This page provides details of how my observations of asteroid 285263 (1998 QE2) are processed.
 
Imaging

A Celestron 11-inch telescope (CPC 1100) is inside an 8-foot dome. Two buried cables provide power and control functions, respectively. The control room is inside my residence. All hardware is controlled by MaxIm DL, v5.24. The observatory is located in southern Arizona at west longitude = 110:14:16, latitude = +31:27:08, elevation 1420 meters. I refer to it as HAO (Hereford Arizona Observatory), and it has a MPC site designation of G95.

The CCD is a SBIG ST-100XME. A 10-slot filter wheel includes filters for bands BVRcIc, u'g'r'i'z' and Cb (clear with blue-blocking). An autoguider chip (next to the main chip) is used for autoguiding. The image scale is 0.605 "arc/pixel, which provides a field-of-view (FOV) of 22.0 x 14.8 'arc. Binning is set to 2x2. Exposure time is currently 15 seconds.

Image calibration includes use of master bias, dark and flat frames. Hot and dead pixels are identified and corrected. Star alignment is achieved using the "Auto - star matching" tool. An artificial star is placed in the upper-left corner of each image.

Here's an example of two consecutive images:

MaxIm DL's photometry tool is used for creating a CSV-file with magnitude readings of a set of images at one FOV location. The photometry aperture radius is set to ~ 2 x FWHM [pixels], where FWHM is the median of a sample of the images for a FOV location. Typically FWHM ~ 3.5 pixels, so the photometry radius is typically ~ 7 pixels. A sky background annulus width of 9 pixels is adopted. A gap width of 3 or 4 pixels is used.

The artificial star is used as a "reference star." Since its flux is constant the fluxes of all star readings serve (during a later phase of analysis) to establish atmospheric extinction (and rate of change), as well as departures from a model extinction versus time - that I refer to as "Extra Losses" (explained below).

Check stars are chosen on the basis of having SNR > ~100. The check stars are candidates for use as reference (described below). Here's an example of a choice of check stars:

Spreadsheet Processing

For each FOV location there is one CSV-file with photometry readings (given as magnitudes) of the moving target (asteroid), the artificial star (constant) and a dozen or more check stars. This file is imported to a spreadsheet used to evaluate suitability of the check stars for use as reference, and modeling of atmospheric extinction and creation of a light curve.

Figure 2. Comparison of calibration star magnitudes with a model fit versus star color.

The above graph shows that the flat field correction achieved an accuracy across the FOV that is good to a few mmag. If, for example, the flat field correction had a systematic error of 10 mmag near the FOV center half area, the calibration stars (found throughout the FOV) would produce a component of scatter of ~ 20 mmag, which would add orthogonally to the APASS uncertainties and the measurement uncertainties to produce an apparent RMS scatter about a best fit that would be at least 20 mmag. The apparent RMS scatter about the best fit is 13 mmag for this star field (N = 13).

Mosaic Flat Field Refinement

On June 18 I began a process of refining the flat field correction using a matrix of empirically determined magnitude corrections based on a set of 54 images of the same star with FOV placements that cause the star to be observed at 54 locations within the FOV. Initial results are described at another web site: Mosaic Flat Field Correction. This is a "work in progress" so I don't know if the new procedure will be used to re-process any asteroid data.

Related Links
    Asteroid 285263 Results
    Individual Light Curves
    HAO Home Page
    Mosaic Flat Field Correction

WebMaster: B. Gary.  Nothing on this web page is copyrighted. This site opened:  2013 May 29.