7.4 Detector and Sky Backgrounds

The primary background sources that affect COS observations are detector dark count, earthshine, zodiacal light, and airglow emission; neither of the COS detectors suffers from read noise. The ETC allows the user to select among several levels of intensity for each of the sky backgrounds, corresponding to a variety of observing environments.

7.4.1 Detector Dark Count

Table 7.1 lists the detector dark-count rates assumed by the ETC as of November 2023 (pyETC v32.1). We assume that a resolution element, or "resel," spans 6 × 10 pixels on the FUV detector and 3 × 3 pixels on the NUV MAMA. The dark rate on the FUV detectors can sometimes vary by a factor of a few on very short timescales, impacting less than 5% of observations. See Sections 4.1.3 and 4.2.3 for more information.

Beginning with Cycle 21, the ETC uses separate dark rates for science observations and target acquisitions obtained with the FUV detector. CalCOS filters the data by pulse height, while the COS flight software does not, so the effective dark rate for science data is lower than that for target acquisitions. In either mode the dark rate on the FUV detector is truly small, due in part to its windowless design. The NUV detector has a window and thus a higher dark rate.

Table 7.1: Assumed Detector Dark Count Rates.

¹ Resel size can vary for the 1055 and 1096 central wavelengths. See Appendix B of the ETC User's Guide for more information.
A website which monitors dark rates can be found at https://www.stsci.edu/hst/instrumentation/cos/performance/monitoring.

7.4.2 Earthshine

Four earthshine intensity levels, with scaling factors of (none, average, high, extremely high) = (0.0, 0.5, 1.0, 2.0), are available in the ETC. Earthshine intensity is a strong function of the angle between the target and the bright Earth limb. The earthshine surface brightness for a target 24° degrees above the limb, corresponding to the "high" level, is shown in Figure 7.1. The limb angle is approximately 24° when HST is aligned toward its orbit pole (i.e., the center of the CVZ). The variation of earthshine with limb angle is shown in Figure 7.2.

7.4.3 Zodiacal Light

Away from the airglow lines, at wavelengths between about 1300 and 3000 Å the sky background is dominated by zodiacal light, which is generally fainter than the intrinsic detector background, especially for the NUV detector. Figure 7.1 shows the zodiacal light for the "average" level in the ETC. Table 7.2 gives the variation of the zodiacal background as a function of helio-ecliptic longitude and latitude. For a target near (50°,0°) or (–50°,0°), the zodiacal light is relatively bright at m_V = 20.9 mag arcsec⁻², about 9 times the polar value of m_V = 23.3 mag arcsec⁻². These limits are plotted in Figure 7.2. The intensity levels and the factors by which they are scaled in the ETC are (none, low, average, high) = (0.0, 0.576, 1.0, 1.738), corresponding to m_V = (none, 23.3, 22.7, 22.1) mag arcsec⁻².

Observations of the faintest objects may need the special requirement LOW-SKY in the Phase II observing program. LOW-SKY observations are scheduled during the part of the year when the zodiacal background is no more than 30% greater than the minimum possible value for the given sky position. LOW-SKY also invokes the restriction that exposures will be obtained at angles greater than 40° from the bright Earth limb to minimize earthshine and the UV airglow lines. The LOW-SKY requirement limits the times at which targets within 60° of the ecliptic plane will be scheduled and limits visibility to about 48 minutes per orbit.

Table 7.2: Approximate Zodiacal Sky Background (V mag arcsec^–2) as a Function of Helio-ecliptic Coordinates.

Note: A value of "SA" denotes positions in the solar avoidance zone.
See Leinert et al. (1998) for more on diffuse sky brightness.

The ETC provides the user with the flexibility to adjust both the zodiacal (none, low, average, high) and earthshine (none, average, high, extremely high) sky background components to determine if the use of LOW-SKY is advisable for a given program. However, the absolute sky levels that can be specified in the ETC may not be achievable for a given target. As shown in Table 7.2, the minimum zodiacal background level for an ecliptic target is m_V = 22.4 mag, which is brighter than both the low and average options with the ETC. By contrast, a target near the ecliptic pole would always have a zodiacal = low background in the ETC. The user is cautioned to consider sky levels carefully, as the backgrounds obtained in HST observations can span a significant range.

7.4.4 Airglow Emission

In the ultraviolet, the sky background contains important contributions from airglow emission lines, which vary from day to night and as a function of HST's orbital position. These features originate mainly from hydrogen and oxygen atoms in the exosphere of the Earth. Airglow lines may be an important consideration for spectroscopic observations at wavelengths near the lines.

The brightest airglow line by far is Lyman α at 1216 Å. The strength of the Lyman-α line varies between about 2 and 20 kilo-Rayleighs (i.e., between 6.3 × 10⁻¹⁴ and 6.3 × 10⁻¹³ erg cm^-2 s^-1 arcsec^-2, where 1 Rayleigh = 10⁶ photons cm^-2 s^-1 per 4π steradians, which equals 3.15 × 10⁻¹⁷ erg cm^-2 s^-1 arcsec^-2 at Lyman α) depending on the time of the observation and the position of the target relative to the Sun. The next-strongest feature is the O I line at 1302 Å, which rarely exceeds 10% of Lyman α. The typical strength of the O I λ1302 line is about 2 kilo-Rayleighs (about 7 × 10^–14 erg cm^-2 s-1 arcsec^-2) on the daylight side and about 150 times fainter on the night side of the HST orbit. The O I] λ1356 and [O I] λ2471 lines may appear in observations on the daylight side of the orbit, but these lines are at least 10 times weaker than the O I λ1302 line. The widths of the lines also vary, but a representative value for a temperature of 2000 K is about 3 km s^-1. Airglow emission from N I λ1200 is also seen, particularly on the day side of the orbit, with fluxes up to 1.6 × 10⁻¹⁶ erg cm^-2 s^-1 arcsec^-2. The N I line is not included in the ETC. Airglow emission lines are essentially unresolved at the resolution of COS, but the emission fills the aperture in the spectral and spatial directions. For the FUV modes, the aperture width is approximately 114 pixels, or 1.12, 1.36, and 9.46 Å for G130M, G160M, and G140L, respectively. For the NUV modes, the aperture width is approximately 105 pixels, or 3.87, 3.46, 4.18, and 41.21 Å for G185M, G225M, G285M, and G230L, respectively.

The COS ETC provides four airglow intensity levels (none, low, average, high), whose scaling factors depend on the airglow line considered: (0.0, 0.1, 0.5, 1.0) for Lyman α, (0.0, 0.0667, 0.5, 1.0) for O I λ1302, (0.0, 0.006, 0.5, 1.0) for O I] λ1356, and (0.0, 0.005, 0.5, 1.0) for [O II λ2471.

It is possible to request that exposures be taken when HST is in the earth's shadow to minimize airglow emission (e.g., if you are observing weak lines at 1216 Å or 1302 Å) using the special requirement SHADOW. Exposures using this special requirement are limited to roughly 25 minutes per orbit, exclusive of the guide-star acquisition (or reacquisition) and can be scheduled only during a small percentage of the year. SHADOW reduces the contribution from the airglow emission lines by roughly a factor of ten, while the continuum earthshine is essentially nil. If you require SHADOW, you should request and justify it in your Phase I proposal (see the Call for Proposals).

An alternate strategy for reducing the effects of airglow emissions is to use time-resolved observations, so that any data badly affected by airglow emission can simply be excluded from the final co-addition. This can be done either by using TIME-TAG mode, the default for all COS observations if the target is not too bright, or by taking a series of short (~5 min) ACCUM mode exposures over the course of each orbit.

As noted, geocoronal Lyman α is by far the strongest airglow feature. On the day side of the HST orbit, when Lyman α is at its strongest, it will produce a net count rate of 20 counts/s/resel, too faint to be a safety concern, but bright enough to make a significant contribution to gain sag on the FUV detector (Section 4.1.7).

7.4.5 Tabular Sky Backgrounds

Table 7.3 lists the high sky background numbers as plotted in Figure 7.1. The high sky values are defined as the earthshine at 24° from the limb and by the average zodiacal light of m_V = 22.7 mag. The quoted values represent the zodiacal and earthshine backgrounds (excluding the contributions from airglow emission lines) averaged over each wavelength interval. The line intensities of some important airglow lines in the COS bandpass are listed in Table 7.4.

Table 7.3: Earthshine and Zodiacal Light in the COS PSA. 

These rates correspond to the high level in the ETC and are listed in units of 10^–15 erg cm^–2 s^–1 Å⁻¹ for the total COS PSA, which is 4.91 arcsec² in area.

Table 7.4: Typical Strengths of Important Ultraviolet Airglow Lines.

¹ This feature is not included in the ETC.