Emission lines observed with Hinode/EIS

Transcription

1 c Astronomical Society of Japan. Emission lines observed with Hinode/EIS Peter Young STFC Rutherford Appleton Laboratory, Chilton, Didcot, Oxfordshire, OX11 0QX, U.K. Giulio Del Zanna University College London, Department of Space and Climate Physics, Holmbury St. Mary, Dorking, Surrey, UK and Helen Mason DAMTP, Centre for Mathematical Sciences, Cambridge, UK (Received ; accepted ) Abstract A recommended list of emission lines observed with the Hinode/EIS instrument are presented. In particular the best lines from a selection of important ions are identified, and notes on line blending given. This provides an important guide to future users of this instrument. There are several excellent density diagnostics in the EIS wavebands, and examples are given from the early data-sets. Key words: Sun: UV radiation 1. Introduction The EIS instrument is rich in emission lines, that provide many useful diagnostics to measure the plasma state. This paper provides a preliminary guide to the most prominent and useful emission lines. A full guide is deferred to future papers, since it would have to include a comprehensive discussion on in-flight calibration, line identifications, line blending for different sources, and of the current status of the atomic physics calculations for each line/ion in the spectra. We have done a lot of work in the past, but much more still needs to be done. Preflight descriptions of the EIS capabilities are given in Del Zanna and Mason (2005a) (where estimated count rates were provided for a selection of sources) and YOUNG (REFERENCE), and it is timely now to reassess the EIS capabilities. The comments we provide are mainly relevant to the use of the narrow slits. The atomic data we are referring to in this paper are available in the version 5 of the CHIANTI atomic package (Landi et al. 2006). 2. Instrument capability A key factor when judging the usefulness of an EIS emission line is the telescope effective area (EA) at that wavelength, which determines the fraction of incident photons that arrive at the detector. The EA curves for the two EIS channels (cf. (Del Zanna and Mason 2005a)) due to the presence of multi-layers, are strongly peaked (in particular the short wavelength (SW) channel, at 195 Å), which means that only spectral lines in the central range are prominent. For example, the Fexii λ line is the strongest line observed by EIS in most conditions. A weak line near 195 Å can yield more instrument counts (data numbers - DN) than a stronger line at the edges of the EA curve. Specific examples of this are discussed below. EIS studies should be designed with this in mind. A second, important factor when considering emission lines is the degree of blending. Even with the high spectral resolution of EIS, many lines turn out to be blended, particularly in the SW channel, crowded with transitions from Fe ions. Some are self-blends of transitions from the same ion, but others are with other species, The blending can be drastically different depending on the temperature (and density) of the source region. There are many cases, however, when contributions from other lines can be estimated quite accurately, by observing other transitions (e.g. that form branching ratios or are densityinsensitive). Important examples are the blending of two of the three EIS core lines, the Heii λ256.3 with Six and Fexii, and the Caxvii λ198.9 with many transitions from Ov and Fexi. A third important issue concerns the EIS nominal data rate of 50 kbits/s, which is not enough to allow the complete spectrum to be downloaded from every exposure while maintaining good spatial coverage and observation cadence. The onboard software thus allows only userselected portions of the spectrum to be downloaded by the choice of up to 25 windows on the detector. Typically these windows are chosen to pick out individual emission lines that the user determines will be optimum for the science they want to perform. The discussions in the following sections provides an overview of the key lines we believe should be included in EIS studies and the science they allow.

2 2 P.R. Young et al. [Vol., 3. Iron lines 3.1. Fe VIII All Feviii lines are found in the SW band, and the strongest lines in terms of DN are the and Å lines. The former is blended with the Nixvi Å line which becomes apparent in active region intensity maps with a mist of emission visible around the hot core of the active region. It is possible, however, to estimate the Nixvi contribution by observing the Nixvi Å line. Feviii is typically found to be strongest in the footpoints of loops at the edges of active regions, where Nixvi is often negligible. The line is a weaker Fe viii transition but is not blended. Alternatively, the Å line is much stronger, however is significantly blended with at least a Caxiv Å line. This Caxiv can be deblended by observing the λ line Fe X The identifications, atomic calculations and diagnostics of this ion are discussed in detail in Del Zanna et al. (2004). We believe that the atomic data calculated in Del Zanna et al. (2004) are reliable. The strongest Fe X lines observed by EIS are at , and Å (the latter actually a self-blend of two lines). λ and λ are comparable in strength in terms of DN. The λ line is partly blended with an unknown line at around Å. The temperature of formation of this blending line is around log T = based on comparisons of images formed in this line with other EIS lines. The / ratio is insensitive to density, but the line is sensitive to density when taken relative to either of these lines (Del Zanna et al. 2004). The ratios are sensitive in the range cm 3. The λ line is in a crowded part of the spectrum and a large wavelength window (at least 50 pixels) is required to obtain a good estimate of the spectrum background. The λ feature is of interest because one of the two blending lines is a rare example of a forbidden line of significant strength in the EUV, and it actually dominates the blend below a density of cm 3. The Fe x λ175.27/λ ratio is an excellent density diagnostic, but unfortunately the low EA in this wavelength region means the ratio is not useful Fe XI Fexi lines are among the brightest ones in EIS spectra, whenever 1 MK coronal plasma is present. Fe xi transitions provide some density diagnostics, and are blending many other spectral lines. Unfortunately, most Fe xi energy levels are uncertain, and strong level mixing is present for some important ones, which produce some of the strongest lines in the EIS wavelengths. There are currently no reliable atomic physics calculations for this ion, and most identifications found in the literature are incorrect. We have been working in the last two years on this ion (in collaboration with P.Storey), but until this work is finished, identifications and atomic data for this ion should be considered unreliable. We believe that the 3s 2 3p 3 ( 2 D) 3d 3 P 2 level decays to the ground configuration 3s 2 3p 4 3 P 2,1 levels to form the Å Å lines. The latter is blended with another Fexi transition, with two Ov lines, and, when temperatures are high enough, with Caxvii. We also believe that the strong Å is mostly due to Fexi. What we know for sure is that the Å line is due to the decay to the ground state from the 3 D 3 (plus a blending of Fex), and that the Å line is due to the decay of the 3 D 2 to the first excited state, 3 P 1. The ratio of the , Å lines is a good density diagnostic in the range Log N= Fe XII Fe xii atomic data, identifications and blending have been discussed in detail in Del Zanna and Mason (2005b). We believe that, after more than 30 years of discrepancies between observations and theory, we can now reliably use Fe xii lines, thanks to the the scattering calculations of Storey et al. (2005). The three strongest Fexii lines are the decays of the 3s 2 3p 2 ( 3 P) 3d 4 P 5/2,3/2,1/2 to the ground state, observed at λ195.12, , respectively. The Fe xii λ line lies at the peak of the EIS sensitivity curve and so, in terms of counts measured at the detector, is the strongest emission line observed by EIS in most conditions. It is one of the EIS core lines, and so every EIS observing study includes this line by default. This line has been found to be broader than the λ line. This could be an indirect confirmation of the identification proposed in Del Zanna and Mason (2005b), i.e. that the λ line is actually a self-blend of two transitions (resolved in laboratory spectra). The only problem with observing the λ195 line is that it is likely to saturate on the detector if long exposure times are used. In active conditions, even a 30 second exposure can lead to saturation in bright parts of an active region. We thus recommend that the λ or λ lines are observed in addition to λ195 as they are around 27 % and 60 % weaker in terms of DN. Their ratio is insensitive to density. Fexii provides some of the best density diagnostics for EIS. The ratio of the or Å lines to any of the above is sensitive to a wide range of densities (Log N = 8 12). The Å line, according to Del Zanna and Mason (2005b), is a self-blend, but also blended with a weaker S x transition, which has a branching ratio with the Sx Å. This latter line should be included in any EIS study that contains the Å line Fe XIII The atomic data for this ion are discussed in Young (2004). The three most important Fe xiii lines are at , and Å which between them form the best coronal density diagnostics available to EIS. The λ line is actually a blend of two Fexiii lines at and Å, that are approximately in the ratio 1:3. Interpretation of the λ line is hampered by a blend with Fexii λ203.72, but fitting the combined fea-

3 No. ] Emission lines observed with EIS 3 ture with two Gaussians can usually separate the Fe xii and Fe xiii components. The λ line is unblended, while λ is just resolvable from the Fexii λ Other strong Fe xiii lines are found at Å, Å (blended with an unidentified line) and Å, but are not as useful as the aforementioned lines Fe XIV Atomic data and comparisons with observations for this ion are discussed in detail in Storey et al. (2000). A number of prominent Fe xiv lines are found in the EIS wavebands, and the one recommended here is at Å. Although there is a blend with Sivii λ274.18, this can be quantified if the Sivii λ line (one of the recommended lines) is also observed as the λ274.18/λ ratio is at most In most active region conditions the blend can safely be ignored. Observing the λ line yields a good density diagnostic (Log N=8.5 11) relative to λ and is recommended for probing hotter parts of active regions. Another strong line is λ270.52, but is weaker than λ in all conditions, and the λ264.78/λ ratio is less sensitive to density. The λ line is the strongest of all Fexiv lines, but is found right at the edge of the SW channel and so has a very low EA and is thus not recommended Fe XV The Fexv Å line is the strongest line from the ion and dominates this part of the spectrum in active conditions. We recommend its inclusion in any EIS study. In quiet Sun conditions the line is very weak or non-existent and an Alix line at Å becomes apparent Fe XVI Three lines are found in the EIS long wavelength band at , and Å. They are all temperature and density insensitive relative to each other, with the λ p 2 P 3/2-3d 2 D 5/2 line the strongest. This latter line is also unblended and is strongly recommended for inclusion in all observations. Note that the λ is much weaker than the main resonance lines of Fexvi, being only 7 % of the strength of λ Fe XVII Fe XXIII Atomic data and identifications for this ion are discussed in detail in Del Zanna et al. (2005). The intercombination 2s 2 1 S 0-2s 2p 3 P 1 λ line should become the strongest line during flares, after the Fe xxiv lines (Del Zanna and Mason 2005a). This line should be unblended, and EIS studies of active regions should include this line Fe XXIV Atomic data and identifications for this ion are discussed in detail in Del Zanna (2005). The doublet 1s 2 2s 2 S 1/2-1s 2 2p 2 P 3/2,1/2 falls at and Å respectively. During flares, these lines become the most prominent lines in the EIS spectra (Del Zanna and Mason 2005a). The Å line should be blended with a relatively weak Sx Å, while the Å line is blended with yet unidentified lines that are present even in quiet conditions. 4. Oxygen lines 4.1. OV The strong 2s 2p transitions of Ov are all at longer UV wavelengths, but there are a number of weaker n = 2 to n = 3 transitions in the EIS bands. The 2p 1 P 1 3s 1 S 0 transition is at Å and is relatively strong. It is important to include this line in any observation because it is necessary to deblend the contribution from two Ov 2p 3 P J 3d 3 D J lines to the Fexi Caxvii Ov feature at Å. See also Young et al. (2007). CHIANTI lists a blend with Al viii λ This Alviii line should only be a minor component, however to estimate its contribution either the or Å lines (which form a branching ratio with the λ248.46) should be observed. The stronger of the 2p 3 P J 3d 3 D J transitions is at Å and can be isolated from the Fexi Caxvii Ov feature at Å. It is an excellent diagnostic of transition region plasma when observed in conjunction with the Å line OVI Two useful O vi lines are available at and Å and the ratio λ183.94/λ is 0.5. They lie close to the recommended line Fex λ and a single broad wavelength window can be used to pick up all three lines. 5. Magnesium lines 5.1. Mg V Only one significant line is found in the EIS wavelength bands: the 2s 2 2p 4 1 D 2 2s2p 5 1 P 1 transition at Å. While very weak in most circumstances, the line is significantly enhanced in loop footpoints (Young et al. 2007) and gives valuable temperature information Mg VI The 2s 2 2p 3 2 D J 2s2p 4 2 D J transitions are at and The latter lies in the wing of Fe XIV which is usually much stronger in active region conditions. The line is unblended. Both lines can be strongly enhanced in loop footpoints as demonstrated in Young et al. (2007, this issue). Note that the λ line is relatively isolated and thus valuable for isolating loop footpoints in 40 slot data Mg VII The λ280.78/λ diagnostic, suggested by Del Zanna and Mason (2005a), is discussed by Young et al. (2007) where it is confirmed to be an excellent diagnostic for coronal loop footpoints.

4 4 P.R. Young et al. [Vol., The λ is blended with Sivii λ278.44, but the two components can be separated using a two Gaussian fit, or by making use of the Sivii λ line which is density insensitive relative to the λ Silicon lines 6.1. Si VII The 2s 2 2p 4 3 P J - 2s2p 5 3 P J transitions are found in the EIS long wavelength band, and the strongest is the line which is unblended. Note that, from visual inspection, Si VII is formed at around the same temperature as Fe VIII but is weaker by a factor two than the Å (blended) line of that ion Si X The six lines belonging to the 2s 2 2p 2 P J 2s2p 2 2 P J and 2s 2 2p 2 P J 2s2p 2 2 S 1/2 transitions are found in the LW channel, with the strongest being λ and λ which form a density diagnostic. Both lines appear to be unblended (the λ272 could be used instead of the λ261.04). The λ forms a branching ratio with the λ256.37, which blends the He ii λ It is therefore recommended that EIS studies include this line. According to the Mazzotta et al. (1998) ion balance calculations Si x is formed at almost exactly the same temperature as Fexii, and so the Six should provide the same results as Fe xii. The Fe xii λ has approximately the same DN as the Si x λ258.37, however the λ is much stronger, so the Fexii ratio is to be preferred. 7. Other ions 7.1. He II While Heii λ is the coolest line observed by EIS, and also the strongest line formed below 10 6 K, appears to be affected by blending at least with Six, Fexii (cf. Del Zanna and Mason (2005b)), and Fexiii SXIII The resonance 2s 2 1 S 0-2s 2p 1 P 1 λ is almost as strong as the Fexvi λ in active regions Ni XVII The 3s 2 1 S 0-3s 3p 1 P 1 λ line is the analogous transition to the λ of the iso-electronic Fe xv ion. The line is unblended and a valuable probe of the hot cores of active regions. It is weaker than the Fexvi λ and Sxiii λ256.68, but according to the Mazzotta et al. (1998) ion balance calculations this line should be formed by slightly hotter plasma, in the log T = range Ca XVII During flares, the resonance 2s 2 1 S 0-2s 2p 1 P 1 transition becomes the dominant contribution to this complex blend, formed by at least four other strong Fexi and Ov lines. Table 1. EIS emission lines recommended to be included in observation studies. A b indicates blending (see text). Ion Wavelength / Å log T max Cool lines (log T max < 6.0) He ii b 4.7 O v b 5.4 O vi Mg v Mg vi Si vii Fe viii b Coronal lines (6.0 log T max 6.4) Fe x Fe xi Si x Fe xii b 6.1 Fe xiii Fe xiv b 6.3 Hot lines (log T max = ) S xiii Fe xvi Ni xvii Flare lines (log T max > 6.6) Ca xvii b 6.7 Fe xxiii Fe xxiv b 7.2 Fe xxiv b 7.2 Table 2. A selection of good density-sensitive line ratios available with EIS. The ion, the line ratios, and the approximate range (log N e, cm 3 ) of sensitivity are indicated. Ion Wavelength / Å log N e Mg VII 280.7/ Fe X 257.2/ Fe XI / Si X 258.4/(261.0,272.0) 7-10 Fe XII (186.88,196.6)/(192.3,193.5,195.1) 7-12 Fe XIII (196.5, 200.0, 203.8)/ Fe XIV (257.4,274.2,270.5)/(264.7,252.2) Further considerations for designing EIS observing studies We strongly recommend that a good selection of lines spanning a wide temperature range is included in most

5 No. ] Emission lines observed with EIS 5 EIS studies. We note here that the EIS planning software allows spectral windows to have variable sizes, and that many important lines are nearby in wavelength, hence data windows should be designed to include them. Window widths should be at least pixels wide. If high cadence is required, then using 24 pixels to reduce the data rate is acceptable, but high velocity events might be missed out. Given the highly peaked EA, strong lines may become saturated, hence it is recommended to have multiple exposures with different durations. The number of lines, exposures, slits, should be carefully chosen, based on the target regions. 9. Summary The EIS instrument is performing, in terms of sensitivity and spectral resolution,very well and as expected. The observed spectra are remarkably similar to those predicted by Del Zanna and Mason 2005a. Here, we have only provided a short summary of some of the most important spectral lines, in terms of their strength and for measurement of electron densities. We confirm the suggestions of Del Zanna and Mason 2005a on which are the most prominent lines in the EIS spectra, but also pointed out that, despite the excellent spectral resolution, many lines are still blended. Table 2 summarises the best density diagnostic ratios. The primary coronal density diagnostics are Fexii λ186.88/λ195.12, and Fexiii λ196.54/λ and λ203.83/λ References Del Zanna, G., Berrington, K.A., & Mason, H.E A&A, 422, 731 Del Zanna, G., Mason, H. E., 2005a Adv. Space Res., 36, 1503 Del Zanna, G., Chidichimo, M. C., & Mason, H. E A&A, 432, 1137 Del Zanna, G., Mason, H. E., 2005b A&A, 433, 731. Del Zanna, G., 2005 A&A, 447, 761 Landi, E., Del Zanna, G., Young, P. R., Dere, K. P., Mason, H. E., Landini, M. 2006, ApJS, 162, 261 Mazzotta, P., Mazzitelli, G., Colafrancesco, S., & Vittorio, N. 1998, A&AS, 133, 403 Storey, P. J.,Mason, H. E., Young, P.R. 2000, A&AS, 141, 285 Storey, P. J., Del Zanna, G., Mason, H. E., Zeippen,C., A&A, 433, 717 Young, P.R. 2004, A&A, 417, 785 Young, P.R., Del Zanna, G., Mason, H.E., et al. 2007, this issue Fig. 1. A sample quiet Sun spectrum of the SW channel.

6 6 P.R. Young et al. [Vol., Fig. 2. A sample quiet Sun spectrum of the LW channel.