EIS Science Notes

These notes emerged initially from the Solar-B EIS (UK) meeting held at MSSL on 2-3 June 1998 (see minutes and science discussion - other formats available see Meetings), in response to Action 4: Wavelength Ranges, and Action 5: EIS Applications / studies . Since then an additional wavelength range has been considered (range 6) around 400 A. The purpose of this document at present is to aid the decision on wavelength range(s) for EIS. This document will develop as the contributions are revised from time to time. More science topics will be added in th near future and more detail in the existing topics once the major decisions on design have been made. Please submit further contributions to me, Matthew Whyndham, at mwt@mssl.ucl.ac.uk and Len Culhane at jlc@mssl.ucl.ac.uk. Contents Wavelength Ranges 170 - 210 Å 240 - 290 Å (baseline range) 308 - 381 Å 517 - 633 Å 1334 - 1428 Å 385 - 415 Å Conclusion EIS Applications / Studies EIS Study Sheet Also available as a separate html file, and in other formats. ->4 August 1998 The CME onset eludes our detailed understanding. We have many models but lack detailed observation of the CME source region at the time of a CME onset. This is partly due to the fact that coronagraphs actually occult the onsets of the CMEs they observe! In addition, since coronagraphs (i.e. CME observations) are tuned to the plane of the sky, the source regions are near the solar limb. Such regions are very difficult to observe because of foreshortening and occultation by the limb itself. Even if the source regions are observed at the time of a CME onset, to obtain detailed diagnostic information of the CME onset process one requires plasma diagnostic information (density, temperature, velocity) on spatial and temporal scales of about a few arcsec and less than a few minutes, over a region of at least several arcminutes by several arcminutes. A good range of temperatures must be observed because the eruption processes include both chromspheric and coronal plasma. Supporting coronagraph (to identify CME) and ground-based (to see any prominence/filament eruption) would be essential. The bottom line is that a spectrometer in the EUV is required. This is an activity being performed using the CDS instrument on SOHO. Whilst much headway may be made using the CDS observations, the fine-scale flow patterns witnessed in prominences and the desire to image over large areas with cadances of minutes are beyond the capability of CDS. Thus, we anticipate great rewards from a close investigation of CME onsets using EIS and the Solar-B payload. Study Details:

         Raster Area:            4x4 arcmin      (* as large an area as is practical)
        Raster Step:            2 arcsec
        Raster Locations:       120
        Exposure Time:          1 sec.          (*Assuming a CDS count of 10 per sec
                                                for the weakest line and an EIS
                                                sensitivity 10x CDS and a desire
                                                for 10% counting stats).
        Duration of Raster:     1x120 plus 10% overhead? = 2.2 min.
                                                (*Overhead for CCD readout time etc...)
        Number of Rasters:      Open, but ideally should monitor region
                                for many hours each day as it aproaches limb.

        Line Selection:         Range of temperatures using bright lines to
                                keep raster repeat times low. Ensure they are
                                well separated to reduce blends and enable
                                good velocity studies. Also, some
                                density capability. e.g. He II 243.03 (256 is
                                blended), Mg VII 278.41, 280.74, Si VII 275.38,
                                Fe IX 244.92, Fe XIII 251.94, Fe XIV 264.78,
                                274.20,  Fe XV 284.16, Fe XVI 262.98 (10 lines).

        Bins Across Line:       25              (* To cover about +/- 250 km/s).

        Telemetry/Compression:  10 lines x 25 bins x 120 bins x 12 bits(?)
                                = 360,000 bits per exposure. At 64 kb/s would take
                                5.63s. Require compression factor of about 5.
                                                (*Assumes we take the slit only,
                                                not the wider parts of the dumbell).

        Solar Feature Tracking: Not required (near to limb)

        Supporting Observations:
                                The EUV observations cannot identify a CME so we
                                require coronagraph support. In addition, the events
                                are undoubtedlt magnetic in nature so detailed magnetic
                                mapping would be ideal. Also, the EUV spectroscopy
                                can only monitor a relatively small region; a larger
                                area coronal mapper/imager would be essential. So the
                                supporting devices would be:

                                1. Coronagraph (to identify CME!) - this could be
                                        the MkIII K-coronameter or event LASCO (if SOHO
                                        is rescued). Needs to monitor the region of the
                                        corona above the EIS region with cadence of
                                        10 min or better.

                                2. Magnetic Mapping of source region - this can be 
                                        from the Solar-B instrument. The EIS region,
                                        and a larger surrounding area if possible,
                                        should be mapped throughout, with a cadence
                                        of under 10 minutes (much less if possible).

                                3. Context Mapping of Coronal Structures - this would
                                        be from Solar-B. Larger area maps of the corona
                                        are required to assess the large area and remote
                                        activity and structure.


        What is New?
                                EIS gives the potential for reasonable area mapping
                                with spectroscopic information down to less than
                                a few minutes. Similar activities using CDS take
                                16 minutes. For eruptive activity, the improvement
                                is very significant. In addition, improved spectral
                                resolution, over CDS, enables a much better mapping of
                                the flows of the EUV plasmas - particularly as
                                prominences are activated. Thus, we can make significant
                                new steps.

Notes: - The basic driver is a large area, quick raster in a number of emission lines over a good range of temperatures with some density capability. - The line selection contains a transition region and coronal density diagnostic, but the weaker Mg VII line is rather weak. However, summing successive rasters to obtain reasonable counts in the weakest lines may be a useful approach. - The observation assumes that we can select data from the slit only and not the slots. Dynamic Events in the Network Richard Harrison (RAL) Coronal heating and solar wind acceleration are ultimately powered by the kinetic energy of the convection layers. Magnetic fields carried in the convection cells migrate to the cell boundaries (the network) and the combination of flux concentrations and newly merging flux in these regions probably drives magnetic transient events which provide the acceleration and heating processes. If this picture is true, the resulting, globally distributed 'disease' of mini-exposive events may hold the key to coronal heating and solar wind acceleration mechanisms. SOHO has been used to observe transient events in the network which fit this picture, i.e. the CDS blinkers and the SUMER explosive events. The former are EUV flashes especially in the few hundred thousand K region lasting typically 10 minutes, with about 3000 distributed over the disc at eny time. No significant velocities are associated with these events. However, the explosive events are UV velocity events seen in the network, maybe 30,000 on the Sun at any time, with speeds of up to 150 km/s. The two classes of network transient events maybe related. In addition to this relationship, we need to invesigate further the effects of these events in the corona and the related magnetic activity in the regions below. The superior EUV spectral resolution and temporal resolution (afforded by the better sensitivity) would allow EIS to take sigificant steps beyond the initial studies made using CDS. The basic requirement is temporal resolutions of under a minute over rastered areas of several tens of arcsec (larger than a cell), with good temperature coverage. Study Details:

        Raster Area:            1x1 arcmin
        Raster Step:            2 arcsec
        Raster Locations:       30
        Exposure Time:          1 sec.          (*Assuming a CDS count of 10 per sec
                                                for the weakest line and an EIS
                                                sensitivity 10x CDS and a desire
                                                for 10% counting stats).
        Duration of Raster:     1x30 plus 10% overhead = 33 sec.
                                                (*Overhead = CCD readout time etc...)
        Number of Rasters:      Minimum 500 (= 255 min)

        Line Selection:         Range of temperatures using bright lines to
                                keep raster repeat times low. Ensure they are
                                well separated to reduce blends and enable
                                good velocity studies. Also, some
                                density capability. e.g. He II 243.03 (256 is
                                blended), Mg VII 278.41, 280.74,
                                Fe IX 244.92, Fe XIII 251.94, Fe XIV 264.78,
                                274.20, Fe XVI 262.98 (8 lines).

        Bins Across Line:       25              (* To cover about +/- 250 km/s).

        Telemetry/Compression:  8 lines x 25 bins x 120 bins x 12 bits(?)
                                = 288,000 bits per exposure. At 64 kb/s would take
                                4.5s. Require compression factor of 4.5.
                                                (*Assumes we take the slit only,
                                                not the wider parts of the dumbell).

        Solar Feature Tracking: YES     (* Must be able to make steps of less
                                        than pixel size to avoid jumpy movies).

        Supporting Observations:
                                1. Magnetic Mapping of source region - This would
                                        be from the Solar-B magnetic imager.
                                        Observations of the same region with a similar
                                        cadence and similar or better spatial resolution
                                        would be ideal, in order to map the magnetic
                                        changes to the transient EUV activity.

                                2. Context Mapping of Coronal Structures - The EUV
                                        images would be relatively small, in order
                                        to have reasonable temporal resolution. Larger
                                        area coronal maps would be useful to assess the
                                        local and remote magnetic structure and activity.

        What is New?
                                This problem has been tackled using a combination of
                                CDS and MDI data from SOHO - with some success. However,
                                to perform the same kind of operation with significantly
                                better temporal resolution and spectral resolution
                                would be essential. In particular, the association of
                                EUV blinkers and UV high velocity events must be sorted
                                out.

Notes: - The basic driver is a quick raster on quiet Sun in a number of emission lines over a good range of temperatures with some density capability. - The line selection contains a transition region and coronal density diagnostic, but the weaker Mg VII line is rather weak. However, summing successive rasters to obtain reasonable counts in the weakest lines may be a useful approach. - The observation assumes that we can select data from the slit only and not the slots. Questions Relating to EIS Performance from these Studies - For these two Studies I would want to use the slit, not the slots. Thus, rather than waste telemetry, I would select on the pixels covered by the slit data. I assume that this is no problem. - I have assume a 'final' full instrument sensitivity something like 10x CDS. CDS has the two reflections off the telescope, the scan mirror and the grating, as well as the detector efficiency to worry about. EIS has the off-axis mirror, the grating and detector. I think I remember quotes of about 5-20x better than CDS. The count rates and, thus, exposure times will depend on this. - I have assumed that solar feature tracking is an option, using the mirror mechanism. It would be nice if such a system had step sizes less than the pixel sizes to avoid jumpy movies. For the CDS smaller rasters this is particularly obvious. I guess pointing should be automatically updated between rasters. - The two studies need a compression of order 5 to avoid the telemetry driving the raster frequency. - I have assumed a 2 arcsec slit and 2 arcsec steps in the rasters. - I have assumed a 21 mA pixel size in the wavelength range. - Other notes are given in the two Studies. Active Region Cool Loop Dynamics Louise Harra-Murnion Recent results from CDS on SoHO and TRACE have confirmed the existence of a separate class of loop system to the more familiar hot coronal loops - that of cool 'transition region' loop systems. The cool loops exist as high as the hot coronal loops but behave very differently.The differences are summarised as follows: 1) There are fewer cool loops than hot loops 2) The cool loops are more sharply defined 3) There is no 'diffuse' cool component 4) Cool loops are extremely dynamic The heating mechanism involved in the cool loop is different to that in the hot loops. CDS has identified the loop structures but to understand them further we must have higher time resolution (the limit on CDS imaging for this type of observation was approx 10-15 mins), and we must be able to observe the flows in the photosphere using OT, and then relating that to what is happening in the cool loops. Measurements of line broadening have been used to try and understand the heating mechanism in the corona. To pinpoint the heating mechanism we need to understand the non-thermal line broadening. There are some hints that the higher non-thermal velocities in the transition region are merely due to multi-directional flows. Study Details Raster Area: as large as practical - 4 X 4 arcmins Raster step: 2 arcsecs Raster Locations: 120 Exposure time: 2 secs Duration of Raster: 4 mins + unknown for overheads Number of rasters: for several hours on an active region on the limb, and then again for an active region on the disk. Line Selection: There are only a few transition region lines in this wavelength range. The first run would study on the cool loop systems only. It would then be extended to combine the information from the hot and cold to understand how they coexist together. He II 243.03 A 19 cts/s O V 248.46 A 9 cts/s Mg VI 268.99 A 17 cts/s Bins across line: 20 Telemetry: 3 lines x 120 bins x 120bins x 12 bits =518,400 bits per exposure. At 64 kb/s would be 8.1 secs. Need compression on order of 4. Supporting Observations: The most important aspect for the disk observations of the magnetic information at the resolution provided by the optical telescope on solar B. One of the biggest mysteries about the cool loops is how they coexist with the hotter coronal loops. If they are completely different structures this should be apparent in the magnetic data. What's new? Fast temporal resolution alongwith good spectral resolution for the line broadening measurements, along with high resolution magnetic information Emerging Flux - Coronal Response Title: EMERGING FLUX -- CORONAL RESPONSE Author: Peter Cargill (ICSTM) Justification: While the role that emerging photospheric flux plays in driving coronal processes is widely accepted, recent observations from the HAO/ASP, Hawaii magnetographs and SOHO/MDI have in fact revolutionised this field. The ground based instruments has indicated that flux can emerge in an already-twisted state, so that free energy is already available to power dynamic phenomena while MDI has shown the remarkable fact that flux is recycled roughly every 72 hours. Given this, flux emergence is likely to be a prime target for the Solar-B optical telescope. Given the pivotal role flux emergence plays, it is essential that its effect on the upper solar atmosphere be measured in conjunction with its emergence. In particular, it is essential to pin down the role that magnetic reconnection plays in the interaction of new flux with pre-existing coronal field. The best characterisation of magnetic reconnection are associated plasma motions and these are best measured by obtaining spectra of a few carefully chosen lines. The temperature choice is critical for this observing sequence, since we expect the temperature attained in reconnection to depend on the magnetic field strengths, where the reconnection takes place, and how fast the new flux is emerging. It is probably preferable to have better coverage at transition region temperatures, but some trial and error will be required to obtain the correct lines. Thus this run will probably need to be repeated. It is anticipated that such a sequence would have the optical telescope picking on a likely region for flux emergence, and staring at it for as long as is practical. We probably want a smaller FOV than OPT to get time resolution so will need advice on where to point. Probably want slot for context. Strategy: 6 - 8 lines covering TZ and corona. He II, OV (if strong enough), some of coronal Fe lines. Maybe a flare line in case there is strong heating. Raster area should be only part of FOV -- 1x1 arcmin say. With 1 sec exposure times (see RAH's entries for arguments for this) can get 30 second raster duration, about same as resolution of optical telecsope. Velocity coverage -- +/- 300 k/s is probably a good start, though this can be increased if required. Need slot for context. Again, similar resolution to OPT will do as a start. Coordination: This sequence requires very close coordination with OPT. XRT probably less necessary, but no harm in having it. What's new: The ability to observe flows in the transition region and corona in response to well-resolved observations of newly emerging flux Flares - Mass Motions in Coronal Lines Title: FLARES -- MASS MOTIONS IN CORONAL LINES Author: Peter Cargill (ICSTM) Justification: Solar flare energy release is believed to involve coronal energy release and subsequent energy transport to the photosphere. The photospheric response, a massive expansion of heated plasma into the corona, is generally referred to as chromospheric evaporation. The phenomenon was first detected by SMM and P78-1, but with no spatial resolution. The BCS instrument on Yohkoh has observed numerous cases of blue-shifted plasma, but of course did not have any spatial resolution. Solar-B provides the first opportunity to infer the location of these mass motions. In addition to blue shifts, line broadening has also been observed in coronal lines during flares. It has been suggested that this is turbulence associated with coronal reconnection and as such could be responsible for coronal particle acceleration. An observing sequence must scan a wide area and do so quickly. The problem is that flares tend to occur in a variety of places. The best bet is probably to make this part of a sequence where SOT studies a likely looking active region. The velocities inferred in these phenomena are quite large, so a large number of bins should be used. Unfortunately I don't see a way around a large raster area since flare loops can be quite long and we don't want to miss the action. So I suggest same FOV as SOT. Can we get exposure time < 1 sec? Are the lines strong? Strategy: Fe Flare lines (XXI and XXIII) are esential. Cooler coronal line and TZ line to see depth of heating in upper chromosphere. Density diagnostics would be nice, but appear to be restricted to cooler plasma, so probably not worth it. Need many bins in each line since we probably need to worry about velocities up to say 800 k/s. FOV is a problem. We probably want maximum in both slit and slot to be sure to catch loop footpoints. But this is a problem for the time resolution and needs some thought. Coordination: Coordination with OPT and XRT essential. What's new: First spatial determination of evaporation and turbulence in flares. Flares - Coronal Reconnection Title: FLARES -- CORONAL RECONNECTION Author: Peter Cargill (ICSTM) and Saku Tsuneta Justification: Since Petschek proposed his famous mechanism in 1964, magnetic reconnection has been seen as a viable mechanism for solar flare energy release. However, it was not until the launch of Yohkoh that good evidence for reconnection in the corona was obtained from the SXT and HXT instruments. Even so, the Yohkoh results could not produce the essential evidence for reconnection namely the appropriate mass motions. The reconnection process involves the conversion of magnetic energy into thermal and kinetic energy, in perhaps roughly equal parts. In addition, reconnection involves the flow of cool, unreconnected plasma into the reconnection site. Thus, spectral lines in the appropriate temperature ranges should be Doppler shifted. EIS provides the first opportunity to observe explicitly such motions. This is in fact quite a difficult observing sequence to construct. The inflow into the reconnection site is likely to be parallel to the surface of the Sun, while the high speed outflow is directed either downward or upward. So one would appear to be restricted to observing one or the other, though tilting of the reconnection site could alleviate this. The inflows would be best observed on the limb and the outflows on the disk. The latter could thus be combined with the previous entry on evaporation. Strategy: Reconnection outflows are expected to be hot, thus require the Fe XXI and XXIII lines. There is no harm in including a couple of cooler Fe lines, but TZ lines would not appear to be useful. For spectral resolution, we would need to measure up to 1000 k/s, so perhaps 100 bins are needed. For the inflows, a range of coronal lines (Fe IX - XVI) would work. Density diagnostics would be nice and this could be done with some of the Fe lines. The velocties expected are smaller than in the outflow region, so +/- 250 k/s would work. The FOV is a problem, since we want a large one in order not to miss the flare, but a high time resolution in order not to miss the peak of the reconnection. For context, we could probably use XRT rather than the slot. Coordination: Disk campaign for reconnection outflows would be in collaboration with OTP and XRT Limb campaign would involve only XRT, with OPT giving advice on likely-looking regions for a flare. What's new: The first accurate observations of mass flows associated with reconnection in flares. First estimate of reconnection rates in flares. Flares - Non-thermal Line Broadening Louise Harra-Murnion 11 August 1998Using the data from the Yohkoh spacecraft we have been able to increase our understanding of the various mechanisms involved in solar flares. The time behaviour of the hard X-ray bursts (which is traditionally viewed as the flare start) and the turbulence (as measured from excess spectral line broadening) was analysed. Interestingly, it was found that the peak of the turbulence occurred well before the peak of the hard X-rays (Alexander et al, 1998). Since turbulence occurs before the start of the flare this requires a major reassessment of what is actually triggering solar flares. Also, small flares occur more frequently than their larger counter-parts. We found that the turbulence of these small flares was surprisingly as large as that previously found in major flares (Harra-Murnion et al., 1997). This suggests that the flare trigger is the same in flares irrespective of their magnitude. The BCS onboard Yohkoh has provided us with many clues to the understanding of the solar flare trigger. However the major drawback is that it is a full Sun instrument with no spatial resolution. Efforts have been made to search for the region of highest turbulence by using the limb to occult the footpoints of the flaring loops (Khan et al, 1995, Mariska et al, 1996). The results were not conclusive. Study Details Raster Area: as large as practical. To obtain as high a time resolution as possible it would be good to have a 'nodding' slit-slot mechanism, by which we could move the slot for context information, and just use a few slit positions for the spectral information. If this was available then we could use 8 X 240 arcmin raster area. Raster step: 2 arcsecs Raster Locations: 4 Exposure time: 0.5 sec Duration of raster: 0.5*4 plus x overhead? Number of rasters: region monitored to get preflare as well as flare. The flare monitor shouldn't be used in this case. Line selection: concentrate on the higher temperature lines initially - then try a do a similar study for the cooler lines as well. Fe XV 284.16 A 1.8e4 cts/s Fe XIII 240.7 A 139 cts/s Fe XVI 262.98 A 3e3 cts/s Fe IX 244.92 A 131 cts/s Bins across line: 20 Telemetry: 4 lines x 20 bins X 120 bins X 12 bits = 115, 200 bits per exposure. At 64 kb/s would be 1.8 s. Need compression about 4. Supporting Observations: Solar B will provide the opportunity for the first time to observe the location of the line broadening in the corona, and also the magnetic motions on the surface of the sun which have been assumed to induce the non-thermal motions via various mechanisms such as waves. So the supporting observations will be; Optical telescope - to provide the magnetic information. X-ray telescope - to give context information. What's new? The main issue is that this will provide the first opportunity to observe flares with spatially resolved spectral information - hence the possibility to resolve the location of energy release. Abundance Anomalies STUDYING ABUNDANCE ANOMALIES WITH EIS The exact specifications of the EIS have not been finalised at the time of writing (29-Jun-98) and so this document discusses necessary conditions for investigations of the FIP effect. The FIP Effect Spectroscopic observations of the solar atmosphere, together with in situ measurements of the solar wind have revealed that element abundances often deviate from their values in the photosphere. A common feature of such abundance anomalies is that elements with a high first ionisation potential (FIP) are underabundant relative to those with a low FIP, although the magnitude of the discrepancy varies from feature-to-feature on the Sun. A useful summary of spectroscopic studies of the FIP effect is presented in Feldman (1992). Instrument Requirements The key requirement, in terms of studying abundance anomalies, is that the wavelength range that will be used for EIS contains emission lines of both low and high FIP elements that have some overlap in temperature. A secondary requirement is that a range of consecutive ions of low and high FIP ions are observed, to help minimise errors in the atomic data. As an example, the Normal Incidence Spectrometer (NIS) of CDS/SOHO observed lines of Ne IV - VII and Mg V - X (where neon is a high FIP element and magnesium a low FIP element), for which there is significant overlap in temperature around logT = 5.4 - 5.7. See, for example, Fig.5 of Young & Mason (1998). One difficulty in studying element abundances is that low FIP ions give rise to many strong lines at temperatures logT > 5.7, but few below this temperature, whereas the opposite is true for high FIP ions. Two high FIP elements that do give rise to EUV lines at coronal temperatures are argon and sulphur, but the lines are often rather weak and so high instrument sensitivity and spectral resolution are required to measure these lines accurately. References Feldman U. (1992). Physica Scripta 46, 202. Young P.R. & Mason H.E. (1998). Proceedings of the ISSI workshop `Solar Composition and its Evolution - from Core to Corona'.

Study Sheet

	EIS Study Sheet
	SUMMARY of TARGET
	Target	:
	Temperature	:
	Emisson Measure	:
	Approx target area	:
	Sequence duration	:
	Interuptible	?	1
	Feature Tracking	:
	Response to brightening, flare etc.	:	2
	Interaction with other Solar-B instruments	:
	SEQUENCE DETAILS:	Spectroscopy		Imaging
	Needed?	:		:	3
	Interval between Spectroscopy and Imaging			:	4
	Require narrow-band imaging (overlap-o-gram)?			:
	NS extent	:		:
	EW extent	:		:	7
	EW scan step	:		:
	Which lines?	:		:	5
		:		:
		:		:
		:		:
	# of lines	:		:
	Spectral Range	:	6	n/a
	Spectral Resolution	:	6	n/a
	NS resolution	:		:	8
	EW resolution			:
	Exposure Duration	:		:
	Desired Cadence	:		:	10
	Compression	:		:	11
Notes
1	i.e. will incomplete sequences have any value?
2	e.g. go and look, do nothing, or abandon sequence
3	Yes or No
4	put 0 for simultaneous
5	indicate which lines should drive the exposures if count-rate limited
6	can specify wavelength or velocity - could be line-dependent
7	scan range
8	could be entire slit integration
9	-
10	readout overhead is TBD
11	What will be acceptable e.g. bit compression, JPEG, NONE

If you have comments or suggestions, email me at mwt@mssl.ucl.ac.uk