7. Instruments IV: FORCAST


7.1 FORCAST Instrument Overview

The Faint Object infraRed CAmera for the SOFIA Telescope (FORCAST) is a dual-channel mid-infrared camera and spectrograph sensitive from 5 - 40 μm. Each channel consists of a 256x256 pixel array that yields a 3.4'x3.2' instantaneous field-of-view with 0.768'' pixels, after distortion correction. The Short Wave Camera (SWC) uses a Si:As blocked-impurity band (BIB) array optimized for λ < 25 μm, while the Long Wave Camera's (LWC) Si:Sb BIB array is optmized for λ > 25 μm. Observations can be made through either of the two channels individually or, by use of a dichroic mirror, with both channels simultaneously across most of the range. Spectroscopy is also possible using a suite of six grisms, which provide coverage from 5 - 40 μm with a low spectral resolution of R = λ/Δλ ~ 200. The instrument has space for cross dispersing grisms allowing for high resolution cross-dispersed (XD) spectroscopy at R ~ 800 - 1200 in the 5 - 14 μm range.The availability of the XD modes during a given Cycle is published in the Call-for-Proposals.

7.1.1 Design

The FORCAST instrument is composed of two cryogenically cooled cameras of functionally identical design. A schematic of the instrument layout is provided in Figure 7-1, below. Light enters the dewar through a 7.6 cm (3.0 in) CsI window and cold stop and is focused at the field stop, where a six position aperture wheel is located. The wheel holds the imaging aperture, the slits used for spectroscopy, and a collection of field masks for instrument characterization. The light then passes to the collimator mirror (an off-axis hyperboloid) before striking the first fold mirror, which redirects the light into the LHe-cooled portion of the cryostat.


FORCAST schematic

Figure 7-1.

The incoming beam then reaches a four-position slide, which includes an open position, a mirror, and two dichroics, one for normal use and the other as a spare. The open position passes the beam to a second fold mirror, which sends the beam to the LWC, while the mirror redirects the light to the SWC. The magnesium oxide (MgO) dichroic reflects light below 26 μm to the SWC and passes light from 26 - 40 μm to the LWC. The light then passes through a Lyot stop at which are located two filter wheels of six positions each, allowing combinations of up to 10 separate filters and grisms per channel. Well characterized, off-the-shelf filters can be used, since the filter wheel apertures have a standard 25 mm diameter (see Section 7.1.3).

Finally, the incoming beam enters the camera block and passes through the camera optics. These two-element catoptric systems are composed of an off-axis hyperboloid mirror and an off-axis ellipsoid mirror that focuses the light on the focal plane array. Also included is an insertable pupil viewer that images the Lyot stop on the arrays to facilitate alignment of the collimator mirror with the telescope optical axis and to allow characterization of the emissivity of both the sky and telescope.

7.1.2 Camera Performance

The SWC and LWC arrays were selected to optimize performance across the 5 - 40 μm bandpass. Both arrays have a quantum efficiency (QE) greater than 25% over most of their spectral range. The cameras can be operated with variable frame rates and in either high or low capacitance modes (with well depths of 1.8x107 and 1.9x106 e- respectively), depending upon the sky background and source brightness.

The best measured image quality (IQ) obtained by FORCAST on SOFIA is in the 7 - 11 µm range with a FWHM PSF of ~2.7''. This measured image quality in-flight is limited by telescope jitter arising from vibrations of the telescope itself (e.g., due to wind loading in the cavity), turbulence, and tracking accuracy. Presented in Figure 7-2 is a sample of FWHM IQ measurements made during a single observatory characterization flight during the winter of 2010 in comparison to the theoretical diffraction limit calculated for a 2.5-m primary with a 14% central obstruction combined with the FWHM telescope jitter, here assumed to be 2.08'' (1.25'' rms).


FORCAST FWHM IQ measurements plot

Figure 7-2: Representative FWHM Image Quality for the FORCAST camera in select filters as measured during Cycle 4. Also shown are the diffraction limited FWHM (solid line; calculated for a 2.5-m primary with a 14% central obstruction) and the modeled IQ (dashed line), which includes shear layer seeing and 1.25'' rms telescope jitter.

7.1.3 Filter Suite


Imaging with FORCAST can be performed in either a single channel or in both the SWC and LWC channels simultaneously (dual channel mode). In single channel mode, any one of the available filters may be used. In dual channel mode, a dichroic is used to split the incoming beam, directing it to both the SWC and LWC.

The dichroic reduces the overall throughput. Table 7-1 shows the throughput with the dichroic in (Dual Channel mode) relative to that for the Single Channel mode. The degradation of the system throughput by the dichroic can have a significant effect on the instrument sensitivities as discussed in more detail in Section 7.1.4. In addition, since there are more short wavelength filters available than slots in the SWC filter wheels, some short wave filters will be housed in the LWC. The final filter wheel configuration for each observing cycle will be driven by proposal pressure. Depending on the final filter configuration, it is possible that not all LWC filters will be able to be used in dual channel mode due to the cutoff in the dichroic transmission at short wavelengths. Additional details are provided below and will be updated once the results of each cycle's Call-for-Proposals are known.

Table 7-1: FORCAST Dichroic Throughput

FORCAST Dichroic Throughput
Bandpass Throughput
5-10 μm 60%
11-25μm 85%
25-40 μm 40%

The filters in the SWC are standard Optical Coating Laboratory, Inc. (OCLI) thin-film interference filters. These filters are stacked with blocking filters to prevent light leaks. The only exception is the 25.4 μm (FOR_F253) University of Reading filter, which is a custom double half-wave (three mesh) scattering filter stacked with a diamond scattering blocking filter to provide blue-light rejection. The 33.6 (FOR_F336), 34.7 (FOR_F347), and 37.1 μm (FOR_F371) filters in the LWC are LakeShore custom double half-wave (three mesh) scattering filters. The 31.5 μm filter is a Fabry-Perot Interferometer filter.

The central wavelengths, bandwidths, and the typical FWHM Image Quality (IQ) in each of the filters are given in Table 7-2 below. The first column under "Imaging FWHM" presents the best measured FWHM IQ in single channel configuration. These values are representative of the IQ observed since Cycle 3. The second column are the average measured FWHM IQ in dual channel configuration. Not all of the filters have measured IQ values, but we expect that they will be comparable to those with measured values that are of similar λeff. Figure 7-3 shows the filter transmission profiles (normalized to their peak transmission) over-plotted on an ATRAN model of the atmospheric transmission. The filter transmission curves (text tables) are available as a zip file or individually from Table 7-2.

Due to the limited number of slots available in the filter wheels, not all of the filters listed in Table 7-2 are available at any one time. For Cycle 5, there will be a "default" filter set (indicated by bold type in Table 7-2). GIs may request that other filters be used for their observations, but they must make a convincing scientific argument to justify swapping out the existing filters for those they desire. We anticipate that a single filter swap may be permitted during Cycle 5 provided there is adequate scientific justification. More details are available in the Cycle 5 Call for Proposals.

Table 7-2: FORCAST Filter Characteristics

FORCAST Filter Characteristics
Channel Filter λeff (μm) Δλ (μm) Imaging FWHM ('') Profiles
SWC FOR_F054 5.4 0.16 b txt file
FOR_F056 5.6 0.08 -
FOR_F064 6.4a 0.14 2.9 2.9 txt file
FOR_F066 6.6 0.24 2.9 3.1 txt file
FOR_F077 7.7 0.47 3.0 3.0 txt file
FOR_F088 8.8 0.41  - -
FOR_F111 11.1 0.95 2.8 2.9 txt file
FOR_F112 11.2 2.7 - - -
FOR_F197 19.7 5.5 2.4 2.5 txt file
FOR_F253 25.3 1.86 2.3 2.1 -
LWC FOR_F113 11.3 0.24 2.6 txt file
FOR_F118 11.8 0.74 2.6 -
FOR_F242 24.2 2.9 2.6 - txt file
FOR_F315 31.5 5.7 2.8 2.8 txt file
FOR_F336 33.6 1.9 3.1 3.3 txt file
FOR_F348 34.8 3.8 3.1 3.0 txt file
FOR_F371 37.1 3.3 2.9 3.4 txt file

a Entires in bold font are expected to be part of the "default" filter set for Cycle 5.

b IQ values for some filters have not been measured at this time, but it is expected that they will be similar to those of similar λeff with measured values.


FORCAST filer transmission profiles

Figure 7-3: FORCAST filter transmission profiles along with an ATRAN model of the atmospheric transmission across the FORCAST band (assuming a zenith angle of 45 degrees and 7 μm of precipitable water vapor). For clarity the filter profiles have been normalized to their peak transmission. SWC filters alternate between green and blue, while LWC filters alternate between red and yellow.

7.1.4 Imaging Sensitivities


The FORCAST imaging sensitivities for a continuum point source for each filter are presented in Table 7-3 and Figure 7-4. The Minimum Detectable Continuum Flux (MDCF; 80% enclosed energy) in mJy needed to obtain a S/N = 4 in 900 seconds of on-source integration time is plotted versus wavelength. The MDCF scales roughly as (S/N) / √(t) where t = net integration time. The horizontal bars indicate the effective bandpass at each wavelength. At the shorter wavelengths the bandpass is sometimes narrower than the symbol size.

Atmospheric transmission will affect sensitivity, depending on water vapor overburden. The sensitivity is also affected by telescope emissivity, estimated to be 15% for Figure 7-4.

Observations with FORCAST will be performed using standard IR chop-nod techniques. The GIs can choose chop/nod amplitudes small enough to leave the source on the array in each position or large enough that the source is positioned off the chip for one of the chop positions. For background-limited observations, as will be the case with FORCAST on SOFIA, chopping and nodding off-chip in nod-match-chop (NMC; see Section 7.2.1) will generally result in the same signal to noise (S/N) as chopping and nodding on-chip in nod-perp-chop (NPC; see Section 7.2.1). Calculations of S/N for various chop-nod scenarios are provided here.

Table 7-3: FORCAST Filter Sensitivities

FORCAST Filter Sensitivities
Filter Channel Single Chan. MDCF (mJy) Dual Chan. MDCF (mJy)
FOR_F054a SWC 65 895
FOR_F056b SWC 74.5 285.2
FOR_F064 SWC 82.1 92.1
FOR_F066 SWC 92.4 120.6
FOR_F077 SWC 85.5 104.1
FOR_F086a LWC 422 -
FOR_F111a SWC 143 155
FOR_F113a LWC 320 -
FOR_F118a LWC 194 -
FOR_F197 SWC 116.2 123.4
FOR_F242 LWC 154.3 -
FOR_F253 SWC 236.2 250.9
FOR_F315 LWC 197.9 269.1
FOR_F336 LWC 414.1 586.9
FOR_F348 LWC 283.6 399.9
FOR_F371 LWC 383 579.3

a MDCF values shown are those measured from Cycle 2 and Cycle 3 data.

b No observations were available to test the theoretical MDCF values.


FORCAST sensitivity plot

Figure 5-4. Cycle 6 continuum point source sensitivities for single and dual channel modes. Values are for S/N = 4 in 900 s under nominal conditions. Investigators are encouraged to use the SOFIA Integration Time Calculator (SITE) for their calculations.

7.1.5 Grisms

The suite of 6 grisms available for FORCAST provide low to medium resolution coverage throughout most of the range from 5 - 40 μm. The grisms are situated in the four filter wheels, two in each SWC wheel and one in each LWC wheel. The arrangement is chosen to minimize the impact on the imaging capabilities of the instrument. The grisms are blazed, diffraction gratings used in transmission and stacked with blocking filters to prevent order contamination. A summary of the grism properties is provided in Table 7-4. Note that during Cycle 5, the cross-dispersed XD configuration will not be available. The information on XD modes below is for informational purposes, only.

Grisms FOR_G063, FOR_G227, FOR_G329, and the FOR_XG063 dispersing grism provided by the University of Texas at Austin, are made of silicon to take advantage of its high index of refraction, which allows optimum spectral resolution. However, these grisms suffer from various absorption artifacts precluding their use in the 8 - 17 μm window. Coverage in this region is provided by the FOR_G111 grism (and its cross-dispersing counterpart), constructed of KRS-5 (thallium bromoiodide) by Carl-Zeiss (Jena, Germany). These latter two grisms have a lower spectral resolution due to the lower index of refraction of the KRS-5 material.

Three slits are available, two long slits (2.4'' x 191'', 4.7'' x 191'') and a short slit (2.4'' x 11.2''). The narrow slits yield higher resolution data. All of the slits are located in the aperture wheel of the instrument. The cross-dispersed spectra are obtained by using the short slit and passing the beam first through the low-resolution grism (either FOR_G063 or FOR_G111), followed by a disperser.

Although grisms are available in both cameras, during Cycle 5 grism spectroscopy will be available only in single channel mode.

It is important to note that due to the fixed position of the slits in the aperture wheels, the lack of a field de-rotator, and the fact that SOFIA behaves in many respects as an Alt-Az telescope, the orientation of the slit on the sky will be dependent on the flight plan and will not be able to be predetermined. Furthermore, the slit orientation rotates on the sky with each telescope Line-of-Sight (LOS) rewind (see Section 3.5). These limitations may be especially important to consider when proposing observations of extended objects.

Table 7-4: FORCAST Grism Characteristics

FORCAST Grism Characteristics
Channel Grism Material Groove Sep. (μm) Prism Angle (°) Order Coverage (μm) R (λ/Δλ)a
SWC FOR_G063 Si 25 6.16 1 4.9 - 8.0 120c/180
FOR_XG063b Si 87 32.6 15 - 23 4.9 - 8.0 1170d
FOR_G111 KRS-5 32 15.2 1 8.4 - 13.7 130c/260a
LWC FOR_G227 Si 87 6.16 1 17.6 - 27.7 110/120a
FOR_G329 Si 142 11.07 2 28.7 - 37.1 160b/170a

a For the 4.7'' x 191'' and the 2.4'' x 191'' slits, respectively

b Not available during Cycle 5

c The resolution of the long, narrow-slit modes is dependent on (and varies slightly with) the in-flight IQ

d Only available with the 2.4'' x 11.2'' slit

7.1.6 Spectroscopic Sensitivity

Tables 7-5 and 7-6 provide samples of the MDCF and Minimum Detectable Line Flux (MDLF) calculated at three different wavelengths across each grism bandpass for each of the available spectroscopic modes – long slit and cross-dispersed. The data are provided for point sources only. The MDCF and MDLF estimates are for the raw integration time of 900 seconds and do not include observing overheads, but do account for a a two-position chop (perpendicular to the slit).

Figures 7-5 and 7-6 present the continuum point source sensitivities for the FORCAST grisms. The plots are the MDCF in Jy needed for a S/N of 4 in 900 seconds at a water vapor overburden of 7 μm, an altitude of 41K feet, and a zenith angle of 60°. The rapid variations with λ are due to discrete atmospheric absorption features (as computed by ATRAN).

To determine the required integration time necessary to achieve a desired S/N ratio for a given source flux, GIs should use the FORCAST on-line grism exposure time calculator. The on-line calculator also allows for calculation of the limiting flux for a given integration time and required S/N. Since FORCAST observations are background limited, the values given in Tables 7-5 and 7-6 and Figures 7-5 and 7-6 can be used to make an estimate of the required integration time using the following relation,


where [S/N]req is the desired signal-to-noise ratio, Fsrc is the continuum flux of the target, texp is the exposure time on source (without taking into consideration observational overheads), and the MDCF is taken from the tables for the point-source sensitivities or estimated from the figures. For emission lines, simply use the line flux for Fsrc and use the MDLF value instead of the MDCF. However, these tables may not contain the most recent or best determined sensitivity values and therefore the on-line calculator results should be used in the actual proposal.

Table 7-5: FORCAST Long Slit Point Source Sensitivities

FORCAST Long Slit Point Source Sensitivities
Grism λ (μm) R = (λ/Δλ) MDCF (mJy) MDLF (W m-2) R= (λ/Δλ) MDCF (mJy) MDLF (W m-2)
  4.7'' Slit 2.4'' Slit
FOR_G063 5.1 120 79 2.3E-16 180 98 2.9E-16
FOR_G063 6.4 120 219 5.2E-16 180 268 6.3E-16
FOR_G063 7.7 120 496 5.2E-16 180 724 6.3E-16
FOR_G111 8.6 130 419 4.9E-16 300 532 6.2E-16
FOR_G111 11.0 130 449 4.1E-16 300 575 5.2E-16
FOR_G111 13.2 130 593 4.5E-16 300 764 5.8E-16
FOR_G227 17.8 110 715 8.6E-16 140 936 1.1E-15
FOR_G227 22.8 110 834 7.9E-16 140 989 9.3E-16
FOR_G227 27.2 110 1979 1.6E-15 140 2586 2.0E-15
FOR_G329 28.9 160 1365 6.5E-16 220a 1899 9.0E-16
FOR_G329 34.1 160 1408 5.6E-16 220a 1994 8.0E-16
FOR_G329 37.0 160 1763 5.6E-16 220a 2439 8.0E-16

a The 2.4'' long slit mode for G329 will not be available during Cycle 5.

Table 7-6: FORCAST Cross-Dispersed Point Source Sensitivitiesa

FORCAST Cross-Dispersed Point Source Sensitivitiesa
Grism λ (μm) R = (λ/Δλ) MDCF (mJy) MDLF (W m-2)
    2.4'' x 11.2'' Slit
FOR_XG063 5.1 1200 238 1.2E-16
FOR_XG063 6.4 1200 703 2.8E-16
FOR_XG063 7.7 1200 918 3.0E-16

a XD modes are not available during Cycle 5


FORCAST Grism sensitivities plot

Figure 5-5. Cycle 6 grism continuum point source sensitivities for both wide and narrow long slits overlaid on an atmospheric transmission model (light blue). Values are for S/N = 4 in 900 s under nominal conditions.


FORCAST grism sensitivities

Figure 7-6: FORCAST grism sensitivities for a continuum point source using the grisms 1 and 2 listed above in Table 7-3. Otherwise as above.

7.2 Planning FORCAST Imaging Observations

As is the case with ground based observations at mid-IR wavelengths, individual FORCAST exposures will be dominated by the sky and telescope background. Therefore chopping and nodding are essential for each observation. Selection of the observing mode, including the distance and direction of chop and nod throws, depends on the details of the field of view around the target. The source(s) of interest may be surrounded by other IR-bright sources or may lie in a region of extended emission, which needs to be avoided to ensure proper background subtraction. Presented in this section is a discussion of how to best plan FORCAST observations in order to optimize the success of GI observations.

7.2.1 Chopping and Nodding in Imaging Mode

FORCAST can used in two different imaging observation modes including: (1) symmetric nod-match-chop (NMC), (2) symmetric nod-perp-chop (NPC), and (3) asymmetric chop-offset-nod (C2NC2). NMC and NPC are both variations of the standard two-position chop with nod (C2N) IR observing mode. NMC is the default C2N observing mode and is the only C2N mode that will be available during Cy 5. A brief description of each observing mode is provided below, but users are strongly encouraged to read the more complete description provided in the FORCAST Observing Modes document.

  • NMC: In NMC mode, the chop is symmetric about the optical axis of the telescope with one of the two chop positions centered on the target. The nod throw is oriented 180° from the chop, i.e. anti-parallel, such that when the telescope nods, the source is located in the opposite chop position. The chop/nod subtraction results in two negative beams on either side of the positive beam, which is the sum of the source intensity in both nod positions and therefore has twice the intensity of either negative beam. This mode uses the standard ABBA nod cadence. An example of an observation taken in this mode is presented in the left panel of Figure 7-7.
  • NPC: NPC mode also uses a chop that is symmetric about the optical axis, but in this case the nod is perpendicular to the chop. The final images produced using NPC show four sources arranged in a parallelogram with alternating positive and negative beams. Unlike NMC, each beam in NPC has the same relative intensity. This mode also uses the standard ABBA nod cadence. The right side of Figure 7-7 shows data obtained using NPC. This mode will not be supported in Cycle 5.
  • C2NC2: The third supported observing mode is asymmetric chop with offset nod (C2NC2). This mode is particularly useful for large extended objects, smaller objects that are situated within crowded fields, or regions of diffuse emission with only limited sky positions suitable for background removal. In this mode, the chop throw is asymmetric, such that one chop position is centered on the optical axis (and the target) while the second (sky) position is off-axis. Rather than nodding, the telescope then slews to an offset position free of sources or significant background and the same chop pattern is repeated. Observations in C2NC2 mode follow a nod cadence of ABA and, by default, are dithered to remove correlated noise.

Figure 7-8 is a cartoon demonstrating how a C2NC2 observation might be designed for a large, extended object. Each source position (solid line) with its associated asymmetric chop position (dashed line) have matching colors. After a full chop cycle at each position, the telescope is slewed to a location off of the source, shown in black and labeled with the coordinates (600, 600). The chop throw and angle at that position is the same as it is for the source position to which it is referenced (not shown in the figure). It is immediately apparent from the figure, that C2NC2 has an efficiency of only ~20%. This is a much lower efficiency than either NMC or NPC since only a single chop position out of a full chop/nod cycle is on source. This should be taken into consideration when designing science proposals.


An example of an observation taken in NMC mode

Figure 7-7: Left: FORCAST image of β And taken in Nod-Match-Chop (NMC) mode. Right: FORCAST image of α Boo taken in Nod-Perp-Chop (NPC) mode.


Cartoon of a C2NC2 mode FORCAST observation with mosaic

Figure 7-8: Cartoon of a C2NC2 mode FORCAST observation with mosaic.

Once a proposal has been accepted, the GI, in collaboration with the SMO instrument scientist, will specify the details of chopping and nodding for each observation using the SOFIA observation preparation tool (SSPOT). Experienced GIs are encouraged to design their observations using SSPOT before writing their proposals to prevent the loss of observing time that might occur if, during Phase II, the observations are discovered to be more challenging than expected.

Following are a few of the most important issues that GIs should consider when preparing a proposal.

  • It is recommended that a near-IR or mid-IR database (e.g., 2MASS, Spitzer , WISE , MSX or IRAS) be checked to see if the target of interest is near other IR sources of emission. In the case of extended sources, where on-chip chop and nod is not possible, it is necessary to pick areas free of IR emission for the chop and nod positions to get proper background subtracted images.
  • If the IR emission from the region surrounding the source is restricted to a region smaller than half the FORCAST field of view (i.e. ∼1.6′ ) then the chop and nod can be done “on-chip”. Observations performed in NMC mode either “on-chip” or “off-chip” yield a S/N equal to or slightly better than that obtained in NPC mode. For additional discussion of this point, see the calculations of S/N for various FORCAST chop-nod scenarios provided here.
  • The chop throw is defined as the distance between the two chop positions in either symmetric or asymmetric chopping modes. When using a symmetric chop, e.g., in C2N mode, chopping and nodding can be performed in any direction for chop throws less than 584′′ .
  • When observing in one of the symmetric chop modes, large chop amplitudes may degrade the image quality due to the introduction of coma. This effect causes asymmetric smearing of the PSF parallel to the direction of the chop at a level of 2′′ per 1′ of chop amplitude.
  • When using an asymmetric chop, the maximum possible chop throw is 420''. However, some chop angles (as measured in the instrument reference frame) are not allowed for asymmetric chop throws between 250'' and 420''. Since the orientation of the instrument relative to the sky will not be known until the flight plan is generated, GIs requesting chop throws between 250 - 420'' are required to specify a range of possible chop angles from which the instrument scientists can choose when the flight plan is finalized.
  • For large, extended objects, it may not be possible to obtain clean background positions due to these limitations on the chop throw.

There are some additional considerations for observations of faint targets.

  • Currently, the longest nod dwell time (that is, the time spent in either the nod A or nod B position) for FORCAST is 30 sec in the SWC-only and dual channel modes and up to 120 sec in the LWC-only modes (depending on the filter). Proposers should run the exposure time estimator to determine if the object will be visible in a single A-B chop-subtracted, nod-subtracted pair, with an exposure time of 30 sec in each nod position. If the object is bright enough to be detectable with S/N greater than a few, it is recommended that dithering be used when observing in C2N mode. The dithering will mitigate the effects of bad pixels when the individual exposures are co-added.
  • If the object is not visible in a single A-B chop/nod-subtracted pair, with a nod dwell time of 30 sec in each nod position (60 sec integration), then dithering should NOT be used.

7.2.2 Mosaicking

If the source has an angular extent large enough that multiple pointings are required, the central position of each FORCAST field must be specified, with due consideration of the desired overlap of the individual frames. Mosaicking can be performed in each of the three available observing modes, NMC, NPC, and C2NC2. A sample mosaic using C2NC2 mode is demonstrated in Figure 7-8. One should keep in mind that for fields requiring large chop amplitudes, the effects of coma may compromise the image severely when in symmetric chopping mode (NMC and NPC).

7.3 Planning FORCAST Spectroscopic Observations

During Cycle 5, grism spectroscopy with FORCAST will only be available in single channel, long-slit modes. By default, observations will be set up using NMC aligned along the slit in long-slit mode and perpendicular to the slit in XD mode. Due to the size of the PSF, neither chopping or nodding along the slit nor dithering are possible for high-resolution XD observations. For larger sources and for targets embedded in crowded fields, GIs are advised to use C2NC2 mode.

The observing efficiency for FORCAST spectroscopic observations depends on a number of factors, including the observing mode, chop frequency and nod cadence, the detector frame rate, and LOS rewind cadence, to name a few. The typical observing efficiency as measured from Cycle 3 and 4 observations is of order 50-75% for most modes. Work is ongoing to optimize the mode-dependent efficiency values. These efficiency estimates are built-in to the SPT and do not need to be specified.

It is important to note that due to the fixed position of the grisms/slits in the filter/aperture wheels, the orientation of the slit on the sky will be dependent on the flight plan and will not be able to be predetermined. Further, the slit orientation rotates on the sky with each telescope Line-of-Sight (LOS) rewind. These limitations may be especially important to consider when proposing observations of extended objects.

7.4 Estimation of Exposure Times

7.4.1 Estimation of Exposure Times for Imaging Observations

The exposure times for FORCAST imaging observations should be estimated using the on-line exposure time calculator, SITE. SITE can be used to calculate the signal-to-noise ratio (S/N) for a given "total integration time", or to calculate the total integration time required to achieve a specified S/N. The total integration time used by SITE corresponds to the time actually spent integrating on-source without overheads. These integration times are used as input for SPT and for SSPOT, both of which will automatically calculate the necessary overheads. The format of the S/N values output by SITE depends on the source type. For Point Sources, the reported S/N is per resolution element, but for Extended Sources, it is the S/N per pixel.

For mosaic observations the total integration time required for a single field should be multiplied by the number of fields in the mosaic to obtain the total time, which is to be entered in SPT.

An important consideration in planning observations is whether FORCAST should be used in single channel mode, or in dual channel mode, since one gains the extra filter observation at the cost of lower system throughput in the individual bands. On the SITE form, the single channel mode is specified by selecting the filter of interest for one channel and "Ignore" on the other channel in the "Instrument properties" section.

7.4.2 Estimation of Exposure Times for Spectroscopic Observations

The exposure times for FORCAST grism mode spectroscopic observations should be estimated using the FORCAST on-line grism exposure time calculator. This calculator can be used to calculate the signal-to-noise ratio (S/N) for a given "total integration time", to calculate the total integration time required to achieve a specified S/N, or to estimate the limiting flux for a desired S/N.

In either case, overheads should not be included, as SPT calculates them independently.