Creation of False Color Ratio Maps of Lunar Features:

Producing your own images:

     The geology and mineral composition of the moon can be studied with a small amateur telescope and multiple broadband filters.  I use broadband filters with band centers at 400, 450, 500, 550, 600, 650, 700, 750, 780, 850, 905, 950 and 1000 nm with average bandwidths of about 30-40 nm.  Silicon based CCD cameras are not typically sensitive much beyond 1000 nm. The images must be very carefully aligned to each other to subpixel accuracy.  This would be difficult to do manually manipulating the images in a program like Photoshop. A program specifically designed for this purpose, like Mira Pro 7 can be used if 16 bit manipulation is required, however a freeware program like Blink Comparator (see links submenu) serves very well for 8 bit images. 

 Image Calibration:

         A good way to begin is to image a lunar feature using broadband filters centered at 415, 750, 900, 950, and 1000 nm. These are commonly called 5 filter images. They should be obtained at a phase angle of between roughly 20 to 50 degrees (i.e. not to far from full moon) because of phase reddening artifact that occurs at larger phase angles. The Apollo 16 landing site must be imaged first (medium power or high power is fine but one selected it must not be changed during the imaging session). Optimal exposure times can be determined for each of the five filters while imaging the Apollo 16 landing site. Exposure time will necessarily increase as the filter wavelength increases. However, the exposure times selected for each particular filter must be used for all lunar features subsequently imaged during that imaging session.

     Images are calibrated by dividing filter images of the lunar feature of interest by the average greyscale value of the Apollo 16 landing site (an area of about 20 square kilometers can be selected for greyscale averaging at the Apollo 16 landing site). They are then multiplied by the value recorded for the filter center wavelength (with 1.0 added) available on the USGS PDS node and listed as the Adam's Apollo 16 specimen 62231 datafile. This is available at: http://pds-geosciences.wustl.edu/missions/lunarspec/index.htm. This data is directional hemispheric data and is used to calibrate telescope images.  Bidirectional reflectance data is also available at this site and is used to calibrate lunar probe data. It is possible to use the bidirectional reflectance data provided that the telescope images are converted to bidirectional reflectance at standard viewing geometry (e=30, i=0). After calibration, the images are co-registered using a program like Blink Comparator or MiroPro7 so that when superimposed on each other the lunar features will be in perfect alignment. It is also possible to use Photoshop for image co-registration, but it must be done very accurately. Calibrated images that have been co-registered can be used to prepare ratio images as described below. The resulting ratio images will be very similar to those obtained from calibrated Clementine images (except that their resolution will naturally be lower). Ratio images produced can be fine tuned aesthetically by slight adjustment of contrast and color saturation of the final image.  One word of caution however... Clementine images were of small areas of the lunar surface and when larger mosaic images were made from Clementine data, an absolute calibration taking lunar curvature, surface roughness, and other factors  into account had to be performed.

 Ratio Images:

      A ratio image is produced by dividing one image by another.  This pixel division between must be done at 32 bit accuracy in Mira Pro 7 if 16 bit images are used.  The division can also be done using the Image Calculator function under the Process menu of the free program ImageJ (see rsb.info.nih.gov/ij/ ). Pixel divisions are made using the various broadband filter images, which should be of similar brightness and contrast following calibration (see above). Commonly used ratio images for studies of lunar geology and mineral composition are 415/750 nm; 750/1000 nm; and 750/415 nm. This scheme is called the maturation composite ratio image because if provides information on lunar soil maturation (and limited information on mafic composition). Another useful scheme is to assign 750/900 to the red channel, 750/1000 to the green channel and 750/950 to the blue channel.  This is known as a mafic composite ratio image because it is designed to show subtle difference in mafic composition. Channel assignments of ratio images are made using a program like Photoshop that allows copying of the individual ratio images to the color channels of a blank color image.  The result is a false color composite ratio map of the lunar feature. Depending on which channels are assigned to specific ratios these channels will impart a specific color to lunar features with certain mineral characteristics such as abundance of iron or titanium.  Iron absorbs light preferentially around 1000 nm while titanium absorbs preferentially at around 415 nm. Thus both maturation and mafic composite ratio images will show strong iron based mafic signatures as being greenish.  Mafic materials with high titanium content (like clinopyroxene or ilmenite) will appear bluish on a mafic composite ratio image. Low titanium pyroxenes such as orthopyroxene (like norite) will appear reddish, violet or purplish on a mafic composite ratio image. Ancient exposed lunar surfaces can also be differentiated from more recently created surface features based upon their increased concentration of glasses as a consequence of long term micrometeorite bombardment. Such surfaces appear reddish or orange red on maturation composite ratio images. Pyroclastic flows containing orange glasses appear deep red on maturation composite ratio images. 

     A low power maturation ratio image of Tycho created from images taken through my 9.25 inch Schmidt Cassegrain telescope at prime focus and calibrated by the method described above is compared to a Clementine maturation ratio image of Tycho obtained from the USGS PDS site:

               Tycho: 9.25 inch SCT (maturation image)

              Tycho: Clementine maturation image

     The two images are quite similar indicating that the calibration of the 9.25 inch telescope data was adequate. The mature lunar highland soil is distinctly red whereas more recently resurfaced areas are blue. The lunar highlands are mostly composed of light colored minerals called felsic minerals which are plagioclase feldspars. By far the most common type is anorthosite.  Fresh anorthosite outcrops show high surface albedo and can be detected by comparing an albedo image with a ratio image of 750/1000 nm. The 750/1000 nm image is useful because it is very efficient at eliminating albedo effects. Although the lunar highlands are mostly composed of anorthosite, plagioclase/mafic mixtures are found in many gabbro outcrops and in crater walls and central peaks. Crater impacts lead to the rebounding of deep crustal mafic layers to the surface in these areas.  Examples include the minerals noritic anorthosite, norite, anorthositic gabbro, gabbro, troctoliteic anorthosite, troctolite, olivine and dunite. To study this mineral composition it is necessary to calibrate the filter images as discussed above. 

Mafic Ratio Images of Tycho (red=1000 nm, Green=900 nm, Blue=415 nm):

                                 9.25" SCT with interference filters (Mafic Ratio)

                                Clementine Mafic Ratio Image

     Lunar minerals can also be studied at infrared wavelengths but most silicon based cameras are not sensitive beyond about 1000 nm, although some have limited sensitivity to about 1200 nm.  I use a Hitachi KPF2A extended NIR camera to capture pyroxene absorption at between 900 nm and 1100 nm. This camera has good sensitivity between 400 nm and 1200 nm, and has a peak response at about 750-800 nm. Cameras with good imaging sensitivity and a wavelength range extending to 2200 nm are extremely expensive and beyond the reach of most amateur astronomers. The usefulness of having mid range infrared imaging available is that this makes it possible to better differentiate olivine from pyroxenes.  Pyroxenes absorb at both 1000 nm and 2000 nm. Olivine absorbs at 1000 nm but not at 2000 nm. Again, to be of maximum use, the images would require calibration.

     Amateurs interested in geologic and mineralogic studies of the moon, but not prepared to take their own filter images or to face the challenges of an absolute or a relative image calibration can simply use calibrated and co-registered Clementine UVVIS images at 415, 750, 900, 950, and 1000 nm available at the USGS PDSMAPS website (see discussion below).  They can also choose to observe the effect of saturating color images of the moon taken with a digital camera.

Using Color Saturation and a Color Digital Camera:

     Amateurs wishing to explore the mineral composition of the moon without the expense of using broadband filters or special cameras can try color saturation imaging using a regular color CCD, DSLR or videocamera (or by using red, blue and green filters with a monochrome camera). For this technique the red, blue and green channels must be balanced so that each contributes equally to the final image.  This must be done before the image is color saturated.  A program like Photoshop CS2 is adequate to perform both functions (use the Auto Color command to balance the color channels and then increase the Color Saturation slider under the Hue & Saturation submenu until the desired level of color saturation is reached).

Working with Calibrated and Co-registered Clementine Data:

     It is also possible to create false color, single ratio images, and the more complex maturation and mafic composite ratio images and other types of false color images using existing calibrated and co-registered data from the Clementine probe available on the USGS PDSMPDS website at: http://pdsmaps.wr.usgs.gov/maps.htm .    To demonstate the effects of image calibration, false color and ratio images made from the calibrated Clementine dataset can be compared to results obtained from raw uncalibrated Clementine images available in CD-ROM format from the National Space Science Data Center http://nssdc.gsfc.nasa.gov/cgi-bin/shop/web_store.cgi  .   A more complete discussion of creating ratio and false color images from Clementine data is available in the third issue of the on-line journal Selenology Today  (Exploring False Color and Color Ratio Images Using Clementine UVVIS/IR data): http://digilander.libero.it/glrgroup/  .