If the redshift value is changed during a setup or in further editing of the AOR then the user is asked if the local oscillator (LO) frequency should be updated -- if previously set. On answering "yes" to accepting the LO frequency change, the local oscillator frequency is changed to compensate for the change in redshift/radial velocity entered. On top of this, the velocity of the LSR with respect to the SSB will be added.

It is important to note that the Geocentric frame reference has been removed.

The orbital motion of the Herschel Observatory itself is accounted for during mission planning once the date and time is known for observing the given target. The observed frequency will therefore frequently be slightly different from that requested. NOTE: Neither the inputs on the frame nor on the redshift are taken into account for moving targets. Therefore it makes little sense to enter a velocity shift for such targets as the corresponding shift observed on the LO frequency will in effect not be the one used (the necessary velocity shift between the spacecraft and target is calculated by the Herschel mission planning system and an appropriate LO frequency shift applied).

When setting up the range of frequencies for each of the spectrometers in use the "Frequency Editor" window is used. There is a lot of information about HIFI and how it works that is illustrated within this window (see Figure 13.3, “The HIFI Frequency Editor window”). At the top of the window -- at left and right -- bars are shown that show the available bandwidths of the WBS and HRS spectrometers given the resolutions selected in the instrument setup window.

Figure 13.3.  The Frequency Editor for setting up the local oscillator and HRS sub-band offset setting for HIFI. In this case, for band 1a, the spectrometers chosen were the WBS and HRS in parallel with both polarizations of the HRS to be made with "nominal resolution." Top left and rught show where requested lines appear in the upper and lower sidebands.

In Figure 13.3, “The HIFI Frequency Editor window” three spectrometer lines are shown, one for the WBS spectrometer and two for the two subbands available using the HRS spectrometer in "nominal resolution."

Emission/absorption lines displayed

The positions of known emission/absorption lines are indicated by lines appearing above the upper frequency scale with different colours (each species has its own colour) being used to indicate different species. HSpot contains a default line list containing most of the stronger lines available in the HIFI frequency range.

	Note
	The displayed/available lines are the redshifted lines available within the band.

The list of lines can be added to from on-line catalogues and/or individual line information provided by the user (see Chapter 18, Spectral Lines for Planning Purposes ). It is also possible to change the width and height displayed of chosen lines in the "Lines" menu in HSpot. However, this is NOT a simulation of the expected spectrum from any given source, but merely an indication of where lines appear and to provide a rough indication of possible "contaminating" lines that might be superimposed from the second sideband that will not contain the main signal of interest.

A left mouse-click near one of the lines appearing in the display will display its identification and stored frequency (see Figure 13.3, “The HIFI Frequency Editor window”).

UPDATE: From HSpot 5.2, the Frequency Editor includes illustration of where emission-lines appear in the upper and lower sideband windows. This provides a simple illustration of what the spectrum being planned will look like when looking at either sideband. Lines associated with a specific sideband appear as solid whereas additional lines that come from the other sideband appear as dotted lines (see top right of Figure 13.3, “The HIFI Frequency Editor window”, for example). These can be used to indicate the likely overlap of spectral lines in the double sideband spectra of HIFI, the standard product of HIFI.

The HRS sub-band sliders can be moved via a mouse-click-and-drag to positions available to them (within surrounding rectangular box) or via the pull-down menus in the table at the bottom of the frequency editor screen (see Section 13.2.6.1, “Frequency Editor Window”). The frequencies being sampled by the WBS and HRS can be seen in the scale below the bandwidth bars.

The LO frequency can be changed via the slider on the second scale shown. The slider has a range over all the available frequencies for the band chosen. As the LO frequency is changed (click-and-drag) the sampled spectrometer frequencies are also changed.

An alternate means of changing the LO frequency, which also has the advantage of automatically placing the line at the best position within the WBS spectrometer, is via the pull-down line and transition menus. The available spectrometer bands have check marks beside them. For these spectrometers, we can optimally place a selected line in either the upper or lower sideband. To select a line and have the spectrometer set up adjusted for you.

We can set up an observation to use the WBS and HRS spectrometers simultaneously to observe more than one emission line at a time spaced close to each other in frequency space. For example the CS 10-9 and CI 1-0 emission lines in band 1a of HIFI.

After selecting the mixer band to be band 1a in the HIFI point or mapping mode AOT window and selecting to "Add" a frequency within the "Frequencies" window we can set up the example by doing the following.

The example is illustrated in Figure 13.4, “ An example of a setup around the CS 10-9 line in Band 1a of HIFI. ” where the line CS 10-9 was chosen in band 1a. It was chosen to appear in the lower sideband (checkbox is blank). We can identify it via a left button mouse-click on the dark red line appearing above the frequency scale. This appears in the figure as a blue box with the line id and frequency in GHz. The id information is placed immediately to the right of the line clicked on.

Note that the line chosen is NOT placed at the centre of the lower sideband IF bandpass, but is placed in a "best IF frequency" position. Lines appearing in other parts of the IF band are likely to have higher noise levels. More information on the variation of sensitivity across the bandpass from the different HIFI mixers is given in the HIFI Observers' Manual (http://herschel.esac.esa.int/Docs/HIFI/pdf/hifi_om.pdf.

Figure 13.4.  An example of a setup around the CS 10-9 line in Band 1a of HIFI.

Note that a warning sign will normally come up to tell you that you are moving the LO by more than 2GHz. Due to the limits of placement of the HRS sub-bands, moving the LO by more than this amount means that all settings (bar the one just changed with the pull-down menus) will be reset to their default "No line" status, otherwise they would be out of the available frequency range of the instrument for the LO setting.

Note too that, when choosing a line from the pull down menus, the position of the line is not in the centre of the IF band (centre of spectrometer range). There are two reasons -- the centre marks a boundary between two of the four CCDs that form the detector of the WBS and the area chosen has the best noise levels within the IF band.

Note

The sensitivities across the IF band do not vary much for bands 1, 2 and 5 but degrade towards the edges of the 4GHz bandpass for bands 3 and 4. Bands 6 and 7 have a steadily decreasing sensitivity towards the upper edges of their bandpass (towards the far left and right on the HSpot screen). Please see the HIFI Observers' Manual (http://herschel.esac.esa.int/Docs/HIFI/pdf/hifi_om.pdf for more details.

The selection of a given line/transition via the pull-down menus can be done for both the WBS and HRS. But -- the HRS sub-bands can not move outside of the 4GHz (or 2.4GHz for the case of bands 6 and 7) sideband ranges. If you try to pick a line that would require a different LO setting a warning will come up and you will be asked if you really wish to proceed (see Figure 13.5, “Warning pop-up that comes up if there is a large LO frequency move requested.”). In the case shown, a warning is given when trying to centre on the CI 1-0 line after having previously setup the spectrometers to look at the CS 10-9 line in band 1a. It is possible to look at these lines simultaneously. An example is presented below.

Note also that when the chosen line cannot be located at the place of the IF band, typically because the LO frequency implied falls outside of the operational range for the chosen band, a warning message will appear to inform about this. Just click "OK", and the LO frequency will be set accordingly.

Figure 13.5. Warning pop-up that comes up if there is a large LO frequency move requested.

An example setup is shown in Figure 13.6, “ Example setup for both spectrometer sets” for a combined WBS & HRS observation of CS 10-9 and CI 1-0 lines, with the HRS in nominal resolution. In this case, there are two HRS sub-bands available to us. To obtain what is illustrated here the following was done.

Figure 13.6.  Example setup for using the Wide-band (WBS) and High resolution (HRS) spectrometers of HIFI simultaneously.

It is also possible to type in a frequency for the frequency positions of both the WBS and the HRS sub-bands. From HSpot 5.3.1 onwards, thereis a new feature allowing you also to enter the line frequency in the Local Standard of Rest (new column "Rest (GHz)"). This allows users to typically enter known rest frequencies from the laboratory, and HSpot will automatically apply the corresponding redshift and set the "Observed" frequency accordingly. Of course, the "Observed" frequency can still be directly filled in, and the correspond "Rest" frequency value will be reflected.

From HSpot 6.0.0 onwards, the selection of a frequency via typed input in the frequency editor results in the creation of a line tick illustrating the location of this line in the frequency slider. This line tick will behave similarly as any other line tick stemming from line catalogues active during the HSpot session.

Note that the frequency chosen for "Observed" frequency is automatically placed in the "best" position within the available WBS band, as for the pull down menu line selection. When the WBS is used first to set the LO frequency with the typed inputs, the placement of subsequent HRS sub-bands will not necessarily follow this rule and simply place the HRS sub-band where the user has instructed HSpot. Note that the typed frequency corresponds to the centre of the HRS sub-band, so the width of the sub-band will be taken into account to check whether the typed frequency is compatible with the LO Frequency assigned via the WBS entry. On the other hand, when the HRS is used first to set the LO frequency by typed inputs, the IF placement optimisation will be applied.

An illustration of this is given in Figure 13.7, “ Illustration of typed-in frequency setting for HIFI spectrometers”, Illustration of a typed-in frequency setting for HIFI spectrometers, which again uses band 1a. In order to provide this setup we did the following.

Figure 13.7.  Illustration of typed-in frequency setting for HIFI spectrometers

During continued monitoring of the HIFI instrument it has been noted that several "spurs" can exist in the data at certain frequency settings of the instrument. These can range from weak spurs that affect a small protion of the observed spectrum to spurs that can completely ruin the data from one or more of the 1GHz bands of the WBS or one or more subbands of the HRS. Typically identifiable as different in nature to spectral line features seen in data, spurs can cause bad settings of the instrument (in the case of very strong spurs) or complicate and overlap with the spectral features the astronomer is looking for.

Further settings can also have general instabilities and small frequency regions available to HIFI may have impure local oscillator frequencies causing line flux loss and the appearance of spurious lines.

The current knowledge of spurs in the system has been built into HSpot so that users receive a warning whenever a new frequency is selected in the frequeny editor that is likely to have a spur visible in the data received. The warning indicates the nature, strength and position (in terms of one or more of the four WBS spectrometer subbands) where a problem may appear. Typically, users should stay clear of these frequency settings, and for single line observations simply selecting to switch the sidebands in which the lines may be the simplest solution for avoiding spurs. Relatively small changes of a few GHz in the local oscillator frequency (i.e., small changes using the slide bar in the frequency editor) may be all that is necessary to avoid spurs too -- still keeping the spectral lines of interest visible in the spectrometer(s).

It should be noted that less than 5 per cent of the local oscillator frequencies have any sort of associated problems.

An example spur warning is shown in Figure 13.8, “HIFI spur warning”.

Figure 13.8.  An example spur warning in HIFI band 4b at 1098.34GHz local oscillator frequency. Information is provided at the level of 1GHz windows. The WBS has 4 x 1GHz wide linear CCDs. In this illustration, the warning indicates that the current setting is likely to provide data that will contain both a weak and strong spur in the first of these WBS 1GHz windows.

After hitting the "OK" button on the frequency editor window the user is returned to the "Frequencies" window. A descriptive line of the frequencies chosen is given in the window (see Figure 13.9, “ "Frequencies" window indicating the frequency settings stored for the current observation.”). Settings can be modified by clicking on the line in the table to be modified and clicking the "Modify" button.

Figure 13.9.  "Frequencies" window indicating the frequency settings stored for the current observation.

If you are happy with the frequency settings, then clicking "OK" in the "Frequencies" window completes the selection and returns you to the "AOT" window.

In dual beam switch (DBS), the HIFI chopper is used to switch the beam from looking at the target object and looking at a position 3 arcminutes from this (along the X axis of the spacecraft -- see the Herschel Observatory Manual). The second half of the cycle has the telescope move 3 arcminutes to place the chopped beam on the target position. Chopping now continues between the target and a position 3 arcminutes on the other side of the target. Calibration measurements are made at regular intervals dependant on the stability of the HIFI band being used (see the HIFI Observers' Manual (http://herschel.esac.esa.int/Docs/HIFI/pdf/hifi.pdf).

The standard (slow) chopping rate provides an approximately 16-second cycle with an ABBA chopping cycle, where A is the off target position and B the on target position. So 4 seconds are used per readout, which allows readout of all spectrometers.

The user has two choices to make in this mode (see Figure 13.11, “ Options for the HIFI Dual Beam Switch mode.”). Note also that the chopper speed is automatically increased when large frequency resolutions are used for the time estimator (e.g., 100MHz or more). The chopper speed is reported in the messages from the time estimator when using this mode. See Section 13.2.8, “Time Estimator Settings” for more details on the use of the time estimator.

1. Fast chop uses a ABABAB structure with frequencies between 0.25 and 2 seconds 
spent at each on/off position (the chopper speed is always provided in the
time estimator messages any time the chopper is used). This is particularly 
useful for measurements of weak broad lines and where stability can be an issue 
(such as any observation in bands 6 and 7). It is very slightly less efficient 
than the standard slow chop, which is the default. Or...
2. Continuum measurement. For accurate continuum level measurements, 
timing/frequency of calibration measurements is adapted to reduce drift 
noise effects on the continuum level measured. 

Figure 13.11.  Options for the HIFI Dual Beam Switch mode.

If, by chopping 3 arcminutes, the off position would not be placed in an emission-free area a position on the sky can be selected for the off reference position in the position switch mode (see Figure 13.12, “ Options for the HIFI Position Switch mode.”). However, the reference position must be within 2 degrees of the target.

The frequency of calibrations and switching is determined by the backend logic of HSpot when the time estimation is calculated for this mode, and is based on the distance between the reference and target positions. Reference positions at greater distances from the target have correspondingly lower observing efficiencies and will take longer to achieve a given noise level.

Figure 13.12.  Options for the HIFI Position Switch mode.

Identifying a reference position

The reference position can either be given as an offset in RA and Dec from the target (in arcminutes) or its position given explicitly.

	Note
	The reference position MUST be within 2 degrees of the target.

Click the appropriate button to identify whether an offset or a specific position is to be used.

If a specific position is to be input, the user has the choice of either typing in a decimal RA and Dec (J2000) in degrees, or by clicking on the "Choose Position" button. This brings up the window shown in Figure 13.13, “OFF position target input window”, which provides a wider range of possible inputs for the reference (OFF) position of the position switch mode.

Figure 13.13.  Dialogue window for identifying the OFF position for the HIFI modes. Note that this uses a similar dialogue to the standard Target input window.

Sometimes we can use the reference of a second spectrum taken with a slightly different LO setting. Doing this has the added advantage that the two spectra taken substantially overlap. Data can therefore be taken very efficiently in this mode (nearly 100%), but gain ripples are likely to remain. It is therefore recommended that an OFF position be used. It is expected that gain ripples will be removed or, at least, reduced by using a reference OFF position in the sky and repeating the same set of measurements.

This mode provides a choice of the frequency switch "throw" (the difference between the two LO frequencies to be used) and the choice of an OFF position (see Figure 13.14, “ Options for HIFI Frequency Switch pointed mode.”). For the frequency throw the user uses a pull-down menu of available frequency throws. Currently there are 4 values allowed, "small positive" (the default), "large positive", "small negative" and "large negative". Small and large are of order 100 and 200-300 MHz respectively, depending on subband in use (see Table 13.1, “Frequency throw values available by HIFI subband.” for details).

The position of any OFF position is done in the same way as for "Position Switch" mode (see the section called “Identifying a reference position”). Note that for frequency switch observations the choice of an OFF position is optional. Use of an OFF position makes the observation more robust against residual baseline ripples after data processing.

Figure 13.14.  Options for HIFI Frequency Switch pointed mode.

This mode allows a reference measurement to be taken against the internal loads of HIFI. This can be particularly useful when a sky reference position is not within 2 degrees of the target or when a large slew would lead to high drift noise (which can be the case for bands 6 and 7 observations).

The HIFI chopper is moved between the on target position and the position of the hot/cold internal calibrator loads of HIFI in order to remove gain variations. Due to the likelihood of residual ripples, this should also be done on a reference OFF position.

The input from the user only requires information about any required OFF position to be defined (see Figure 13.15, “ Options for HIFI Load Chop pointed mode.”). The OFF position is included in the same way as for the "Position Switch" mode (see the section called “Identifying a reference position”) and, likely the Frequency Switch case, is an optional input.

Figure 13.15.  Options for HIFI Load Chop pointed mode.

This is done via the pull-down menu at the top of the time estimator window (see Figure 13.16, “HIFI time estimator settings window.”). The default is "Time" where the user indicates how much time he/she wishes to spend on the observation.

To change the minimum and maximum resolution width (Res. Min. Width and Res. Max. Width respectively) values, simply click on the box in the table below each of the headings and input the required value. For chopping modes, a higher chopping speed can be obtained when a larger resolution width is required, so unchecking the 1GHz box and putting in a large goal resolution..

The maximum and minimum widths should be representative of widest line widths expected (Res. Max. Width) and the highest resolution/smallest line width wanted for the science goals when (for instance) looking at line profiles. The noise estimates provided will then be the noise across the width of a profile and the noise for the smallest resolution bin you have put in (Res. Min. Width).

To input a time goal (how much time to be spent on the observation), the user clicks on the table cell in the "Time (s)" column. The goal for the observation should be entered, in seconds.

If "Noise" is chosen from the pull-down menu in the time estimator settings screen, a noise goal (in degrees Kelvin antenna temperature for a single sideband and the average of the two polarizations) can be input. Click on the table cell below "Noise (K)" and input a value.

The HIFI AOT logic uses a code to compute optimal observing sequences based on user input parameters. Information on the sequence chosen is displayed with the "Show Messages" button. Doing so provides a message window to show sequence/observation information (see Figure 13.18, “HIFI message information”, messages returned for our position switch observation).

Figure 13.17. Time return information from the HIFI time estimator for a Dual Beam Switch observation. In this case a total observing time of 180 seconds was requested as a user goal. Total instrument observing time returned is 370 seconds with 180 seconds observatory overhead added for the initial slew to the target. Noise values for the upper and lower sideband and each of the minimum and maximum goal resolutions (0.48 and 10 MHz) are returned (in the standard HIFI units of degrees Kelvin for combined H and V polarizations).

Figure 13.18.  Time return information extended to show HIFI messages. This indicates the sequence of measurements taken by HIFI -- chop/switch cycles in each position and nod positions for the example case -- during the course of the observation request and time taken both ON and OFF the target. The chopper speed is also indicated. The efficiency of the mode indicates the time/noise compared to an ideal instrument looking at the target throughout the observing time.

Such parameters in the message information differ slightly for each observing mode, but provide fundamentally similar sets of information.

Always provided for each mode are the following:

Message information can be saved to disk using the "Save messages" button at the bottom of the messages window. Messages are also automatically saved in the AOR file of observations after a time estimate has been made, including when batch estimates are made using the "Recompute All Estimates..." selection under the "Tools" menu within the main window of HSpot.