Background and stray radiation from sources other than the target increase the noise and thus affect the detectability of a continuum or line flux.

Aspects which are related to background radiation and which constraint the sensitivity are:

The last three items in this list have already been addressed in detail in Chapter 5.

If the target object was close to strong infrared objects, but not confused with the stronger infrared source, then the background radiation was dominated by emission in the tail of the point spread function of the stronger object. In this case chopping and beam switching techniques generally did not work well. It was then recommended to the observer to consider e.g. observations in scan mode (see Section 4.5) to obtain sufficient information of the stronger source to remove its contribution from the data at the analysis stage.

A.2.1 The celestial background

The two components which contribute most to the celestial background in the infrared are the zodiacal light and the diffuse Galactic emission. Zodiacal light dominates at the shorter ISO wavelengths $\lambda<$ 50 $\mu$ m with a peak around 25 $\mu$ m. The diffuse Galactic emission is more important at the longer wavelengths $\lambda>$ 100 $\mu$ m with a peak around 200 $\mu$ m.

The amount of zodiacal emission depends on wavelength and on the ecliptic coordinates of the object. The closer to the ecliptic plane the more background emission is to be expected. In addition to the dependence on celestial coordinates, zodiacal emission depends also on the satellite orientation: The smaller the solar elongation angle, which for ISO ranges from 120 to 60 degrees, the more zodiacal emission is received. The results obtained with PHT-S, PHT-P and ISOCAM suggest that the zodiacal emission spectrum can be well approximated by blackbodies of 260-290K, depending on the solar elongation and on the ecliptic latitude (Ábráham et al. 1999, [2]). In addition, ISOCAM CVF and ISOPHOT-S measurements demonstrated that the spectrum is featureless between 5 and 16 $\mu$ m. The diffuse Galactic emission has a dependence on Galactic coordinates: Toward the Galactic centre the background radiation is increasing. While the Galactic emission has a significant fraction of intensity in lines and broad spectral features, the zodiacal light is expected to be dominantly continuum radiation.

Both emission components affected the detection of faint sources and it was often desirable to determine the background flux via a reference measurement at a position in the neighbourhood of the source position. The most common methods to obtain reference measurements were beam switching and chopping, which were offered in several AOTs. For some observations the CAM field of view was large enough to image both the source and its background in one frame, thus avoiding beam switching altogether. If a suitable reference position was relatively far from the source position (but still within 3 degrees) the proposer usually prepared two separate observations which were then concatenated (see Section 4.5).

For imaging and spectroscopic observations one of the parameters an observer had to provide was the peak flux density. This included both the emission from the target and the background. For small apertures and strong sources the background contribution was usually negligible. For weaker objects observed with larger apertures (at longer wavelengths) it was necessary to take into account the background emission as discussed in the paragraph below.

For PHT AOTs the background emission was an explicit parameter required for the observations. IRAS maps were considered to be the best source to obtain estimates of the Galactic emission. For estimates at wavelengths outside the IRAS wavelength range, COBE results were recommended. Table A.1 gives some very rough estimates based on COBE data. The values are relative numbers, normalised to the 100 $\mu$ m flux, and should be used for extrapolation from the IRAS fluxes.

It should be noted that the values in Table A.1 apply to the diffuse interstellar medium only. In molecular cloud complexes the surface brightness at 200 $\mu$ m may be factors of 5 to 10 higher compared to the diffuse clouds.

**Table A.1:** *Typical infrared fluxes of interstellar clouds detected with COBE. The results are averages from 10 diffuse clouds and are normalised to the 100 $\mu$ m flux.*
Wavelength	Surface Brightness
[ $\mu$ m]	[MJy/sr]
3.5	0.0016
4.9	0.0015
12	0.043
25	0.058
60	0.42
100	1
140	1.99
240	1.40

A proposer had to be aware of all offset corrections made to the data products used. E.g. IRAS maps are often provided with zodiacal emission subtracted. As zodiacal emission might have been the main contributor to the background, it was necessary to take it into account for the total background level estimate. This was not exactly possible as the satellite orientation was not known prior to the actual observation. Therefore a conservative estimate had to be made to avoid saturation. Table A.2 contains for various wavelengths estimates of the maximum zodiacal light contribution as a function of the ecliptic latitude. The ecliptic latitude $\beta$ can be obtained from Right Ascension $\alpha$ and Declination $\delta$ by equation:

**Table A.2:** Maximum zodiacal light contribution in MJysr $^{-1}$ with the smallest solar aspect angle possible for ISO (60 $^{\circ}$ except for the ecliptic pole) as a function of the ecliptic latitude $\beta$ and wavelength $\lambda$ . The estimates are based on measurements by ZIP (2.5-7.5 $\mu$ m, Murdock & Price 1985, [128]) and IRAS (12-100 $\mu$ m), an examination of full IRAS scans obtained with a solar aspect angle $60^{\circ}<\varepsilon<61^{\circ}$ except for the poles for which scans with about 90 $^{\circ}$ solar aspect angle were used). The 200 $\mu$ m estimate is an extrapolation of the IRAS data.
$\lambda$	2.5	3.5	7.5	12	25	60	100	200
	$\mu$ m	$\mu$ m	$\mu$ m	$\mu$ m	$\mu$ m	$\mu$ m	$\mu$ m	$\mu$ m
$\vert\beta\vert$
$0^{\circ}$	0.2	0.2	25	75	140	45	25	15
$10^{\circ}$				55	100	30	15
$20^{\circ}$				40	70	20	10
$30^{\circ}$				30	50	12	8
$45^{\circ}$				20	35	10	6
$60^{\circ}$				15	30	8	5
$90^{\circ}$	0.2	0.15	4.5	14	25	7	1	0.26

A.2.2 Source Confusion

Source confusion is an additional noise factor closely related to the astronomical background. However, there is a fundamental difference. The sensitivity limit due to confusion is determined by the telescope aperture, observing wavelength and position on the sky. After reaching the confusion limited level, sensitivity cannot be improved by increasing the integration time. With the ISO mission properties and depending on the instrument configurations the confusion limit was reached within a relatively short integration time. Therefore users had to be aware of the limiting source densities as a function of wavelength and the position on the sky. The most important contributors to the confusion limits in the ISO wavelength range are listed below.

For an object away from the Galactic plane ( $\vert b\vert>20^{\circ}$ ) and at wavelengths below 25 $\mu$ m it was necessary to consider confusion if the estimated flux of the target was less than about 1mJy. On the Galactic plane the situation is worse (i.e. the confusion limit is higher). At longer wavelengths the approximate limits are: 10mJy at 60 $\mu$ m, 50mJy at 100 $\mu$ m and 0.1Jy at 200 $\mu$ m. However, the level of cirrus may easily change these limits by an order of magnitude depending on the location of the target in the sky. Model dependent estimates of confusion limited flux levels have been presented by Franceschini et al. 1991, [55] and by Gautier et al. 1992, [61]. Results on the confusion problems, based on ISO results, can be found in Herbstmeier et al. 1998, [79]. A discussion of the models and the consequences for observations with ISO can be found in Puget 1992, [138].

When the expected source flux suggested that confusion noise would be a significant factor, it was important to obtain observations with the highest available spatial resolution. This was achieved by setting the reference measurement (or chopping for PHT) as close as possible to the target position. The best observing strategy for sources with a flux level close to the confusion limit was to obtain fully sampled maps. This was, of course, very expensive as far as observing time is concerned.

A.2 Astronomical Background Radiation

A.2.1 The celestial background

A.2.2 Source Confusion