Optimizing visual daytime satellite observations

With increasing number of space resident objects there is a growing
interest in their optical monitoring, partly as an augmentation to
expensive radar facilities. One of disadvantages of visual and infrared satellite tracking is the difficulty to provide meaningful astrometric data
during daytime. In recent years several authors provided simulations
and experimental results of detecting up to 8-10 mag targets during the
day using visual or near IR systems. In order to optimize visual
systems for daytime observations the use of a long-pass filter seems
advisable as the way to lower the background level and noise in
images. At the same time Earthshine, which is an important contributor
to satellite brightness, is dominated by blue light. It is therefore
unclear whether and to what degree the use of a filter will actually
improve the SNR of satellite images. Using two identical 12.5 inch
telescopes B and C from PST3 telescope cluster we compared the
performance of several long-pass filters (with cut-off wavelengths
from 400nm to 780nm) to a non-filtered system both
on stars and satellites at different orbital regimes and at different
illumination conditions. We investigated the possibility to improve
SNR also for stellar images in an attempt to increase the chance
to use them as reference for astrometric measurements.
We also compared the reference system to a
long focal ratio configuration in order to establish the optimum
optical system for daytime observations with a fast visual CMOS camera.