To: EUVE Science Advisory Board
From: Nancy S. Brickhouse and Jeremy J. Drake, AXAF Science Center
Date: April 24, 1999
Subject: Joint AXAF/EUVE Calibration and the Emission Line Project
The EUVE Science Advisory Board has made a commitment to support the AXAF calibration (NRA Appendix A). We propose here a specific plan for the EUVE support of cross-calibration. The proposed EUVE exposure times are within the allotted EUVE calibration time. We ask that a decision on this proposal be made before the Guest Observer proposal deadline.
Goals for AXAF in-flight grating calibration include effective area with HZ 43 and the line response function and Emission Line Project (ELP) with three emission line sources. Table 1 lists the proposed cross-calibration targets for AXAF and EUVE. The routine EUVE observation (which cannot be made simultaneously) of HZ 43 (non-variable) supports the AXAF LETG effective area calibration. We propose simultaneous observations for Procyon, Capella, and HR 1099 (variable sources) for both direct spectral overlap between EUVE and LETG, and for the ELP involving both LETG and HETG.
While the three stellar targets will be used to calibrate the line-response function across the waveband of each AXAF grating, the AXAF exposure times (Table 1) are supplemented by the needs of the ELP, with the goal of benchmarking critical atomic data used as tools for interpreting spectra from a diverse range of astrophysical sources. The atomic data are inputs to plasma emission models, widely used for diagnosing the physical conditions of the source: temperature, emission measure, density, ionization state, abundances. The ELP will produce a catalog of line identification and intensity measurements from EUVE and AXAF for these sources which will serve as a spectroscopy reference guide and as feedback on the reliability of atomic data.
The three ELP targets have been chosen to span the broadest range of ions producing line emission. With the three selected targets of different coronal temperatures, critical ions will provide X-ray emission lines with sufficient counts (for the listed exposure times) to test the atomic models at the level of their expected accuracy.
The consensus that has developed in the AXAF community for ELP is based on the understanding that in order to make plasma emission models useful for X-ray spectroscopy at both modest and high resolution, they must be tested and improved. Having put tremendous effort into AXAF calibration, there needs to be an equally serious effort to ensure that modeling tools of general use are optimal for spectral interpretation. The AXAF ELP grating spectra will themselves be unsurpassed in the combination of excellent bandpass, calibration, spectral resolution, and signal-to-noise (S/N).
Recognizing that testing analysis tools (in this case, atomic rates) is not a traditional calibration objective, the AXAF community has nevertheless found the scientific objectives sound, the time investment commensurate with those goals, and the need great. The ELP has received broad endorsement from astrophysicists, the AXAF Science Center and NASA Project Management, plasma code developers, and theorists and experimentalists in the atomic physics community.
What is the role for EUVE in the Emission Line Project? By extending the wavelength range for emission line observations into the EUV, the Emission Line Project will produce a catalog that gives accurate total line cooling on an ion-by-ion basis, line intensity ratios for lines of different energies from the same ion, and connections to theoretical predictions. It is not generally appreciated that the accuracy of theoretical predictions for line intensities for many strong EUV lines far exceeds the accuracy of intensity predictions for their X-ray counterparts. As an example, the strong Fe XVIII lines observed in the SW spectrometers are believed to have collision strengths accurate to about 10%. These same ions produce lines around 1 keV, which are not accurately predicted to better than 30%, and may in fact be far worse than that.
The computed EUVE exposure times in Table 1 are based on the benchmarking goal that the most accurate EUV emission lines (primarily Fe IX, XIV, XV, XVI, XVIII, XXI, XXIII, and XXIV) have sufficient counts to at least ``match'' the theoretical expectations (S/N > 10). Times are scaled from existing EUVE observations. The EUVE exposure times provide good coverage for simultaneity with AXAF.
The use of the EUVE archives by high energy astrophysicists will be greatly enhanced if predicted line intensity ratios between EUV and X-ray lines are confirmed by observation. For example, column densities could be determined by comparing predicted to observed line ratios for well separated lines (e.g. the Fe XVIII line ratios mentioned above).
Both AXAF and EUVE calibration products are to be made immediately public, and thus the ELP data will receive widespread scrutiny and use. We expect these data sets to provide guidance to the astrophysics community on the appropriate use of plasma emission models, as well as much needed feedback to the atomic physics community; here again, the EUV line intensities are critical in that they are intrinsically more accurate. In supporting the Emission Line Project, EUVE can add to its legacy a major contribution to improving our understanding of high energy spectroscopy, via a line intensity catalog for major ion species emitting throughout the EUV to X-ray band.
(1.) How did the AXAF community make the decision to conduct the Emission Line Project? The idea arose from work being done to understand discrepancies between EUVE and ASCA data. Recent results (e.g. Liedahl & Brickhouse) show that fits to ASCA data may fail because of inadequacies of even the newest plasma emission codes. Brickhouse & Drake brought these results to the leadership of the AXAF Science Center (H. Tananbaum), with the idea that long grating exposures would allow great progress to be made in this area. The idea for the project was discussed with the AXAF Users' Committee, the Science Working Group, and the Project Manager (M. Weisskopf). Key astrophysicists supporting ELP include A. Fabian, J. Linsky, R. Mushotzky, C. Canizares, B. Brinkman. Plasma modelers such as J. Raymond, R. Mewe, J. Kaastra, and K. Dere are supportive, as are atomic physicists such as H. Mason and F. Keenan. D. Liedahl, P. Beiersdorfer, and W. Goldstein have pledged their support of the project as a complement to the LLNL joint theoretical and experimental studies of atomic processes in plasmas.
(2.) How were the AXAF exposure times calculated? The atomic benchmarking goals for the LETG were to provide a survey of lines, line intensity ratios, and tests of the ionization balance models. The LETG band includes L-shell emission from Ne, Mg, Si, S, Ar, and Ca in addition to the overlap with the SW Fe lines. By comparing the shapes of emission measure distributions derived from different elements, problems with the ionization balance may appear. Since the dielectronic recombination rate coefficients are particularly uncertain, this test is critical. Thus the critical exposure times were based on providing coverage of sufficient numbers of overlapping emission measure curves based on lines with S/N > 10.
The HETG benchmarking goals are to determine the overall fluxes of bands of weak lines which affect global fitting of non-grating data and to provide 1#1 to 2#2 line ratio measurements of Fe L-shell ions. These line ratios also affect fits to moderate resolution X-ray data. Thus the critical exposure times were based on obtaining sufficient counts in the weaker lines ( 1#1 lines) from most of the Fe ions.
(3.) Why not use solar observations for benchmarks? Aren't they photon rich? Solar observations ARE used for benchmarks; however, they are generally limited in their usefulness by narrowness of spectral range, spectral resolution (in some cases), poor calibration and other instrumental effects (the need for scanning across the waveband), time-dependent effects (flares being the only source of high temperature lines). Recent EBIT measurements of Fe XXIV disagree with measured solar line ratios by about a factor of two.
Furthermore, the solar community has had the luxury of working in the EUV spectral range, where the atomic rates are generally better known. SOHO does not contain any X-ray spectroscopic instruments.
(4.) Aren't laboratory experiments better for testing data? Yes, in most cases, but laboratory experiments are expensive and limited to only a few transitions, whereas we need benchmarks for hundreds of lines (themselves the culmination of thousands of atomic processes). Furthermore, laboratory measurements are done under carefully controlled conditions (such as at a single electron beam energy), and thus cannot reproduce the plasma conditions in astrophysics (such as a Maxwellian plasma). Finally, laboratory measurements will only be useful for identifying line blending when line identifications of all elements are completed, a project expected to take many years.
(5.) How can we be sure that effects seen are atomic rather than astrophysical? We have chosen sources that have been well studied with EUVE as well as with X-ray missions such as ASCA. There is no evidence to suggest that these sources are other than ideal coronal equilibrium conditions with negligible optical depth. The data themselves will either continue to confirm the basic models or will disprove them. EUVE measurements will provide critical tests, since the comparison of lines formed at different energies is a good test for non-equilibrium ionization. If we find a consistent pattern of discrepant line ratios, we will know that we have to look again at our simplifying assumptions; however, individual lines that are discrepant would indicate atomic physics (or line blending) problems that can be given a high priority for laboratory tests.
(6.) Shouldn't this be a Guest Observer proposal, since it is not instrument-based? For AXAF the Emission Line Project is part of the calibration plan. The LETG observations do provide directly overlapping spectral coverage with EUVE, so that piece of the EUVE participation is an instrument calibration project.
The point at issue is the coordination of EUVE and the AXAF HETG. The goal of the ELP project is to produce atomic data benchmarks, not astrophysical results, and thus, is not a traditional Guest Observer proposal either. An EUVE GO proposal would need to be judged, not in the category of Late Type Stars, but either by a separate Time Allocation Committee panel, or as part of many panels representing different types of emission line sources. The benefits of the Emission Line Project cross interdiscplinary lines, including supernova remnatns, galaxies, clusters of galaxies, AGN, the interstellar medium, as well as late-type stars. Furthermore, the data products are to be public, like other calibration products. Finally, like calibration data, the ELP will provide a service to the scientific community at large.
