert here doesn’t know that. It’s also telling that they don’t like the accuracy of the surface model, except where it suits them: the Apollo sites.
They then attempt to confirm that Google is an appropriate tool to do their ‘analysis’ by comparing it with other sources. They include in this the Chang’e-
They also use an Apollo 17 topophoto map (a contour map superimposed on an orbital surface image) to validate Google. They fail to point out that this map was generated using a combination of Apollo 15 mapping images and Apollo 17 panoramic camera images.
Weird that they left that out, right?
They then get to the meat of their discussion, the images themselves. They look at actual Apollo panoramas as presented in Google Moon, at their own reconstruction of those panoramas, and reconstructions using Google Moon’s own terrain model. So far so good. Here’s how they present it.
What they do next is point out some odd features in the Google Moon simulation, saying:
“It is difficult to explain why the irregularities shown in figure 5 are not recorded on the actual photos, unless Google Moon elevations or the indicated position of the photographer are wrong.”
There’s a big clue there boys. You should maybe check if that’s true. They didn’t, but I will. Superimposing a labelled LRO view on Google Moon;’s version gives you this:
The black circle identifies the location of the panorama in Google Moon, and it is clearly incorrect.
To be fair, in this case it’s unlikely to have much influence on matters -
They do the same with another panorama, this time one taken the landing site itself, and this one at least is accurately plotted, and they do this to see if the same ‘inconsistencies’ can be found. Let’s see what they’re identifying.
These features aren’t visible in the Apollo photos, but it doesn’t seem to occur to them that the reason for this is that they aren’t actually there, and that they’re a product of the inbuilt inaccuracy of low resolution data. Let’s put those irregularities into some regional context.
Hopefully it’s evident that the ground is much more uneven in these views than is presented in the Apollo photographs, but that, in part, is the author’s point. Is it universally true depending on the model used? Well, no. Here’s the view north using different models -
Looks nothing like as uneven here. How about some nice oblique views of the site using LRO and Danuri (South Korea’s probe) images?
Sure it’s uneven, but is it as uneven as presented in the Google Moon model? No.
To illustrate their point, the authors pick two mounds identified here.
Which their calculations suggest should be 3m high. )R 5m. Depending on which one they want to believe.
You might want to remind yourself of the accuracy level of the underlying JAXA model here.
Now, you would think if such a thing was present, then they would be easy enough to spot in the images taken from orbit -
Next to it is the oblique LRO, and while the degree of shading isn’t as great, the relatively shallow craters have some. There’s no such shadowing where these hills are.
In the overhead view, the 3 metre high LM descent stage (bottom centre) casts a clear long shadow. Anything 3 metres tall elsewhere should be doing the same.
There’s no such shadow at the location of the supposed hills.
The authors also look at a walking traverse map of the area available to the astronauts at the time. Their conclusion is that the material available to them at the time showed only basic irregularities, implying that they didn’t know about these lumps and bumps shown in their preferred model. The corollary of that argument is that the could not have known about all manner of details that the Apollo image record shows in great detail!
For their next trick they pick taken during Apollo 16 looking towards South Ray from a spot near Cinco crater. Their claim is you can’t see South Ray crater from this vantage point. There’s just one small problem: the marker on Google Earth is completely wrong, and they didn’t bother checking.
We can demonstrate this by looking at the details for the panorama included in Google Moon. Below left is the relatively low resolution imagery in Google Moon superimposed on the LRO view. The black circle identifies the viewpoint the authors have used. To the right is a labelled map from the ALSJ.
The panorama in question is this one, which the ALSJ (a source they quote and with which they must therefore be familiar) states consists of AS16-
The diagram above right shows where those photographs were taken, and that location (a circle south of the LRV location) is inside the black square on the image above left. It is very obviously in the wrong place. Google Moon is wrong, and no amount of the visibility analysis they do will compensate for that. You can even position yourself on the LRO quickmap and see that the view is correct.
They then move on to Apollo 17, this time identifying pronounced mounds behind Bear Mountain. You can see the mounds in their paper, and I’ve generated a different view to show that Google Moon does, indeed, show them.
The problem is that those mounds simply aren’t there -
You also don’t see them in any of the models produced below using CNSA (top left), JAXA (top right), ISRO (bottom left) or LRO (bottom right) data. See here for details on those.
All those models were made using freely available data that the authors could have used themselves. They did not. They went with what confirmed their bias.
They conclude from this that the Apollo photos are of a model, and compound their level of stupid by showing Google Earth failing to reproduce a terrestrial scene. Almost as if it’s not entirely accurate.
They then briefly mention the poor resolution of the lunar orbiter images available to the Apollo mission planners, stating that this poor resolution (ie, nothing smaller than 20m was recorded) was the reason:
“small land features are not well recorded on photos”
This is, not to put too fine a point on it, bullshit.
Small land features are clearly recorded on Apollo images, those features are confirmed by modern probes but were unknown at the time. Let’s have a look at what we see of the views in Google Moon draped in Lunar Orbiter photographs for Apollo 15 (again, something they could have done).
Compare that with the same view using LRO data and the SEVA panorama taken by Scott.
And for the avoidance of doubt, let’s identify some of those ‘not recorded’ features.
Good luck finding those on any pre-
There then follows a load of nonsense about spotlights that is pretty much just them seeing what they want to see, but they make this comment concerning two photos from Apollo 17 where the camera was rotated 90 degrees:
“But why would anyone need to rotate the fixed on the chest camera (if that was possible anyway) with a square film format?”
The question over the camera being fixed to the chest shows their woeful ignorance. This image from the time of the photo sequence shows Cernan not only with the camera detached, but at an angle.
The final part of their delusion comes from them trying to claim that an Apollo 8 image has a hair grip on it.
Here’s their artefact compared with an actual hair grip.
It’s quite obviously a fibre, and even if it was a hair grip, an object 2 inches long would make the photograph huge!
They also query whether the other white dots in the photograph might be stars, which, as supposed photography experts, they should know they can’t be.
In summary: their conclusion that Google moon is the best source of terrain models is flawed at best, dishonest at worst. They have not checked the veracity of the Google Moon model, or its data sources, nor have they compared it with any other sources that were definitely available when they wrote their embarrassment of a paper. The irregularities they find do not exist in the Apollo record, or indeed anywhere else, for one simple reason: they don’t exist. Even JAXA’s own data, used to underpin Google Moon’s model, contradict their findings.
Their assumptions are flawed, their exploration of alternative explanations is lacking, and their analysis is based on a priori reasoning -