I wonder if all you knowledgable people can help me find out how to calculate the corner coordinates of a lunar orbiter photo please?

I suspect it is a fairly basic astonomical calculation but I do not know how to do it.
I want to program (calculate for myself not find out from other websites) the corner coordinates of the old Lunar Orbiter 1-5 photographs, both medium and high resolution ones for a new web project I am starting.

I guess the formula for doing this has something to do with the height of the spacecraft, the angular coverage and the centre point of the photo. I have found this information and I know that the angular coverage of medium resolution photos taken with the 80mm lens is 44.4 degrees x 38 degrees and the coverage for the 610mm lens for high-resolution photos is 20.4 x 5.16 degrees.

I have found diagrams explaining it all but it is a long time since my school days and geometry was never my best subject, so I am hoping that some good soul has a formula somewhere so that I can just plug in the appropriate values and get the answer.

Any help would be appreciated.
Thanks. Paul

??? good question
this is just a guess . The first thing I would try

using ISIS3
http://isis.astrogeology.usgs.gov/index.html
download the lo spice data
"rsync -azv --delete isisdist.wr.usgs.gov::isis3data/data/lo data/
"
and import the PDS files into isis with lo2isis
from - http://pds-imaging.jpl.nasa.gov/search/search.html -
run "spiceinit "
and the corners will have the lat/long set

BUT i would also use qtie and qnet on the lo cub files to match them up to the NEW LOLA moon data

Thanks for your reply, however, I dont know how to use ISIS3 or any of the other things you mention( although I do know linux).

I really wanted to write a program to calculate the corners rather than getting someone else's program to do it for me.
If I can do that, then I can calculate which features are on which LO images.
I want to produce similar output to this page in the Digital Lunar Orbiter Photographic Atlas of the Moon but I want to go further than they have.

Hi - this is what you want:

http://nssdc.gsfc.nasa.gov/nmc/datasetDisp...o?id=PSPG-00160
You should contact NSSDC to see if they have the data in hard copy which could be scanned and OCR'd into a text file. See the email link at the bottom of that page.

It was my understanding that the Orbiter 1 data were missing, however. That's why the global photomosaic has a hole in it.

Phil

Thanks Phil, Thats great. I would have searched forever to find that! It seems to be what I am looking for.
Paul

I have found what I was looking for in excel format and in csv format.
In case anyone else is looking for this, I found it here
I hate it when people on other forums say they have found an answer and dont give the link where they found it!

This csv file is from here
This data says in it "The following list has been transcribed and transposed from the list of
Lunar Orbiter photos appearing in the book by Bowker and Hughes as it appears at:
http://www.lpi.usra.edu/resources/lunar_or...ok/table2.shtml "

I know that I do not know very much about this kind of thing, so please be gentle with me and allow me some lee-way. :-)
It seems to me that these corners are all over the place.
Could you briefly see if you can make head or tail of these please and set me on the right course?

To illustrate this, I have extracted the first 12 lines (from 178) the med-res corner part of the line and placed the frame number at the start.
The fields go
frame lat long lat long lat long lat long
1102 -46.57 113.57 0 0 -0.07 139.64 -41.08 127.43
1115 -19.44 125.39 32.87 122.27 29.48 -179.83 -23.83 160.99
1116 -18.4 134.3 35.49 131.32 25.37 -166.39 -23.91 172.83
1117 0 0 0 0 0 0 0 0
1136 -22.45 109.03 21.25 111.21 16.63 151.72 -25.98 145.33
1028 -29.5 -177.15 11.84 -169.64 8.27 -135.48 -32.21 -135.68
1030 -29.72 173.1 12.05 -178.79 7.81 -144.81 -31.96 -145.24
1035 -27.32 175.11 14.88 -179.11 11.43 -142.33 -30.85 -144.39
1036 -27.24 175.77 14.95 -178.6 11.76 -141.55 -30.91 -143.64
1037 -26.59 -178.72 15.64 -174.18 15.17 -133.99 -31.49 -136.7
1038 -26.52 -178.11 15.73 -173.68 15.64 -133 -31.57 -135.85
1039 -25.38 -168.67 17.5 -165.67 23.99 -108.39 -33.76 -118.42
1040 -25.34 -168.32 17.58 -165.36 22.16 -110.44 -33.92 -117.48
2196 -32.35 73.54 17.76 79.69 15.56 123.48 -34.55 124.8
2033 -32.48 148.02 14.72 155.06 12.26 -164.61 -34.95 -163.11
2034 -11.92 153.27 35.8 143.25 24.49 -153.77 -17.24 -173.11
2075 -48.59 111.79 -1.17 140.42 -0.68 174.27 -48.27 -156.88

I have worked out looking at these corner coordinates that there appears to be no pattern to any of these.
BL=bottom left, TL=top left, BR=bottom right, TR=top right
1102 BL=-46.57,113.57 TL=0,0 TR=-0.07,139.64 BR=-41.08,127.43
1136 BL=-22.45,109.03 TL=21.25,111.21 TR=16.63,151.72 BR=-25.98,145.33
1028 BL=-29.5,-177.15 TL=11.84,-169.64 TR=8.27,-135.48 BR=-32.21,-135.68
1030 BR=-29.72,173.1 TL=12.05,-178.79 TR=7.81,-144.81 BL=-31.96,-145.24
1035 BR=-27.32,175.11 TL=14.88,-179.11 TR=11.43,-142.33 BL=-30.85,-144.39
1036 BR=-27.24,175.77 TL=14.95,-178.6 TR=11.76,-141.55 BL=-30.91,-143.64
1037 BL=-26.59,-178.72 TL=15.64,-174.18 TR=15.17,-133.99 BR=-31.49,-136.7
1038 BL=-26.52,-178.11 TL=15.73,-173.68 TR=15.64,-133 BR=-31.57,-135.85
...
2196 BL=32.35,73.54 TL=17.76,79.69 TR=15.56,123.48 BR=-34.55,124.8
2033 BR=-32.48,148.02 TR=14.72,155.06 TL=12.26,-164.61 BL=-34.95,-163.11
2034 BR=-11.92,153.27 TR=35.8,143.25 TL=24.49,-153.77 BL=-17.24,-173.11
2075 BR=-48.59,111.79 TL=-1.17,140.42 TR=-0.68,174.27 BL=-48.27,-156.88
As you can see they are all mixed up - if I have interpreted the data correctly.
I am assuming that east longitude is positive and west longitude is negative,
so that a more negative longitude is further 'left' or west.

Have I totally got the wrong end of the stick? :-)

Do they match what you would expect if you plot them on a map? If you compare them with the printed index maps? - see links on the bottom of this page:

http://www.lpi.usra.edu/resources/lunarorbiter/


Phil

Phil, I am not questioning the values.
I am saying that there does not appear to be any consistency with respect to the value being the lat and long of a particular corner.
I would have expected for each pair of lat/long columns in the table to represent a particular corner. That is why I am wondering if my thinking is straight or whether I have misunderstood the system used.

I can sort them out programmatically, thats no problem at all, but for people who are following along behind and wishing to use this data in their own projects, I am enquiring whether what I suspect is in fact the case.

Understood! But still, plotting them might teach you something useful, including possibly explaining what's going on with the corners. Nothing teaches us about spatial relationships better than a map!

For instance, plot the corners of 1-102, the first line in your list. (it's the famous setting Earth image) The 0,0 coordinate pair is nowhere near in the right place. Maybe it was code for 'off the disk', though not a very good code since it could also be a real location. Also, looking at pictures, I see that the nearside images of LO1 are north-up, the farside ones are south-up (maybe a spacecraft orientation for thermal management issue).


So, seriously, map them - is my advice.

Phil

my thoughts too
it is very likely that the spacecraft was rotating while taking the images
so without mapping them to check , they would look a bit odd


a good example of something recent would be the Huygens images as it came down
those corners would be all over the place

The point is that there must be good data around somewhere, because the maps and indexes show the positions of the medium and hi-res images on a map of the Moon.

It is not very good for scientists to have to deal with half-correct support data when the completely-correct data is obviously available somewhere. It really does not matter for me because I am not writing peer-reviewed papers and doing research which matters to the scientific community.

I realise I may be raising an awkward question, but it should just be a matter of dumping a few fields of a database somewhere for the correct data to be available. That is hardly an onerous task is it and would take someone literally a few minutes of their time.

you do know just how old the lo data is
Lunar Orbiter V August 1, 1967
a few days after i was born

Yes, I know.However, if the data is there for the maps and charts posted on official websites, it should be available for our scientists too. AS I said, some people are probably relying on this data to be correct, so should we know if there is more complete set available? Isn't one of the base maps built on Lunar Orbiter photgraphs?

I am sorry, but I do not see what age of data has to do with accuracy and completeness of data?

I am trying to be constructive and to find ways to make our future understanding of this data set better, but this thread is turning more towards defending the data we have already been presented with.

well most of the data was really tossed out in the trash
there are even some real to real tape storage that have not been looked at since 1967
and are in the process of being read on OLD computers from 1967
http://www.nasa.gov/topics/moonmars/features/LOIRP/

I was going to suggest that you ask on the ISIS/GIS discission forum at USGS, but I see you have done that. I don't use this pointing data myself (I warp images to fit modern control using multiple tie points).

The Orbiter images were used in a new version of the lunar basemap at USGS, as you say.

http://astrogeology.usgs.gov/Projects/Luna...rameMosaics.pdf


My reading of that suggests that they fitted reseau mark locations to the ULCN and used them as tie points to warp the images to the required projection. No pointing data needed. I suspect that the old coordinate data which you have are only predictions based on the orbit data and viewing geometry. It would be reliable enough for the first exploratory maps of the far side (know to have errors as large as 20 or 30 km) and the published index maps, but not for modern mapping.

So I'm going to suggest you do an experiment to assess the quality of the data you have. Take one specific image from your list, one with all four corners on the disk. Find the image here:

http://www.lpi.usra.edu/resources/lunarorbiter/


Then, as accurately as possible, locate the corners (by comparing nearby features) on these maps:

http://planetarynames.wr.usgs.gov/Page/moo...edRelief?map=lo


Then, compare those four coordinate pairs (which will be in ULCN, the best coordinates we have until LRO data are fully processed) with the numbers in your list. If they are very close, your numbers are OK. If not, you may have to resort to the other approach and 'georectify' the images (warp them using many tie points) in a GIS.

What I'm describing will not take long for one image but it will teach you - and everyone else - something very useful. I could do it easily, but I have my own projects. This is not defending old data, it's being realistic about limitations of old data.

Phil

OK, thanks to all of you for your help on this. For my project, I do not think it is worth getting to know how to match up points on the images with other points on other images and to stretch them to fit. Besides, the project does not really involve placing images onto a larger map but more about placing features onto images - similar to what others have done.

I will do some experiments and probably use a 'best-fit' approach.