关于CCD平场快门效应的改正及有关问题

SOME NOTES ON THE SHUTTER-FREE FLATFIELDING FOR CCD CAMERAS

本文对文献中报道的两种测定快门效应改正的方法进行了讨论．我们指出，对反光镜卡焦，即使面光源本身亮度分布均匀，照射在ＣＣＤ上的辐射也可以不均匀，允许有一个二维分布。但如果这种亮度的相对分布在晨昏蒙影过程中也变化，则Ｓｕｒｍａ方法的应用便要受到影响，而这是许多人忽略的．从单纯测定快门改正函数的角度看，圆顶平场法测定更简单准确．可惜，大多数情况下精确改正快门效应的目的是同时得到精确的平场．如果要求１％或更高精度的平场，单测定快门改正还不够．对每架给定的反光镜，如何测定平场本身都是需要仔细研究的事． 作为例子，给出了佘山１．５６米反光镜加Ｔｈｏｍｓｏｎ ＲＣＣＤ（１０２４ ｘ １０２４像元）的快门改正测定．

There are in literature two kinds of methods of determining the shutter effect with dome and twilight flatfield exposures. We note that: (1) While determining the shutter function with dome flat-fields, the illumination of the dome screen is unnecessary to be homogeneous, i.e., the constant in formula (1) may have a 2D distribution (C = Cij≠const), if the flat-field function itself is not needed. The key point here is that the illumination (light source) must be stable during the whole process of experiment. (2) In order to use Surma's method, the exact shutter time must be known indepe- ndently. Unfortunately, few users have at hand equipment to determine the shutter time. Because the shutter time and the shape of the shutter function may change from time to time due to aging and the ambient temperature, it should be measured just before or after flat-field exposures. (3) An easy way for the users to determine the shutter time is to supplement dome flat-fields after or before twilight flat-field exposures. (4) In Surma's formula(2), C may have a 2D distribution too. The twilight sky must be flat, but the method does not need a constant C. However, the relative distribution must be stable. (5) The critical point is that for a given reflector plus a CCD camera system even the relative distribution Cij may be changeable during twilight exposures, so the application of Surma's method may give rise to errors. The scattered light may be the main cause. (6) While pointing the reflector to a blank night sky region and making a series of exposures successively from 900 to 4 seconds, the night sky flat-field turns to be the twilight flat-field. Comparing this series of frames, systematic variations of Cij clearly appear for the 2.16m reflector of Beijing Observatory and the 1.56m reflector of Shanghai Observatory. We are afraid that these two telescopes are not exceptions among reflectors. (7) If a precision of 2-3% or lower in the flat-field is enough for some users, all the troubles mentioned in this paper may be ignored.