The advent of far infrared detector arrays will allow qualitatively new approaches to sensitive astronomical detection at these wavelengths. By applying source extraction techniques similar to CLEAN or DAOPHOT, the effective beam size can be reduced substantially with a resulting reduction in confusion noise compared with predictions for instruments with a small number of detectors or for aperture photometers. The same benefits can be achieved for operating in regions where there is strong infrared cirrus. To operate in this mode, it is essential that far infrared telescopes have highly reproducible, preferably diffraction limited, beam profiles.
Table I shows as an example the sensitivity that can be
achieved with a cryogenic 85 cm telescope using a CLEAN algorithm
and operating at the average level of IR cirrus at high Galactic
latitude. At 60
m, the instrument will not become confusion
limited unless integrations longer than 10,000 sec are attempted.
Because the three types of limiting noise scale with almost
the same power of the telescope aperture, their relative
contributions will not change significantly for telescopes of other
sizes. As a result, the well-recognized benefits of operating
meter-class telescopes at temperatures cold enough that they are
natural background limited also apply to larger instruments. For
example, although a telescope operating at < 5K at 160 -
200m will become confusion limited in about 60 seconds, the same
telescope at 8K would require 1000 seconds to reach the same
sensitivity limit; at 15K it would require 100,000 seconds.
Therefore, designs for future far infrared telescopes should
emphasize operating them at low temperatures.
We greatly appreciate the hospitality of the conference organizers. The work reported here was supported by NASA as part of the SIRTF project.