Telescope Reviews: LCROSS Project

Using photometry to find skybrightness above the LCROSS impact site and Brightness of Impact Cloud

In the spring of 2009, the LCROSS Team issued a single frequency estimate at 550nm of the impact ejecta curtain brightness of approximately 3.0 mpsas during the first between 10 and 60 seconds after impact. Heldman (2009). A version of Heldman (2009) is reproduced as a low-resolution frame on the LCROSS Team Observation Campaign webpage (2009). url: http://lcross.arc.nasa.gov/impact.htm . For amateur CCD or video imaging purposes, this understates the likely mpsas, since CCD cameras respond to the broad integrated wavelengths between approximately 380nm and 1000nm. CCDs do not respond at a single wavelength.

The LCROSS-EDUS impact will generate a dust ejecta curtain about 10 km x 5 km high between 40 to 60 seconds after impact according to the NASA-LCROSS Team conservative best estimate impact model (CBEIM). Bart(2008) at slides 19-20; Wooden (2008) at slide 2. This current conservative LCROSS CBEIM translates to a 3 arcsec x 2.5 arcsec ejecta dust cloud at a mean lunar distance of 384,400km. A larger hypothetical 20 x 20 km ejecta curtain converted to a circular shape of equal 400 sq km area has a diameter of about 12 km. As seen from Earth at a mean lunar distance of 384,400 km, the 20 x 20 km square billboard would be 11 x 11 square arcsecs, or equivalent to a circle with a diameter of 12 arcsecs.

At those ejecta curtain sizes and an irradiance at a single wavelength of 550nm of 3.0 mpsas, the corresponding integrated magnitudes of the curtain would be 1 integrated magnitudes for the 10km x 5km scenario and -1.9 integrated magnitudes for a 20km x 20km, 12 arcsec diameter scenario.
This apparent brightness (3.0 mpsas) provides a favorable contrast index against typical Earthshine (dark limb) irradiance between 12 to 17 mpsas for the dark limb (mean value 15.44 mpsas) but not against Moonshine (bright limb) at 4 to 6 mpsas.

The LCROSS ejecta curtain will present a uniquely difficult extended object for amateurs to image. The impact is hoped to occur when a permanently shadowed region (PSR) is on or near the bright limb terminator with the crater surrounding the PSR being partially lit at a very low sun angle. The dust ejecta curtain will rise into the sunlight and reflect light back to Earth-based telescopes against the background darkness of the night sky immediately adjacent to the dark limb.

At the moment of projected impact, 2009-10-09 11:30UT, the Moon will have the following geocentric empheris characteristics:

Phase angle 65.2984
Perscent illumination 71% illumination
Terminator colongitude 158
Lunar age 20.7 days

The average sky brightness under moonlight scattered illlumination (expressed in mpsas) at various distances from the Moon was modelled with precision by Krisciunas and Schaefer (1991) based on earlier measurements by Krisciunas (1990) and an earlier rough model by Walker (1987). See Krisciunas (1990) at Figure 8, reproducing the rough model of Walker (1987).

The effect of moonlight is additive. Under the rough Walker model, the full Moon adds about -1.9 mpsas to whatever your base sky brightness is. At a lunar age of 20 days, about -1 mpsas is added. If you are observing from a light-polluted 3 integrated magnitude urban sky (equivalent to about 16.9 mpsas), the light from a 20 day old Moon adds on average about -1 mpsas to your sky brightness. That is it would reduce a 16.9 mpsas sky to a 15.9 mpsas sky - equivalent to about 2.1 integrated magnitude sky.

Conversely, if you travel to your favorite 21 mpsas - 6.1 integrated magnitude rural dark sky site, the same Moon reduces sky brightness to 20 mpsas or 5.5 integrated magnitudes.

The effect of moonlight is dependent on the lunar phase and the degrees of distance between the Moon and the observed target. Krisciunas and Schaefer (1991) (at Table 2) give the following rough delta mpsas (as measured through a Johnson V standard filter) for the lunar phase angle and distance between the Moon and the observed target. These results are for their improved model over the Walker (1987) rough model:

Phase angle Angular distance Moon-target
05 30 60
30 -4.5 -2.9 -2.2
60 -3.7 -2.1 -1.5
90 -2.7 -1.4 -0.9
120 -1.6 -0.6 -0.3

The value in Krisciunas and Schaefer's table closest to LCROSS impact circumstances is a phase angle of 60 and an angular distance of 5 degrees. They report a delta mpsas of -3.7 for those circumstances.

(Lunar phase angle is the number of degrees between the Sun, the Moon and the observer. It varies between 0 and 180 degrees. It is the S-T-O angle reported by the NASA/JPL Horizon's Ephemeris system. Phase angles less than 90 degrees imply backscattering of light; phase angles greater than 90 degrees imply forward scattered light.)

For an urban 16.9 mpsas (3.0 integrated magnitude) sky, 60 degree phase angle moon light would increase sky brightness near the Moon by -3.7 mpsas to about 13.2 mpsas - or -0.5 integrated magnitudes. At the rural dark sky site with an excellent 21 mpsas - a 6.1 integrated limiting magnitude - moonlight was reduce the limiting magnitude to 17.3 mpsas or a 3.3 integrated limiting magnitude.

Equating (roughly and perhaps inapproriately) the LCROSS team's single 550nm wavelength mpas of 3.0 to the LCROSS ejecta curtain to the full Johnson V-band mpsas, indicates that the LCROSS ejecta curtaion might have a positive contrast index even in an urban light polluted environment:

Sky
brightness
5 degs from Moon Curtain MPSAS Contrast Index
Urban 13.2 3.0 4.1
Rural 17.3 3.0 5.7

Even though the LCROSS ejecta curtain will be seen rising against a bright background sky, at an average curtain brightness of 3.0 mpsas, it still may be visible or subject to imaging.

An obvious implication of the above is that travelling a dark sky site may improve the contrast between the ejecta curtain. Dark skies - even though they are washed out by moonlight - still relatively increase an amateur's likelihood of visually detecting the impact curtain and of imaging the curtain.

An extended object fainter than 3.0 mpsas - such as Saturn - has been widely imaged as it was occulted by the Moon. Saturn has a computed mpsas of 6.7 based on an diameter of 14 to 20 arcsecs and an integrated magnitude ranging between 0.4 to 1.2 mags. But unlike Saturn, the LCROSS ejecta curtain will be a low-density dispered dust cloud.

In an abbreviated literature search, no articles were found concerning the additive effect of moonlight between 0 degrees and 5 degrees from the lunar limb.

Lunar amateurs and lunar occultation amatuers are familiar with this effect of moonlight. Only the brightest stars can be seen next to the bright limb of the full Moon. Magnitude 4 to 7 stars disappear within an arcminute of the bright lunar limb.

Amateurs in the LCROSS Google observation group ( url: http://groups.google.com/group/lcross_observation ) have been attempting to gain some understanding of the brightness of the night sky above and within a quarter-degree of the lunar limb by observing grazes and imaging stars during lunar occulations of clusters. Example images are:

Chris Kitting. Moon occulting Pleaides 2009-08-14 9;46UT
http://01227941410742638900-a-g.googlegroups.com/web/CKitting_20090814_0946Pleiades.jpg
http://tinyurl.com/oopegs
Image details: http://groups.google.com/group/lcross_observation/msg/80c0f58084faec69

In Kitting's image, magnitude 7 to 8 stars in Merope's Tail are co-exposed close to an overexposed lunar terminator.

Derek C. Breit April 25, 2007 6.6 mag star graze on dark limb
http://www.poyntsource.com/tmp/April_25th_2007_Graze.mpg (24 mb mpg)

These experiments have had limited success, principally because the few number of such occultation events and uncooperative weather limits the ability to gather useful data using those techniques.

An alternative technique would be CCD photometry analysis. In this technique, the south pole of the Moon is imaged at high focal ratios (f/30 plus) at various exposure times from underexposure to high over-exposure. The camera and scope are then retargeted without changing focal length or exposure on an open cluster at a similar altitude as the Moon. The V magnitudes of open clusters are well-known and plots of clusters limited by a specific magnitude are easily made using the Webda's online database. url: http://www.univie.ac.at/webda/

The open cluster provides reference stars to determine simple transform coefficients for the CCD camera at a given exposure setting through standard B and V or C and V filters. Modern imaging processing packages like AIP4WIN include photometry utilities by which the differential magnitude of stars can be found and correlated to CCD well ADUs. See single star photometry in Exercise C.5 in the AIP4WIN handbook and the extractive photometry utility on the AIP4WIN menu (Measure | Photometry| Extractive Photometry). The open cluster image also provides an image scale that can be applied to your image of the south pole.

This poster unsuccessfully attempted a preliminary test during the August 14, 2009 lunar occultation of Pleaides. An image of the south pole was taken, followed by an image of the Alcyone triple. Unfortunately, weather conditions and an unanticipated glare problem unquie to the Mak imaging scope prevented getting a useable result. The Meade ETX 125 Mak has a curved corrector plate. When slewing from the south pole to Alcyone, the Mak's curved surface diverted off-axis lunar glare into the tube and washed out the image. No useable photometry information could be gleaned from the image, but a demonstration information on image scale could be found.

A panel summarizing this technique demonstration test can be seen at:
http://members.csolutions.net/fisherka/astronote/observed/LCROSS/2009_8_14_0246UT_KafTestPanel.jpg
http://tinyurl.com/n4l8sn

The next best - and last - analogous lunar south pole illumination to the LCROSS impact will occur on September 9 11:30UT:

Lunar libration data for images and impact

Source: LTVT ephemeris data, topocentric W111.8 N41.8
Date-TimeUT libr_lat lib_long colong illumfrac
20090907 1130 -6.2 -5.5 128 92
20090908 1130 -6.2 -5.5 140.2 85.5
20090909 1130 -6.1 -5.4 152.3 77.3
20091009 1130 -3.5 2.8 158.2 70.9 Impact day
20090910 1130 -5.4 -5.1 164.4 67.7
20090911 1130 -4.4 -4.3 176.5 57.1
20090912 1130 -3.2 -3.3 188.7 45.8

Date-TimeUT Lunar age (days)
20090907 1130 18.1
20090908 1130 19.1
20090909 1130 20.1
20091009 1130 20.6 Impact day
20090910 1130 21.1
20090911 1130 22.1
20090912 1130 23.1

On the early morning of Sept. 8, there are no appropriate open clusters near the Moon. NGC752 (Caldwell 28) is higher in the sky. The Double Cluster is also at a higher altitude, but visible.

On the early morning of Sept. 9, the Moon and M45 will both be visible, but the Pleaides will be a higher altitude.

On the morning of Sept. 10, the Moon will be near the Pleaides and at the same altitude. On the evening of the 11th, a number of open clusters might be used for baseline photometry - M45, M36, M37, M38 or NGC1647.

As we approach full Moon on September 4, the Moon will be 68% illuminated on August 29 (from the opposite direction). The night of August 29 provides an opportunity to set up and test equipment under test analogous illumination before September 9.

On the evening of August 29, the Moon will be low in the southern sky just below the apex of the Sag "teapot". Open clusters M21, M23 and M25 are also visible, but at higher altitudes.

Amateur photometry to determine sky brightness within one quarter degree "above" the lunar south pole on September 8 through 11 (particularly on September 9) will help determine if the LCROSS impact ejecta cloud will have sufficient contrast against the moonlight night sky to be seen and imaged. Imaging through filter combinations (V-C, V-I, B-V) will give the most accurate results. To confirm whether the background sky brightness will not overwhelm the ejecta curtain brightness, better photometry data might be collected by amateurs within zero to 5 arcminutes "above" the south lunar pole.

Clear Skies - Kurt

Disclaimer: This is an amateur note. Criticisms and corrections to the above are welcomed.

Fisher, Kurt A. (amateur). 2008. Conversion Calculator for NELM(V) to MPSAS (B) systems. (Web calculator). url: http://members.csolutions.net/fisherka/astronote/plan/tlmnelm/html/NELM2BCalc.html (last accessed 26 Aug. 2009)

Heldman, J. (LCROSS Team). Email Feb. 11, 2009, Slide 4 (slide4.gif).

Krisciunas, K. 1990. Further measurements of extinction and sky brightness on the island of Hawaii. PASP 102:1052-1063. Bib. Code 1990PASP..102.1052K url: http://adsabs.harvard.edu/abs/1990PASP..102.1052K (last accessed 26 Aug. 2009)

Krisciunas, K. and Schaefer, B.E. 1991. A model of the brightness of moonlight. PASP 103:1033-1039, Bib. Code. 1991PASP..103.1033K url: http://adsabs.harvard.edu/abs/1991PASP..103.1033K (last accessed 26 Aug. 2009)

LCROSS Team. 2009. Average and Edge Brightness of Ejecta Curtain (Figure). LCROSS Observation Campaign website. url: http://lcross.arc.nasa.gov/impact.htm and http://lcross.arc.nasa.gov/observation.htm Image: observation05.jpg (last accessed 26 Aug. 2009)

Walker, A. 1987. ________________. NOMO Newsletter. 10:16.

Efls for imaging the LCROSS impact

The following is amateur work product. Criticism and corrections are appreciated.

This opinions put forward here on recommended effective focal lengths to image the LCROSS impact differ slightly from the official Citizen Science LCROSS recommendations on the "About" page. See url:

http://apps.nasa.gov/lcross/about/

This note is based on a discussion of Roger Sinnott's nomogram for choosing an effective focal length at url:

http://media.skyandtelescope.com/images/Linked.gif

Imagers can be best prepared to capture the impact by brainstorming and discussion of techniques before this unique one-shot impact event.

I. Optimal resolution theory

Modern imaging theory suggests that the optimal resolution images can be obtained by magnifying the full-width half-maximum (FWHM) size of the atmospheric seeing disk so that it subtends two pixel elements. Optimal resolution is achieved by matching the current seeing conditions to pixel element dimensions and to the target type.

This guiding principle suggests focal lengths suitable for imaging the LCROSS impact depending on camera type, e.g. DSLR, CCD, high-end lunar planetary imagers (LPIs) or low-end LPIs.

For discussion purposes, the LCROSS Team recommended 254mm (10 inch) aperture is assumed.

For long-exposure amateur photography without the benefit with adaptive optics, the effective seeing disk size is equal to the current seeing conditions. If the seeing disk as disturbed by the atmosphere is 2 arcseconds that is the minimum resolution that can be achieved. This is because the exposure time for faint extended objects like nebulae are above 15 seconds. The disturbed atmosphere constraint controls, even where the theoretical resolving limit for a telescope (e.g. the Dawes limit) signficantly is less than disturbed seeing limit, i.e. 2 arcseconds.

For long-exposure astrophotography, the imager seeks to take this 1, 2 or 3 arcsec disk and magnify it sufficiently to cover 2 pixel elements on their camera.

For lunar and planetary imaging, the highest resolution exposure strategy is different because the objects are very bright. Exposure times can be corresponing low, for example around 0.036 second exposures (28 frames per second (fps) using a Meade DSI or 0.02 secs (60 fps) using an ImageSource mono camera). With the Meade DSI, I take very short exposures (0.01 seconds), but spaced at longer intervals.

These low exposure times bring into play the possibility of the applying the poor man's adaptive optics - "lucky imaging". In lucky imaging, hundreds or thousands of exposures are gathered over an extended period in disturbed air. The hope is that luck will smile on the imager and that a few of those images will occur during subsec periods when the air is absolutely still.

Lucky imaging brings into play the theoretical resolving power of the amateur's telescope.

Recall that under optimal resolution theory, the full-width half-maximum (FWHM) size of the seeing disk covers two pixels. When lucky imaging is involved, that seeing disk is the theoretical resolving power of your telescope.

The theorectical resolving power of a telescope can be related to the full-wdith half-maximum (FWHM) size of the telescope's Airy disk. The Airy disk is the diameter of the diffraction disk to the first diffraction dark ring and contains 84% of a point source's light energy. Dawes limit measures a radius, not a disk diameter. The Dawes limit is the radius of the diffraction disk to the first diffraction dark ring and contains 42% of a point source's light energy. The FWHM represents diameter of the Airy disk that contains 50% of the Airy disk's energy.

For a 254mm (10") aperture, the Airy disk is 0.47", the Dawes limit is 0.46" and the FWHM is 0.40". to achieve optimal resolution, it is this FWHM diameter that the lunar planetary imager seeks to magnify across two physical pixel elements.

A brief bit of math - the basic equation for finding the magnified size of an object on a flat plate at prime focus is:

A_meters = lambda_radians * Fl_meter [Eq. 1]

where:

A_meters = the size of the object on the flat plate at prime focus in meters
lambda_radians = the angular size of the object in radians
Fl_meter = the focal length of the telescope

Two assumptions are implicit in the the lucky imaging strategy for lunar and planetary imaging:

1) That are there brief subsecond moments when the image of the object is crisp.

2) That you can image sufficiently long enough that there is a good probability you will capture a sufficient number of these crisp images to be deconvolved into an even sharper imaging with image processing software.

II. Pixel element sizes of modern LPIs, CCDs and DSLRs

Modern LPI, CCD and DSLR cameras come in a wide variety of total chip size. The total chip size governs the total true field of few captured on the image and is not related to the size of the pixel elements on the chip. It is the size of the pixel elements that governs resolution.

Modern LPI, CCD and DSLR chips can be divided roughly into classes by individual pixel element sizes in microns (um=microns), e.g. -

----------------------------------------------

Table 1: Physical pixel element sizes in microns

Pixel Class | Type | Camera | Chip | Pixel size um | Pixel diag. um

Small sq | High LPI | Imagesource DMK 31AU03 | ICX204AL | 4.6x4.6 | 4.6*

Small sq | Mid rng LPI | ImageSource DMK 21AU04 | ICX098BL | 5.6x5.6 | 5.6*

Small sq | Low end LPI | NexStar Celestron LPI | Unknwn | 5.6x5.6 | 5.6*

Small sq | Low end LPI | Meade LPI imager | Unknwn | 6x6?? | 6*

Small sq | High end LPI | Orion Starshoot | Aptina MT9V032 CMOS | 6x6 | 6*

Small sq | Mid rng CCD | Orion Starshoot III | ICX285AL | 6.4x6.4 | 6.4*

Small sq | DSLR | Canon ESO350D | Unkwn | 6.4x6.4um | 6.4*

Small sq | Mid CCD | Meade DSI III | ICX285AL | 6.4x6.4 | 6.4*

Mid range | High end CCD | Lumenera 2-1 | ICX205 | 7.6x6.2 | 9.8

Mid range | High end CCD | SBIG ST7 | Kodak KAF-400E | 9x9 | 9*

Mid range | Old LPI | Quickcam LPI | Unkwn | 10x10 | 10*

Mid range | Low end CCD | Meade DSI Pro I | Unknwn | 9.6x7.5 | 12.2

Large |High end CCD | Apogee AP8 | Kodak KAF-1000 | 14x14 | 14*

Large | High end CCD | SBIG ST4 | Unkwn | 13.75x16 | 21

Large |High end CCD | Starlight | Sony ICX083AL | 23.2x22 | 22*

* For square pixel elements, one side traditionally is taken as the controlling dimension. For rectangular pixel elements, the diagnol's length is used.

----------------------------------------------

III. Finding your effective focal length to match the optimal resolution for your camera with Sinnott's nomogram

As a general rule-of-thumb for DSO imaging, the smaller the pixel element and the larger the seeing disk, a smaller focal length is applied to reach optimal resolution. Only larger pixel elements can match long focal lengths to a large seeing disk. This is why high end CCDs can have big pixel element sizes - they can stand high levels of magnification. This principle also explains what "binning" is about. Binning takes two small pixel elements and combines them into one pixel - effectively creating a larger pixel element that better matches higher magnifications.

But in general, most amateurs are using cameras with smaller 12 micron and below pixel element sizes. The optimal resolution matching principle explains, in part, why SCT owners who want to use their 2000 to 3000mm focal length scopes to image DSOs first buy a focal reducer. Focal lengths under 2000mm best match long-exposure 1 and 2 arcsec seeing with the pixel element sizes of common CCDs and DSLRs.

For lunar planetary imaging the rule-of-thumb is the opposite: a longer focal length is used with smaller pixel element sizes.

Applying the theory of optimal resolution, for a Meade DSI Pro I with a 12.2 micron pixel (or 24.4 microns for two pixels), I would want to magnify the 0.40 arcsec FWHM disk (same as the 0.5 arcsec theoretical lucky image seeing disk) to about 22.4 microns. But based on experience, I'm going to use a 1 arcsec seeing disk with a FWHM of 0.6". Using the image scaling equation, I compute that I need about a focal length of 4700mm to achieve the optimal sampling resolution.

Rather than deleve into imaging scaling math, a simple to apply nomogram that relates seeing disk size, the type of imaging (DSO or lunar planetary), chip size and effective focal length was published in S&T in 1997.

Dennis di Cicco. _____. Of Pixel Size and Focal Reducers. Sky & Telescope.
url: http://www.skyandtelescope.com/howto/astrophotography/3304356.html?showAll=y&c=y (last

Roger Sinnott's nomogram printed in the diCicco article is particularly helpful to understand these principles:

http://media.skyandtelescope.com/images/Linked.gif

Sinnott's nomogram is divided on the left side into imaging objectives - DSOs and Planetary-Lunar. Note the seeing disks for Planetary-Lunar are in the theoretical range of telescope performance above 10 inches of aperture, expressed in terms of the Airy disk size. The DSO left-hand scale is grouped around atmospheric constrained seeing at disk sizes of 1 or 2 arcsecs.

On the right-hand side of Sinnott's nomogram is the pixel element size of a chip ranging from 30 microns down to an LPI's tiny 5 micron elements. See Table 1, above.

For my Meade DSI Pro I example, drawing a line between 0.6 arcsecs on the left and a 12.2 micron chip element on the right, suggests a focal length of somewhere above 4000mm.

For my 1200 mm focal length 10" aperture DOB, that translates into applying projection magnification of about 3.75x and 4x and an f-ratio of 4700mm/254mm or about f/18.

IV. Applying optimal resolution and effective focal length to the LCROSS impact ejecta curtain.

LCROSS is a hybird event that falls between a DSO imaging strategy and a lunar planetary lucky imaging strategy. On the one-hand, the impact is basically a lunar planetary imaging - invoking application of the lucky imaging strategy. On the other, the short-time frame and continuous change nature of the event - an ejecta that changes in size and brightness over a 60 second time frame - invalidates underlying assumption of the lunar lucky imaging strategy. Lunar lucky imaging assumes a stationary unchanging target over three or four minutes - sufficient to capture enough crisp images in still air that can be stacked and deconvolved using image processing software.

Note Sinnott's nomogram with respect to the remaining discussion in this section:

http://media.skyandtelescope.com/images/Linked.gif

A) The LCROSS ejecta curtain and the lucky imaging strategy

The conservative 95% probability simulation for the LCROSS ejecta curtain is an extended object about 10km by about 3km sticking up above a crater rim. This translates into an angular size of 5.6 arcsecs by 1.7 arcsecs.

For a 10 inch (254mm) recommended aperature with an experience based resolution of 0.7 arcsecs FWHM and a 1.0 arcsec seeing disk, mentally, one would divide this linear object into 5 or 6 one arcsec "point" objects. To achieve optimal resolution, these five or six point objects get magnified so they subtend 10 or 12 pixel elements on your CCD chip or LPI of choice.

If you are using a lunar planetary camera with 0.5 or 0.6 micron sized pixel elements, Sinnott's nomogram suggests using about a 2000mm efl. If you are using a CCD or DSLR camera with 12.2 micron sized pixel elements, Sinnott's nomogram suggests using about 4000mm of effective focal length.

B) The LCROSS ejecta curtain and the traditional DSO strategy

The conservative 95% probability simulation for the LCROSS ejecta curtain is an extended object about 10km by about 3km sticking up above a crater rim. This translates into an angular size of 5.6 arcsecs by 1.7 arcsecs.

For a 10 inch (254mm) recommended aperature with seeing at 2.0 arcsecs with a 1.5 FWHM disk, mentally, one could divide the LCROSS linear ejecta curtain into 4 or 5 two arcsec "point" objects. To achieve optimal resolution, these four or five point objects get magnified so they subtend 8 or 10 pixel elements on your CCD chip or LPI of choice.

If you are using a CCD or DSLR camera with 12.2 micron sized pixel elements, Sinnott's nomogram suggests using about 1500mm of effective focal length to acheive optimal resolution matching.

C) The LCROSS ejecta curtain and a hybird strategy

The hybird nature of the impact might justify a hybird strategy, e.g. - assume 1.5 arcsec seeing and 1.0 arcsec FWHM disk. For a 5 or 6 micron sized pixel elements of a planetary camera, Sinnott's nomogram suggests a very low 1000mm efl. For a larger 12.2 micron sized pixel elements of a DSO camera, Sinnott's nomogram suggests about 2000mm efl. This is the elf recommended on the LCROSS Citizen Science "About" page for DSLR cameras.

V) Lunar glare and imaging the LCROSS ejecta curtain

Large chip arrays in DSLRs and high-end CCD cameras may create a unique barrier. In a couple of prior notes, I mentioned articles by Bradley Schaefer on his modeling of lunar glare and its effect on the faintest magnitude star that can be seen during a lunar occultation.

Post 8-29-2009 re: Schaefer's lunar glare effect on lunar occultation articles
http://tinyurl.com/lkfv4c

Post 8-30-2009 on S&T QBasic Program of Schaefer's lunar glare model
http://tinyurl.com/mr2j2v

Post 8-31-2009 Modified Schaefer occultation program
http://tinyurl.com/kw6t88

The take-away point form playing around with Schaefer's lunar glare model, the faintest star visible - and thus the brightness of the sky just above the Moon - are significantly sensitive to the fraction of the Moon that is visible in your eyepiece. The larger fraction of the lunar disk that you have in your eyepiece or on your CCD chip, all the dark areas on your chip will read relatively brighter.

Extrapolating this relationship to imaging the LCROSS impact, the lunar glare effect seems to weigh in favor of using higher magnifications and smaller chips that will capture a smaller fraction of the sunlit side of the Moon's terminator on CCDs chip. More glare will make the dark unlit portions of craters on the sunlit side of Moon (e.g. Caebus A and B) or the night sky above the dark limb of the lunar terminator relatively brighter and might reduce the contrast between the fainter ejecta curtain and its background.

VI. More on differential photometry of dark crater areas

Here's an image that I took this morning (9-8-2009) that is overexposed and at an elf of about 4000mm. Seeing was poor - the Moon was at a high altitude but the jet stream was right over my observing point and ran across the Moon's disk. The purpose of the image was to look at the differential photometry in the dark crater holes as compared to the night sky. In terms of ADUs from a raw image, the dark craters floor read about 0.4 mags brighter than the dark sky:

http://members.csolutions.net/fisherka/astronote/observed/LCROSS/20090908_8UT_SP_verD_label_kaf.jpg
http://tinyurl.com/n27edc

VII. Conclusion

Using an internet search, find the size of your camera's pixel elements. Use Sinnott's nomogram to estimate a recommended effective focal lenght for your telescope and camera. Then devise a projection magnification setup to magnify the 5 or 6 "points" that will match the LCROSS impact linear ejecta curtain object to the scale of your pixel camera estimates.

Make a judgment call on which strategy you feel will be best - traditional DSO, lunar lucky imaging, or a hybrid between the two. See Sinnott's nomogram to choose a focal length that matches your strategy.

Consider the effect of lunar glare on the TFOV covered by your image. Larger chips and lower magnifications will bring more of the Moon's brightly illuminated disk into the frame and increase the brightness of the background sky.

Utlimately, there is no "right" answer and experienced imagers have to make a snap or gut judgment call on the best setup for there local seeing conditions.

As always, this is guideline. Test and tweak for your particular gear. On the mornings of Sept. 9 and 10, the Moon will have an illuminated fraction similar to that which will be seen during the impact. It's a good time to test your setup. Considering seeing is a variable, you may want to test a long and short efl setup suitable for your camera's pixel element size. That way you can seemlessly pop in the right set up on impact day - the morning of Oct. 9.

Clear Skies - Kurt

References:

Berry, R & Burnell, J. 2005. 2d. Handbook of Astronomical Processing. (HAIP). Willman-Bell. at page 8

Kitchin, C. 2003. 2ed. Telescopes and Techniques. Springer. ISBN 1-85233-725-7 at page 33

Schaefer, B. E. A star's visibility just before occultation. Sky Telesc., Vol. 85, No. 1, p. 89 - 91 (S&T Homepage) Bib. Code 1993S&T....85...89S