1. HST Spectroscopic Legacy Working Group Areas for Recommended Work in FY2014 and Beyond
Meeting 26 Jun 2013

2. Overview
Group has tried to identify main threads from the “Enhancing the Legacy of HST Spectroscopy” Workshop held at STScI on Nov 15-16, 2012
Membership of the group is as follows:
Alessandra Aloisi (chair) Jerry Kriss
Azalee Bostroem Cristina Oliveira
Ken Carpenter Rachel Osten
Stefano Casertano (co-chair) Lou Strolger
Scott Fleming Jeff Valenti
Andrew Fox Rick White
5 major areas of discussion were identified and explored:
Combining spectra
Spectral tools that are already available but people are not aware of
Extracting multi-object spectra automatically
Extracting information from spectra
Target information and classification
Identified the following tasks as priorities for the next couple of years (ordered from the highest to the lowest priority):
Make combined 1D spectral products available and create the pipeline to do it
Visualization tools for spectra
Characterization/classification of sources
Feature identification and measurement (develop equivalent of source extractor for spectra)
Confluence used for collecting and storing material, and for tracking discussion

3. 1) Make Combined 1D Spectral Products Available and create the Pipeline to do it (Part 1)
High priority, but will require ~ 2 years
Ultimate purpose of this activity is to combine 1D spectroscopic data together. In order to do so a spectroscopic pipeline is required.
Task should be done incrementally by increasing the difficulty/complexity of the 1D spectroscopic data that are handled by the combination process
Several pipelines already exist (developed by individuals or science teams) and can be used to facilitate design and requirements
Typically existing pipelines are tailored to specific science goals.
A more general design is required.
The writing of the pipeline should be done in steps:
The tool can start as an off-line tool manually triggered to create new high-level products for our own HST data into the HLA. This will allow us to get something quickly and with high visibility.
It could evolve into part of the OPUS/MAST.
Ultimately it could become a stand-alone tool offered to the community, similarly to an ASTRODRIZZLE for spectroscopy.

4. 1) Make Combined 1D Spectral Products Available and create the Pipeline to do it (Part 2)
For combination of 1D homogeneous data (HST similar instrument and resolution):
Phase 1 (6 months, FY 2014-Q1,Q2): design and requirements
0.3 FTE-year scientist
0.2 FTE-year RIA
0.5 FTE-year developer
Phase 2 (18 months, through FY 2015): development and testing
0.5 FTE-year scientist
1.0 FTE-year RIA
1.5 FTE-year developer
Phase 3 (partially overlapping with Phase 2, continuing effort): implementation, data processing and ingestion
1.5 FTE-year RIA/scientist (HLA FTEs) until the bulk of the data is ingested (start midway throughout second year of effort)
For combination of heterogeneous data, add 50% or more
could be partly done in parallel
For extended sources, the second dimension needs to be added (2D spectroscopy)
High-level products (and possibly pipeline) need to be advertised with the community

5. 2) Visualization Tools for Spectra
Medium-high priority, mostly based on adaptation of existing tools
We envision these tools as web-based and included in the Archive portal.
The tools will allow for a quick look and basic measurements of the spectroscopic data available in the archive
A lot of tools already exist spread throughout MAST, HLA, and VO
Existing tools are OK for a quick look and some basic measurements
Existing tools need improvements to become more user friendly and for more in depth analysis
Everything should be consolidated into the portal
Timeframe is driven by integration into the portal
Resources needed:
1.0 FTE-year scientist
0.5 FTE-year for a scientist responsible for the project
0.5 FTE-year for several other scientists that would provide requirements and test the tools
0.5 FTE-year developer
These tools need to be advertised with the community

6. 3) Characterization/Classification of Sources
Medium priority, relatively lower cost, may bring results in less than a year
applicable to all HST data (imaging and spectroscopy)
COS and STIS spectroscopy can be given higher priority if desired
Short-term approach based on literature analysis and dynamic search tool
literature analysis: associates datasets and publications
also includes PI-provided keywords
Effort ongoing, 1.0 FTE-year of an RIA (Archive) needed to complete analysis of past literature
Dynamic search tool to be integreted into the portal
based on VO lexicon
0.5 FTE-year from developer
0.5 FTE-year from science aware user(s) to test and refine interface
Future developments could include catalog-based classification and crowd sourcing
This capability within the portal needs to be advertised with the community

7. 4) Feature Identification and Measurement (Develop the Equivalent of Source Extractor for Spectra)
Medium priority, partly based on adaptation of existing tools
May require multiple options to handle different types of data (emission/absorption, bands, forests; high S/N vs. low S/N; continuum structure)
Task needs to be started after the process of classifying the spectroscopic sources has been completed
Useful tool may be obtained in ~ 1 year
Resources needed:
1.0 FTE-year scientist
0.5 FTE-year developer
This capability needs to be advertised with the community

8. Other Areas of Possible Interest
Multi-Object Spectral Extraction and Characterization (e.g., Grisms)
Not championed within working group
May be best led by instrument teams (e.g., WFC3 + JWST instrument teams)
Tools to visualize and analyze time dependence
Common framework yet to be established
HST data in meta-catalogs

9. Summary of Resources Year 1 Year 2
Scientist
(INS)
RIA
Developer (OED)
HLA staff (HMO)
Combined 1D spectra and pipeline
0.4
0.6
1.0
0.5
1.5
Visualization tools
Source classification
Feature identification and measurement
TOTALS
1.9
1.6
2.0
Resources amount to a total of ~ 6 FTEs per year spread between INS (COS/STIS team & RIAB), OED, and HLA over a couple of years to get the bulk of the high priority items advanced to a significant level.