The American Association of Amateur Astronomers

Visit

Home
Weighing Jupiter
Eugene Lanning
Downloads

 Project Jupiter
CCD Imaging Method

AstroMax
The AAAA
Online Store

Serving the Amateur Astronomy Community ONLINE since 1996


www.AstroMax.com

Up
Abstract
I. Purpose
II. Background
III. Orbits
IV. Period  Determination
V. Methods
VI. Kepler's Laws
VII. Observing Suggestions
IX. Data Processing
X. Observer's Data Results
XI. Other Quad-A Results
XII. Conclusions
XIII. Attachments


This Project Jupiter Report was prepared by
Mizar Consulting
Eugene A. Lanning
130 Hillside Terrace
Nebraska City, NE
68410-3740
ealanni@alltel.net
Member of AAAA


AAAA
The American Association of Amateur Astronomers
P.O. Box 7981
Dallas, TX
75209-0981
e-Mail:
aaaa@astromax.com

www.AstroMax.com


Project Jupiter

Back Up Next

VIII. Data Gathering Methods

C. CCD/Astrophotography Method

1. Accuracy

The basic techniques used here are the same as the Sketch Method, but with the inaccuracies of the hand sketching removed.

2. Basics

During each observing session the observer takes an astrophoto or CCD image(s) that contains the moons of Jupiter. Later the separation on the photo/image is measured and is then used as one data point in the Jupiter Project data analysis.

3. Camera lens / CCD equipment Selection

Because the separation of the satellite in any given observation is to be compared to other observations, there is a need to have a consistent FOV from one observation to the next. The observer needs to select equipment that as that enables the maximum separation of the moon to be still on the images made during subsequent observations 14 without having to change your equipment.

4. Magnitude Differences

Because the magnitude of Jupiter and its satellites varies considerably, it is recommend that one image be taken of Jupiter with its moons and a separate and lesser-exposed image of Jupiter be taken. To image the satellites adequately the image of Jupiter is over exposed and pixel bleeding makes the image of Jupiter larger than is really the case. This contributes to position uncertainties.

It should also noted that the magnitude of Callisto is nearly a full magnitude less than the other satellites. Thus, the image that contains Callisto may need a slightly longer exposure. Particular challenges with Jupiter are the low-contrast image and the strong limb-darkening. Therefore, points to consider are:

  • Adding many short exposures can be better than taking single exposures.
  • CCD chips are most sensitive in the near-infrared, so an unfiltered image looks like a red-light image (with rather low contrast though potentially fine detail), and may suffer particularly from chromatic dispersion. Thus CCD users are encouraged to use an infrared exclusion filter.
  • Because of limb-darkening, some form of digital unsharp-masking is needed to bring out the limb. Most observers process their images to reduce limb-darkening and enhance contrast. Such image-processing should be done judiciously with awareness of the artefacts that it can create; check that there are not conspicuous rings around satellite shadows, nor any saturated white areas in the image.

5. Image Scale

It is helpful if the field of view (FOV) of your equipment is determined before the observing sessions begin. This is not a required step, but enables fewer assumptions to be made in the processing of your data. A variety of techniques may be used here, including creating star trails on the images from a timed exposure with the drive turned off.

To determine the FOV, select a star near the celestial equator (within ±5° of zero declination ) and make a timed exposure. Multiply the exposure time by 15 to convert that clock time into arc-seconds (or arc-minutes if timed in minutes), then divide by the trail length to obtain the number of arc-seconds per pixel on the CCD electronic image or per mm on astrophotos.

6. Data Preparation

Measure the separation of the selected moon from the center of Jupiter on each of the images made. It is recommended that the familiar inches and fractions be avoided by measuring the separations on the images directly in pixels on the computer from CCD images, before prints of the images are produced. Measuring in mm for astrophotos is recommended. Avoid the tendency to "smooth" the data at this stage. If on observation #5 the separation is 80 pixels, on observation #6 it is 95 pixels, and on observation #7 it is 90 pixels (went back down), that is OK. Report the data as imaged. For each observation, record the separation measurement (in pixels or mm ) on a separate line on the data sheet (Attachment A).

[Home]
[Abstract]
[I. Purpose]
[II. Background]
[III. Orbits]
[IV. Period  Determination]
[V. Methods]
[VI. Kepler's Laws]
[VII. Observing Suggestions]
[VIII. Data Gathering]
[IX. Data Processing]
[X. Observer's Data Results]
[XI. Other Quad-A Results]
[XII. Conclusions]
[XIII. Attachments]
[Up]
[Jupiter Diameter Method]
[Sketch Method]
[CCD Imaging Method]
[Eyepiece Method]

 

The image of Jupiter on the Project Jupiter cover page is courtesy of AAAA member Charlie Warren of Texas. Used by permission. Jupiter and three of its moons - right to left are the moons Europa, Io and Ganymede. Callisto is not on the image. CCD Image taken February 2, 2002.

AAAA
The American Association of Amateur Astronomers
P.O. Box 7981
Dallas, TX 75209-0981
e-Mail: aaaa@astromax.com

www.AstroMax.com