Automated HST/STIS reference file generation ...

Automated HST/STIS reference file generation pipeline using OPUS Rosa I. Diaz, Michel Swam and Paul Goudfrooija a Space

Telescope Science Institute, Baltimore, USA; ABSTRACT

Bias and Dark reference files are part of the basic reduction of the CCD data taken by the Space Telescope Imaging Spectrograph (STIS) aboard the Hubble Space Telescope (HST). At STScI, the STIS team has been creating these reference files using the Bias and Dark Pipeline. This pipeline system starts with automatic retrieval of bias and dark exposures from the HST archive after they have been ingested. After data retrieval, a number of automatic scripts is executed in a manner compatible with the OPUS pipeline architecture. We encourage any group looking to streamline a stepwise calibration process to look into this software. Keywords: OPUS, HST, pipeline, STScI, STIS, PyRAF, calibration

1. INTRODUCTION Bias and Dark reference files are a key element of the calibration process of any Charge Coupled Device (CCD) science observation. In case of CCD data taken outside the Earth’s atmosphere (such as STIS aboard HST), a proper removal of bias and dark current is especially important, given the strong effects of a high flux of powerful cosmic rays. Bias and dark reduction involves the removal of the 2-dimensional structures of the ”bias” and dark current from the science image. This includes ”hot” pixels and columns which change on short time scales, as well as a more extended structure which changes more slowly with time elapsed since the installation of the instrument on HST1 . For STIS, and other HST instruments, these files are not taken contemporaneous to science observations but rather as part of the internal STScI calibration programs, which allow the execution of these observations during the periods where the instrument is not in use. (Usually during occultations or in parallel to another instrument’s observations.) Making these reference files independent of science observations is practical and results in higher return for HST observers, who can dedicate more time to their scientific programs and therefore increasing the scientific return of the observatory. Since the installation of STIS and in support of HST observations, dark and bias images have been taken on a regular basis. The frequency with which these files are taken, depend on the number of images needed to produce a high signal-to-noise (S/N) ratio reference file and to the rate with which hot columns and hot pixels appear between anneals. A high S/N reference file is necessary in order for these to have a negligible contribution to the effective noise of the science data. Given that the opportunities to schedule calibration exposures are limited, frames are accumulated until a S/N = 1 per pixel is achieved. Although the process followed to create these files involves the use of well defined steps and scripts, it can take considerable amount of time to walk through each of them because of the volume of data used, the frequency in which these calibration files are created, and the variety of modes in which the STIS CCD operates. Therefore, in order to provide observers with the best possible calibration products in an acceptable amount of time, a Bias and Dark Pipeline was created using the STScI OPUS pipeline architecture2 ; which is the backbones of the HST calibration pipelines used by the STScI Data and Archive Distribution System.3 This pipeline supplies the automation of the different stages needed to create these files and a user friendly interface that the STIS team can use to monitor their creation. In section 2 of this paper, we give an overview of the procedures followed for the creation of STIS Bias and Dark reference files and how these vary for the different CCD observing modes. Further author information: (Send correspondence to R.D.) R.D.: E-mail: [email protected], Telephone: 1 410 338 4372 M.S.: E-mail: [email protected], Telephone: 1 410 338 4593 P.G.:E-mail: [email protected], Telephoe: 1 410 338 4981

Advanced Software and Control for Astronomy II, edited by Alan Bridger, Nicole M. Radziwill Proc. of SPIE Vol. 7019, 701931, (2008) · 0277-786X/08/$18 · doi: 10.1117/12.790275

Proc. of SPIE Vol. 7019 701931-1 2008 SPIE Digital Library -- Subscriber Archive Copy

Section 3 describes the OPUS system, and why it is used for this pipeline. It also describes the details of the STIS Bias and Dark Pipeline, and discusses future development of the pipeline and applicability to other projects here at STScI.

2. STIS CCD BIAS AND DARK CALIBRATION FILES The STIS CCD bias and dark reference files are used during the basic two-dimensional image reduction of STIS data. In this stage, the bias files are used to subtract the two-dimensional bias structure present in the CCD data, while the dark reference files remove the dark signal of the data; i.e., the count rate created in the detector in the absence of photons from the sky. More detailed information about the basic calibration process has been discussed elsewhere,4, 5 as well as the details on the Bias and Dark creation and testing procedures.1, 6–8 Here we will only provide a summary of the process. Bias reference files or superbias images are the combination of a well defined number of input bias frames for a given gain and binning combination. Each of the files is corrected for the bias level indicated by the overscan areas in the detector. These files are then combined together to facilitate cosmic ray removal. Cosmic rays are removed using an iterative rejection threshold of σ = 4, 5, and 3, respectively. The resulting average image is flagged with the location of the hot pixels; which is done by first creating a median-filtered version of the averaged bias (using a kernel = 15 x 3 pixels). This image is then subtracted from the averaged bias, leaving a ”residual” image containing hot pixels and columns. The pixels hotter than 5 σ of the effective RMS noise in the residual bias image are then flagged as hot pixels in the Data Quality (DQ) extension of the output reference bias file. Darks, on the other hand, are a combination of 2 images. The first is a ”baseline dark” image which is a (typically) monthly average of (cosmic-ray-rejected) dark images, normalized to an exposure time of 1 second. The second image is constructed using darks spanning a period of a ”week”. These darks are combined together (cosmic- rays are rejected in the combination), and normalized to an exposure time of 1 second. Hot pixels in that normalized dark are then flagged in the baseline dark DQ extension. In this case, the hot pixels have a value higher than the baseline dark current + 5 σ of the new dark. The pixels hotter than 0.1 electron/sec in this dark are being assigned a DQ value of 16. Previous hot pixels that have fully annealed out between the observing dates of the baseline and the new dark are being assigned a value equal to that in a median-filtered (kernel = 2x2 pixels) version of the baseline dark. The frequency with which the Bias and Dark need to be created depend on the particular combination of gain and binning of the STIS CCD. In particular, for the Gain = 1 un-binned data, Bias and Dark reference files are created on an approximately weekly basis (+ − 2 days). In the case of Gain = 1 binning 1x2, 2x1, and 2x2, these reference files are created on an approximately biweekly basis (+ − 2 days); while the less used gain-binning combinations (Gain = 1, binning = 4x1, 4x2 and un-binned Gain = 4) are produced only biannually. These reference file creation frequencies were arrived at by taking into account the time available to schedule bias and dark exposures, the usage of a given data taking mode, and the number of frames needed to achieve a S/N = 1 per pixel. Table 1 of ISR 99-089 provides information about the number of STIS CCD bias frames to be combined together in order to obtain S/N = 1 per pixel for superbias images taken in all supported gain and binning settings.

3. THE PIPELINE The complexity and well-defined procedural steps used to create the STIS Bias and Dark reference frames, makes these steps a perfect candidate for automation. To speed up the process with which these files are created and made available to users, a pipeline was constructed using the OPUS architecture. While the OPUS pipeline was developed for the general calibration of HST data, it can easily be extended to any systematic set of calibration steps for data from any telescope or instrument. The Reference Files pipeline is also robust enough to provide easy handling of schedule changes or re-triggering of the pipeline steps at any stage of the process. This is particularly useful when sudden changes in the observing schedule occur and the Dark and Bias frames have to be redistributed across ”week” or ”biweek” boundaries. Using the pipeline makes the process for recreating the reference files quite simple.

Proc. of SPIE Vol. 7019 701931-2

OPUS is an event-driven pipeline that provides the right architecture to automate the creation of the bias and dark reference files starting from the retrieval of bias and dark frames immediately after these are available in the STScI archive, and ending with the creation of ”weekly”, ”biweekly”, and biannual bias and dark reference files for all gain and binning combinations. This process is accelerated by the capabilities that the software provides. The OPUS blackboard interface allows to run multiple instances of the same processes, in a network of machines, against different set of data and disk directories, detect errors, as well as to re-trigger stages of the pipeline processes as needed. The pipeline currently operates on a UNIX-based set of workstations (Solaris and Linux now and soon for Mac environment) and implements all its stages either as UNIX shell, Perl, or Python/PyRAF scripts. These scripts use a combination of system commands, utilities, OPUS programs and STSDAS tasks to perform the necessary processing and database access to the OPUS, Science Planning and Scheduling System (SPSS), and Data Archive and Distribution System (DADS) databases. In order to better understand the advantages provided by the OPUS pipeline architecture, we describe briefly the stages involved in the creation of bias and dark reference files below.

3.1. The Bias and Dark Reference File Pipeline There are nine main stages for the creation of STIS Bias and Dark reference files, shown in Figure 1. Each of them is awakened periodically and triggered whenever an execute condition is found. This execute condition is set up by the success of a unique pre-required stage, some of them visible through the Observation Manager Graphical display (OMG) and others running behind the scenes. Those processes that are controllable, and therefore visible in the OMG (Figure 1) via an Observation Status File (OSF), are those stages that correspond to actual scripts written by the STIS team in order to produce the bias and dark reference files. Table 1 lists all the pipeline stages and provides a brief description of the steps involved in each stage. Those processes not showing in the OMG pipeline stages (e.g. fill anneals, fill ref asns, polarc) are those that require special communication with the different STScI Data Management Systems (e.g. scheduling database, archive database) as well as those populating the STIS Dark and Bias reference files database (OPUS DB). The latter contains the information on the data used for the creation of the reference files (ref asns table) and how the dark and bias reference files should be subdivided by ”week” and ”bi-week” (ref period table) for any given anneal period. An anneal is a procedure used to minimize the number of hot-pixels in the CCD detector. Many of these hot pixels can be repaired or annealed by warming the CCD from its normal operating temperature near -83oC up to the ambient instrument temperature of about +5o C for several hours10 Figure 2 shows the different resources the pipeline uses in the process of creating the Bias and Dark reference files; while Figure 3, provides a flow diagram for the Bias and Dark pipeline. Upon receipt of a new Science Mission Schedule (SMS) the first stage of the pipeline is triggered (fill anneals). This task starts by looking at the SPSS database for the dates of anneal occurrences. (This particular process does not appear in the OMG.) Using this information the task sets the boundaries for the creation of the appropriate number of ”weekly” or ”bi-weekly” reference files, depending on the CCD gain and binning mode, and the information is loaded into the ref periods table. Here we refer to ”weeks” as the period of time that spans between 5 to 9 days, while a ”bi-week” can be between 12 and 16 days. For example, for gain 1 bin 1x1 it usually means four ”week” periods between 5 to 9 days each. A successful completion of this process will trigger the fill ref asn process that looks again into the the SPSS database. This time it looks for scheduled bias and dark observations that fall within the anneal period. Polarc, on the other hand, is a time poller that awakens periodically and determines if all the requisite exposures needed to produce a particular Bias or Dark reference file, as documented in the ref asns table, are present in the archive. If so, a working directory is created in the reference file pipeline area and an OSF is created on the reference file blackboard to start production of the reference files. The Bias and Dark pipeline creates one OSF per gain, bin and week combination. It verifies the data available for each of them and executes the different steps as needed. It also supports running several instances of the OMG stages simultaneously if necessary. The number of instances that a given stage can run is unlimited, and set via the OPUS Process Manager (PMG; see Fig 4). This can improve the processing throughput by keeping the CPU and I/O systems loaded. It can also improve the robustness in the event of hardware problems; that is, by having different copies of processes running on different machines it is possible for other machines in the cluster to continue to process the data if one machine goes down with a hardware failure.

Proc. of SPIE Vol. 7019 701931-3

0MG: opusObservation 0he,tinn Mng: stJef OMG: OPUS Manager: stiref File Helpp Foe Edit Edit Manage ManageView ViewTools 10015

OSFS OSFs

time_stamp time stamp 2001 09/13 16 50:23 2001_09/13 16:50:23 2001_09/13 16:50:24

I

dataset dataset

bibias_111_07635_01_wk1 5511]. 07635 01 skI

obs_cmd - data data_idwi obs_em

aol sti

flati

dark_111_07635_01_wk1

sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti

bies 112 a7635 01 bsk1 bias_112_07635_01_biwk1 blea 121 0.7635 01 biskl bias_121_07635_01_biwk1 blea 122 a7635 01 biskl bias_122_07635_01_biwk1 blea 411a763501b1sk1 bias_411_07635_01_biwk1 20u1 09/1316:50:29 1.5: 5O:2Wb1ea111a7635 01 th2 2001_09/13 bias_111_07635_01_wk2 2001 09/13 16: 50:25 2001_09/13 16:50:25 20u1 09/13 1.5: 50:26 2001_09/13 16:50:26 20u1 09/13 1.5: 50:27 2001_09/13 16:50:27 20u1 09/13 1.5:50:2W 2001_09/13 16:50:28 2001_09/13 16:50:30 20u109/131.5:5O:aU:

20u1 09/13 1.5: 50:31 2001_09/13 16:50:31 20u1 09/13 15: 5O:3Z 2001_09/13 16:50:32 2001_09/13 16:50:33 20U109/1315:5O:W3 20u1 09/13 15:50:34 2001_09/13 16:50:34 20u1 09/13 15: 50:34 2001_09/13 16:50:35 20u1 09/13 15: 50:35 2001_09/13 16:50:37

20u1 09/13 1.5: 50: 37 2001_09/13 16:50:38 20u1 09/13 1.5: 50:3W 2001_09/13 16:50:40

ati ati ati ati sti sti sti sti sti sti sti sti sti sti sti sti sti sti sti

daflk 111 a7635 01 th2 dark_111_07635_01_wk2

bias 111 a7635 01 k3 bias_111_07635_01_wk3 daflk 111 a7635 01 k3 dark_111_07635_01_wk3 bias 112 a7635 01 biakz bias_112_07635_01_biwk2 bias 121 0.7635 01 biakz bias_121_07635_01_biwk2 bias 122 a7635 01 biak2 bias_122_07635_01_biwk2

bias •411a7635o1biak2 bias_411_07635_01_biwk2 bias 111 a7635 01 &4 bias_111_07635_01_wk4

aak111a763so1 bias_111_07635_04_wk1 20u1 09/1316:50:41 1.5: SO:3Wbias111a763504 th1 2001_09/13 dark_111_07635_04_biwk1 2001_09/13 15:50:40 16:50:42 20U109/1315.50: 41 2001_09/13 16:50:42 42 2001_09/13 16:50:39 42 2001_09/13 16:50:43 2001_09/13 16:50:44 44 2001_09/13 16:50:45 45 2001_09/13 16:50:46 2001_09/13 16:50:46

ti

daki11a763S041 bias_121_07635_04_biwk1 bias 112 a7635 04 iak1 bias_112_07635_04_biwk1 bias 121 0.7635 04 biakl dark_111_07635_01_wk4 bias 122 a7635 04 biakl bias_122_07635_04_biwk1 •411a7635o43isk1 bias_411_07635_04_wk2 bias 111 a7635 04 k2 dark_111_07635_04_wk2 daflk 111 a7635 514 k2 bias_111_07635_04_wk3 bias_111_07635_04_wk3

tccccccc fl

RQ

RT

VE

RB

AV

c c c c c c c c c c c c c c c c c c .q c c c w ww ww w

c c c c c c c c c c c c c c c c c c c c c _ _

c c c c c c c c c c c c c c c p c c c c c _ _

c c c c c c c c c c c c c c c _ c c c n c _ _

c c c c c c c c c c c c c c c _ n n -. n w n _ _

WK UD

c c c c c c c c c c c c c c c _ c -.c c _ c _ _

c c c c c c c c c c c c c c c _ p c c _ c _ _

CA

c c c c c c c c c c c c c c c _ _ c — c _ c _ _

A

c c c c c c c c c c c c c c c _ _ _ — _ _ _ _ _

ccc ccccnc P— ccccnccc

ccccnccc ccccnccc — —

OSfs: c:19,r.245p:2total:266 OSFs: c: 19 w: 245 p: 2 total:266 00i24.12:38:03-D-MGRFrame Statusinfo: C: w: 19w: 2 total: 266 09/24.12:38:03-D-MGRFrame StatusInfo: c: 19 245246 p: 2p:total:266

Log Log

j

Figure 1. Observation Manager graphical display to control the stages of the Bias and Dark reference file production and list of the files being created.

Proceedure fil _ref_asns.py HST Science Mission Schedule (SMS) Delivery Newly scheduled exposures for HST

Identify planned exposures within the date boundaries for collection Verify mode of exposures Replace asociation members with parent.

HST Planing Database Tables

Perl and Python interfaces to Sybase databases

HST Archive

OPUS Time Poller (periodic trigger)

Procedure fil _anneals.p Identify CCD Anneal Proposals Form date boundaries for BIAS & DARK collecion Periods

HST Archive Catalog

STIS Dark & Bias Pipeline Database Tables

Identify uncollected reference file groups Check if needed exposures are archived YES - mark period collected and create Observation Status File (OSF) and start pipeline.

Figure 2. HST databases and systems that support the automatic retrieval of the data used by the STIS Reference File pipeline to create the Bias and Dark reference files.

Proc. of SPIE Vol. 7019 701931-4

fill_anneals

New SMS

fill_ref_asn

POLARC

Uncollected Periods

ARCREQ

ARCCHK

/work/dir/

VERBIA/VERDRK

REFBIA

multiple groups

REFAVG

BASJOI

WEEKBI

one group

Enough datasets?

no

yes multiple groups

UPD_RB/UPD_WD

REFAV2

CALSTIS

APPROVE

Figure 3. Diagram flow for the Bias and Dark reference files Pipeline.

Proc. of SPIE Vol. 7019 701931-5

Table 1. OMG Stages

Stage Name

ID

Action

ARCREQ

RQ

Submit archive request for on the fly reprocessing (OTFR) raw FITS files

ARCCHK

RT

VERBIA

VE

VERDRK

VE

This task polls the archive retrieval directory for the exposures required to construct the reference files Verify the mode of BIAS exposure files retrieved from the archive to make sure they apply for the type of reference file to be constructed. Imsets are separated into separate files and organized into subdirectories (groups of 29) for processing as above but for DARK exposure files

REFBIA

RB

REFAVG

AV

BASJOI

BJ

WEEKBIA

WK

WEEKDRK

WK

REFAV2

A2

UPD RB UPD RD

UD

This task combines bias exposures into a bias reference file. It runs Pyraf.REFBIAS task, which performs bias overscan correction and cosmic-ray rejection. It searches for hot columns and those are marked in the data quality array. Run Pyraf.REFAVER task. If there is more than one group, it combines each separate group of files into a single, averaged, bias reference file. Run Pyraf.REFAVER task. This routine produces a baseline bias file, a combination of all the bias files for a given anneal period. Run Pyraf.WEEKBIAS task. Required when insufficient number of imsets for high S/N criteria. Uses a monthly baseline bias, created with BASJOI, as input. Run Pyraf.WEEKDARK task. Uses the BIAS reference file as input to create a ”weekly” Dark. Combines separate bias file products produced in groups into a single, averaged, bias reference file. Updates header keyword and history comments

CALSTIS

CA

Runs calibration test using new reference files for verification

APPROVE

AP

Submit approved and verified reference file to CDBS for use in pipeline calibration

The first stage appearing in the OMG is ARCREQ, which submits the archive requests for the bias and dark raw FITS files.The list of requested exposures is written to a file in the local working directory so the next pipeline stage, ARCCHK, determines if the archive retrieval has completed successfully. To do this, the task searches the archive retrieval directory for all the {exposure} raw.fits files, where exposure is filled for every exposure listed in the local file created by the previous step. Once all the required files are delivered by the archive, these are moved to the working directory. The successful completion of the archive retrieval will trigger the VERBIA or VERDRK stages of the OMG. These stages verifie that all the retrieved files correspond to the mode in production. The files that don’t match the particular mode are moved out of this directory. Once a complete list of all the requested data is available, this process splits the set into groups of 29 or less exposures; a limit that is imposed by IRAF. A successful completion of VERBIA triggers the REFBIA stage; which runs the task Pyraf.REFBIAS. This task performs bias overscan correction and cosmic-ray rejection for all the available data. It also performs an statistical analysis of the data to mark the location of the hot-pixels in the data quality extension of the files. If the data had to be subdivided into more than one group, the successful completion of this stage will trigger the stage called REFAVG which takes all the groups and averages them to produce one single file. When there is only one group available, this stage appears as completed and the next step into the chain starts. If within a given ”week” period there are not enough frames to produce a high S/N bias, then the reference bias is created in the stage WKBIA, which runs the task Pyraf.WEEKBIAS. This task requires

Proc. of SPIE Vol. 7019 701931-6

—

=J2sJ

II1G: OI'US Process Maoaoec PMG: OPUS Proces Manager

repro

File Edit EditManage Manage View Tools Foe View Tools HewHelp

_______ ____________________________ _________________

stiref g2f repro stiref g2f repro quick quick

OPUS OPUS

Processes Processes

fl • OTFC

pid pid

CDB

OTFC

SAMPLE SAMPLE

•ALL

ALL

GIF O- OlE TEST •TEST Pipellines pipeiines

D stiref

____________

.:

00000716 0000312c 00000761 00003143 00000745 Yfl9245. 000531sf 000031a8 000007c2 Y22°7S. 0000072d 0000072d 0000316b 000006ef 00003175

I

process process polarc polarc polarc poiarc refavg refavg refavg refevg

reehia refbia reehia refbia reqsti reqsti splsti splsti

dddi splsti )ti updkey iipdJey updkey iipdJey

proc_stat proc_stat I

idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle idle

stat_time .start_fime 2001 2001 09/24 13:14:29 13:14:29 2001 2001 09/24 13:13:17 13:13:17 2001 2001 09/24 13:14:30 13:14:30 2001 2001 09/24 13:13:16 13:13:18 200109/24 09/24 13:14:31 13:14:31 .°± .°± 200109/24 09/2413:13:19 13:13:19 2001 09/24 13:14:31 .°± 2001 2001 09/24 13:14:31 13:14:31 doi : 2001 09/24 13:13:18 2001 2001 09/24 13:14:31 13:14:31 200109/24 2001 09/24 13:13:19 13:13:19

ñi

:•J:

path pathj

stiref stiref stiref stiref stiref stiref

stiref stiref stiref stiref stiref stiref

node node artichoke artichoke aardvark aardvark artichoke artichoke aardvark aardvark artichoke artichoke aardvark aardvark

proc_cmd p. roc_crnd I

stiref artc•g] artichoke stre stiref artichoke stiref

artichoke

Is streE aaai stiref aardvark

stiref stiref artichoke artichoke

stiref stiref

aardvark aardvark

11 11 ________ tiref: stiref:idle: idle:lItotal: 11 total:

09/24. 13:17:22-I-Monitoring Q924.13:1Z22-I-Monitoring stirefsiref

Loud

Figure 4. OPUS managers provide control of the stages of the Bias and Dark Reference Files Pipeline that are running.

a high-S/N ”baseline” bias file that is created in the BIASJOIN routine using the Pyraf.BASEBIAS task. In this case, the baseline bias is created using all the biases within an anneal period. Since the Dark reference files need a reference bias file, throughout this process the WKDRK stage had been waiting for the Bias reference file corresponding to the particular ”week” to become available for use. Therefore, once the reference bias has been successfully created by Pyraf.REFBIAS or Pyraf.WEEKBIAS, the stage WKDRK runs and creates the Dark reference file using the task Pyraf.WEEKDARK. When either the Bias or Dark reference files have been created, the next step in the pipeline starts execution. The stages, UPD RB or UPD WD, update the Bias or Dark reference file image header history and keywords, respectively, to reflect how these files were created. Testing of the files are performed by running the STIS calibration pipeline, Pyraf.CALSTIS, and an statistical analysis of the data for its validation. Figure 5a shows a STIS bias created using the task REFBIAS while Figure 5b shows a bias reference file created using the WEEKBIAS task. Although these two reference files correspond to consecutive ”weeks”, it is clear that the WEEKBIAS task results in a more noisy reference file. Once the Bias and Dark Pipeline successfully ends this last step with no error conditions throughout the stages, the reference files are ready for inspection and these are delivered to the Calibaration Database System (CDBS) for their use with science data. The last stage of the reference files, i.e. delivery of the files to CDBS, is still done manually. However, the STIS team plans to extend the Pipeline to execute this last step automatically. This would imply that the pipeline will fully create and deliver these files; thus making them available to the scientific community in a minimum period of time.

3.2. Future Development and Other Instrument Pipelines Following the experience gained in building the STIS Bias and Dark Pipeline, the Advanced Camera for Surveys (ACS) developed a BIAS and Dark Reference File Pipeline for their CCDs.12 In this case the pipeline was slightly extended to produce a variety of standardized statistics, which are used by the ACS team for long-term analysis. After the HST Servicing Mission #4 (SM4), during which STIS as well as the CCD channels of ACS are planned to be brought back to life, these pipelines will be revamped to further streamline the process to

Proc. of SPIE Vol. 7019 701931-7

1

4

Figure 5. The left panel shows a STIS bias file created using the task Pyraf.REFBIAS, while the right panel shows the product of the Pyraf.WEEKBIAS task. The noise level is lower in the former, while the hot pixels and columns are more contemporaneous with ongoing CCD observations in the latter. See discussion in Section 3.1.

create and deliver the BIAS and DARK reference files. The STIS team is planning to add new stages that would help to further assess the accuracy and correctness of these files and that will deliver these reference files directly to CDBS for their immediate use with STIS and ACS science data. Furthermore, similar pipelines will be developed to support the work to provide calibration reference files for The Wide Field Camera 3 (WFC3) to be installed in HST during SM4.

REFERENCES 1. P. Goudfrooij, and Diaz-Miller R.I., STIS Technical Instrument Report, STIS TIR 2000-06, 2000 2. J. Rose, R. Akella, S. Binegar, T. H. Choo, C. Heller-Boyer, T. Hester, P. Hyde, R. Perrine, M. A. Rose, K. Steuerman, ”The OPUS Pipeline: A Partially Object-Oriented Pipeline System”, in ASP Conf. Ser. ADASS IV, bf 77, R. A. Shaw, H. E. Payne, and J. J. E. Hayes, eds., (San Francisco: ASP), pp. 429, 1995 3. D. A. Swade,E. Hopkins, and M. Swam, in ASP Conf. Ser. ADASS X, 238, F. R. Harnden, Jr., F. A. Primini, and H. E. Payne, eds. (San Fransisco: ASP), pp. 295, 2001 4. P. Hodge, S. Baum, and P. Goudfrooij, STIS Technical Instrument Report, STIS TIR 1998-26, 1998 5. R. M. Katsanis, and M. A. McGrath, STIS Instrument Science Report, STIS ISR 1998-14,1998 6. J. Christensen, and P. Goudfrooij, STIS Technical Instrument Report, STIS TIR 1998-05, 1998 7. R.I. Diaz-Miller, and P. Goudfrooij, STIS Technical Instrument Report, STIS TIR 2000-05, 2000 8. R.I. Diaz-Miller, and P. Goudfrooij, STIS Technical Instrument Report, STIS TIR 2004-07, 2007 9. P. Goudfrooij, STIS Instrument Science Report, STIS ISR 99-08, 1999 10. J. J. E. Hayes, J. A. Christensen, and P. Goudfrooij, STIS Instrument Science Report, STIS ISR 98-06-Rev A, 1998 11. M. Swam, P. Goudfrooij, and R. I. Diaz-Miller, ”Automated ACS Bias and Dark Reference File Production, and Improved Detector Trend Analysis ”, in ASP Conf. Ser. ADASS XI, 281, D. A. Bohlender, D. D. Durand, and T. H. Handley, eds., (San Francisco: ASP), pp. 277, 2002 12. R. A. Lucas, M. Swam, M. Mutchler, and M. Sirianni, ”Automated ACS Bias and Dark Reference File Production, and Improved Detector Trend Analysis ”, inThe 2005 HST Calibration Workshop, A. M. Koekemoer, P. Goudfrooij, and L. L. Dressel, eds. , Space Telescope Science Institute, 2005, p.61

Proc. of SPIE Vol. 7019 701931-8