Dark Current and Bias

A dark frame (that includes the bias) can be measured from our telescope by using a cooled blank filter which blocks any incoming light from entering the detector. When taking science frames, it is necessary to subtract off a dark frame with the same exposure time. In order to reduce the noise contribution contributed by error propagation when combining images, we can average a set of dark images with the same exposure time. For this lab, we wanted the contribution of the noise from the dark frame subtraction to be less than 10%.

To ensure this, we need to follow how the error propagates when average our dark frames. We will use the following equations to determine how many dark frames we need to average to get an adequately small enough noise contribution.

$$\bar{x}_{dark,avg} = \frac{1}{N}\sum_i^N{\bar{x}_{dark}}$$

$$s_{dark,avg}^2 = \frac{s_{b}^2}{\sqrt{N}}$$

Further analysis also shows that the bias of our infrared array is almost negligible. The bias for our camera is 0.7$\pm$.7 DN. The same analysis also found the dark current to be 78.46$\pm$.12$\frac{DN}{s}$ at a chip temperature of 81.4 K.¹