NOVEMBER 1, 2018
STIS NEWSLETTERS

November 2018 STAN

This STAN provides various updates to STIS proposing, including changes to the STIS review process for PIs, common acquisition errors and how to avoid them, and how high jitter can impact coronagraphic data.

Changes to STIS Technical Review Procedure

HST is a limited resource and as such, all PIs have a vested interest in ensuring that safety protocols are followed and that programs are executed successfully. Beginning in Cycle 26, the STIS team is implementing new policies that are aimed at increasing the efficiency that M dwarf observations are screened for safety and facilitating a higher rate of finding and fixing errors during technical reviews.

1. Providing M dwarf characteristics for BOT clearing

PIs observing M dwarfs with the MAMA detectors are responsible for demonstrating the safety of their proposed targets both during quiescence and during flaring states, following the guidlines in ISR 2017-02. Contact scientists on the STIS team are required to make an independent calculation to verify safety. To facilitate this process, we will be sending PIs of M dwarf programs a spreadsheet to fill in the information required to make these safety checks. This policy is in line with what the COS team currently does. Ideally, the spreadsheet should be returned by the Phase II deadline so that safety violations and mitigation strategies can be identified as early as possible. For large or complex programs where such a deadline would present an undue burden, supplying this information on an extended timeline is acceptable. However, no visit can be cleared for observation until the requested information for that visit's target is provided. 

2. Supplying ETC IDs for target acquisitions

We are requesting that PIs provide ETC run numbers for their target acquisitions. Although APT will allow a Phase II submission without these numbers, PIs are expected to comply with requests for missing ETC IDs from their instrument reviewers before programs are cleared for observation.

Why the change? The STIS team does not have the resources to independently verify all specifications made in the Phase II, and such in-depth verifications are restricted to health and safety reviews of the proposed science target and surrounding field to prevent accidental damage to the MAMA detectors. STIS instrument scientists do perform a basic technical feasibility review, and providing ETC IDs for target acquisitions allows reviewers to quickly assess whether the minimum S/N will be met and whether the inputs to the ETC are broadly reasonable. Improper target acquisition specifications has lead to loss of science and could risk the safety of the instrument if HST acquires the wrong target before a MAMA observation.

As always, PIs still retain the ultimate responsibility of correctly designing their observations to meet their science goals.

 

 


 

 

How to Avoid Common Acquisition Errors

While all of the information required to specify a successful acquisition sequence is documented in Section 8.2 of the STIS Instrument Handbook, there are some common misconceptions about how the target location algorithm works that can cause users to overlook relevant sections of these instructions. Often, the trouble stems from insufficient knowledge of the area surrounding the science target. Sometimes ancillary observations, such as HST pre-imaging, are required to be successful.

The target location algorithm is simplistic. After an on-board CR-split combination, the algorithm sums the total counts within a checkbox centered at every pixel in the image and chooses the checkbox location with the largest summed counts. This is essentially smoothing with a 2-D boxcar defined by the checkbox size and selecting the brightest pixel in the smoothed image. The target location is refined by either computing a flux weighted average (all point source acquisitions and diffuse acquisitions with DIFFUSE-CENTER = FLUX-CENTROID) or using the geometric center (diffuse acquisitions with DIFFUSE-CENTER = GEOMETRIC-CENTER) of the selected checkbox.

Three common situations that cause the algorithm to miss the desired target are:

1. The science target is not the brightest object within 5'' of the coordinates.

No matter how accurately the science coordinates are specified or how precisely the telescope points in the blind pointing stage, the centering algorithm will find the brightest object in the ACQ image. Note for diffuse acquisitions with large checkbox sizes, a field larger than 5x5'' is checked. If a brighter object is nearby, acquiring an offset object can be used to avoid this pitfall (see STIS IHB 8.2.3). Consult Section 3.2.4 of the Phase II instructions to properly specify the coordinates of an offset target, and use the target confirmation charts in APT to check that the magnitude and direction of the offset is correct.

2. Neglecting nearby diffuse emission for point source acquisitions and vice versa.

The algorithm does not perform a consistency check that the type of object found matches the type of acquisition. A bright point source summed over a large checkbox can still generate more counts than extended emission filling the entire checkbox. Similarly, extended emission can generate more counts in a small checkbox than a bright point source, as illustrated in the image below. For targets located in a crowded or complex field, the STIS team provides a TA Simulator  to help predict how the algorithm will behave. Users will have to modify their input images to mimic a STIS ACQ image (i.e., match the plate scale, image size, and approximate bandpass).

Common Acquisition Errors Fig. 1
Figure 1: The intended target is the brightest point source in the image. However, the extended galaxy emission summed over the 3x3 checkbox is brighter still.

3. Inappropriate checkbox sizes for diffuse objects.

Even when the field is free of other bright sources, faint emission or random noise can lead to centering errors if the checkbox size is too large. The checkbox size should be as small as possible to center the object. This means it should be roughly the size of the object (or substructure) of interest.

Common Acquisition Errors Fig. 2
Figure 2: Example of how different checkbox sizes find different targets in a complicated scene. In this example, assume the desired target is the bright, extended but relatively compact feature in the upper middle of the image. The brightest checkbox locations (x's) and corresponding checkboxes (squares) for different checkbox sizes of 3x3 (red), 5x5 (yellow), and 19x19 (blue) are indicated.
Common Acquisition Errors Fig. 3
Figure 3: These three images show the summed checkbox fluxes at each pixel location. These are to help visualize how the algorithm identified the best checkboxes in Figure 2. (Left:) Because the target of interest is larger than 3 pixels on a side, its full flux is never captured in any 3x3 checkbox. Thus, the algorithm finds the relatively bright point source in the lower right. (Middle:) The 5x5 checkbox size captures the target object's full flux when roughly centered on it. The point source in the lower right has its flux diluted by nearby empty pixels. Therefore, the algorithm correctly finds the target. (Right:) The 19x19 checkbox is much bigger than the object of interest. Therefore, the full flux of the target is included in many checkbox locations. In this example, the diffuse background emission to the right of the object pulls the centering in that direction. Random noise can have a similar but more unpredictable effect.

 

 


 

 

Impact of High Jitter on Coronagraphic Observations

Between April 21,2018 and October 5, 2018, the Gyro 2-4-6 combination on HST resulted in elevated spacecraft jitter. In response to this increase, the STIS team has carefully assessed which observing modes are directly affected. A previous STAN article addressed the impact of elevated jitter on spectroscopic modes, but coronagraphy, especially with the smallest inner working angles, can be impacted by spacecraft jitter as well.

The spacecraft jitter in the April to October, 2018 timeframe was as high as 15 mas along the V3 spacecraft direction, which is oriented ~45 degrees clockwise from the detector Y direction. Local spacecraft excursions of up to 70 mas were observed over short periods of time. To directly address the impact that jitter might have on a coronagraphic observation, the STIS team executed the special calibration Program 15603 on September 23, 2018. This program targeted HD 38393, a bright solar-type star used to obtain high contrast images using the BAR5 occulter in a previous special calibration program (14426).

The program replicated a single orbit of 14426 by placing the star behind the mask and taking 0.2s exposures over an entire orbit. Using the diffraction spikes of the star to determine the stellar centroid on the detector, we measured the instantaneous stellar position on the detector as a function of time. Relative to the average centroids, root-mean-square (RMS) jitter over the orbit was measured to be 16 mas (see Figure 1), consistent with Fine Guidance Sensor telemetry during that time period.

Impact of High Jitter Fig. 1
Figure 1

Initial analysis of the program data (Figure 2, left) reveals that the raw noise within ~0.75” of the mask edges have enhanced systematic noise compared to the 14426 data (Figure 2, right). The noise is up to 2 times higher than what was seen in 14426, where average jitter was ~3 mas. Users who obtained coronagraphic observations during this time period of elevated jitter should assess if any noticeable impact to their data was seen. It is likely that this impact is smaller for other coronagraphic aperture locations. It should also be noted that the flux levels in each pixel near the mask edge are highly correlated with the exact position of the star. Users are encouraged to contact the help desk if they have further questions, or they can download the publicly available 14426 and 15603 datasets from the archive.

Impact of High Jitter Fig. 2
Figure 2

Please Contact the HST Help Desk with any Questions

https://hsthelp.stsci.edu.