OBSERVATION FEASIBILITY ESTIMATES

Using the calibration information presented in the previous section, we now discuss estimating the feasibility of a variety of possible science observations using HUT. Quick estimates of expected signal-to-noise ratios (S/N) can be made using the sensitivity curves presented in Figures 5-1 and 5-2. These show the continuum flux or surface brightness required to achieve a S/N ratio of 10 per Å in a single 1800 s observation for a point source and for extended sources which fill the long slits. For comparison we show similar sensitivity curves for IUE observations using the SWP camera and the large aperture.

**Figure 5-1:** Point source sensitivity to achieve S/N = 10 per Å in a typical 1800 s exposure. The dashed curve shows the sensitivity of the IUE SWP for the same assumptions.

Figure 5-2: Extended source sensitivity to achieve S/N = 10 per Å in a typical 1800 s exposure. Diffuse emission is assumed to fill the apertures indicated. For IUE, the emission is assumed to fill the 10x20 arcsec large aperture.

More detailed estimates can be made using the following procedures. For a continuum source, the number of detected counts per $\sim$ 0.5 Å pixel can be estimated using the anticipated effective area curve presented in Figure 5-4 and the relation

$\begin{displaymath} N_\lambda~=~{\lambda F_\lambda \over hc} A_\lambda \Delta t \Delta \lambda~~{\rm counts~pixel^{-1}},\end{displaymath}$

(1)

where $F_\lambda$ is in ${\rm erg~cm^{-2}~s^{-1}~\AA^{-1}}$ , $A_\lambda$ is the effective area (from Figures 4-5, 4-6, or 4-7), $\Delta t$ is the integration time (typically 1800 s for a single pointing), and $\Delta \lambda =$ 0.51336 Å for first order, or 0.25668 Å for second order.

For an extended source, one must use the surface brightness for $F_\lambda$ in ${\rm erg~cm^{-2}~s^{-1}}$ Å ${\rm ^{-1}~arcsec^{-2}}$ and the angular size $\Delta \Omega$ of either the slit or the source, whichever is smaller, as in

$\begin{displaymath} N_\lambda~=~{\lambda F_\lambda \over hc} A_\lambda \Delta t \Delta \lambda \Delta \Omega~~{\rm counts~pixel^{-1}}.\end{displaymath}$

(2)

=-2 Count rates for continuum sources must be less than 2.5 counts s^-1 pixel^-1 and less than 40 counts s^-1 pixel^-1 at the peaks of isolated emission lines. For sources with rates in excess of these values, the full aperture of the telescope must be stopped down. Closing one door gives a reduction of a factor of two. The 50 cm² small aperture door gives a reduction of a factor of 102, and the 1 cm² small aperture door gives a reduction of a factor of 5120. (Use of the 1 cm² small aperture door requires special procedures to avoid undue pressure buildup in the telescope and spectrograph due to outgassing. Use of this door state should be avoided if possible.)

To evaluate the signal-to-noise ratio (S/N) for a specific observation, one must also take the background into account. The dark count due to charged particle events is quite low, but scattered light from geocoronal Ly $\alpha$ makes a significant contribution to the general background, particularly during orbital day or if large apertures are chosen. For a continuum point source the S/N is given by

$\begin{displaymath} S/N~=~N_\lambda \sqrt { {\Delta t \Delta n \over (N_\lambda + B_\lambda)} },\end{displaymath}$

(3)

where $B_\lambda$ is the background rate per pixel and $\Delta n$ is the number of pixels. Table 5-1 also shows the anticipated background rates due to dark counts plus scattered light in each of the HUT science apertures for both orbital day and orbital night. We assume a geocoronal Ly $\alpha$ intensity of 2 kR (1 Rayleigh $=~10^6 / 4 \pi~{\rm photons~cm^{-2}~s^{-1}~steradian^{-1}}$ ) during night, and 20 kR during day though the actual values will differ and also depend on the shuttle orbital position and the pointing direction. These are conservative estimates based a flight near solar minimum using the models of Meier (1991).

Discrete airglow lines also substantially increase the background in their immediate vicinity, especially during orbital day. Typical airglow spectra for an 1800 s observation using the 20 $^{\prime \prime}$ aperture are shown in Figures 5-3 and 5-4 for orbital night and orbital day, respectively. Both spectra have been truncated to show the fainter features. In the night spectrum, Ly $\alpha$ peaks off scale at 4500 counts pix^-1. In the day spectrum Ly $\alpha$ is off scale at 37,700 counts pix^-1, and O I $\lambda$ 1304 peaks at 1100 counts pix^-1. Again, these estimates are based on Meier's (1991) models near solar minimum.

While orbital night clearly offers advantages, about half the scheduled observations must take place during orbital day. To reserve the lower background of orbital night for the fainter targets, bright point source observations are normally done during the day.

**Figure 5-3:** The HUT spectrum illustrates the expected airglow for an 1800 s observation through the 20 $^{\prime \prime}$ circular aperture during orbital night on Astro-2. Ly $\alpha$ is off-scale at a peak of 3500 counts pix^-1.

Figure 5-4: The HUT spectrum illustrates the expected airglow for an 1800 s observation through the 20 $^{\prime \prime}$ circular aperture during orbital day on Astro-2. Ly $\alpha$ and O I $\lambda$ 1304 are off-scale with peaks of 29,300 and 2500 counts pix^-1, respectively.

Simulated HUT data for a wide variety of input source spectra can be generated with the HUT simulator hutsim, which is a task in the hut package in IRAF. The following instructions will enable a user to acquire and install a local version of the hut package. These instructions are not for a system-wide installation. Rather, the individual user obtains his/her own copy of the programs and files via anonymous ftp. Once installed, the package is not visible when starting up IRAF, but it can be loaded by typing its name, hut.

To install the IRAF package hut containing the hutsim tasks, take the following steps: