JOP 52: Lock-In Detection of Network-Driven Waves in the Corona

Craig DeForest


I hope to identify signatures of coherent wave motion on a five-minute time scale, or to reject the hypothesis that motion of the chromospheric network excites macroscopic waves that, in turn, heat the corona. Especially in light of recent developments, such as the measurement by UVCS of very high, mass dependent ``kinetic temperatures'' that may be associated with small scale, time dependent bulk flows; and such as SOHO observations of polar plumes, which appear to be relatively steady on a five-minute time scale and not to be heated by micro-impulsive events on arcsecond scales at their footpoints, there is a pressing need to investigate signatures of alfvén and other mechanical waves in the lower to mid corona.

The chief trade-off in measurements of wave motion in the lower to mid corona has been one of spectrometer slit width vs. exposure time: If a spectrometer slit is narrowed enough to provide the necessary spectral resolution to identify the doppler shifts from coronal wave motion, then the required exposure time to get good photon statistics precludes the measurement of dynamically changing doppler shifts as anything but a general line broadening.

By tracing a feature through the EIT field of view to the bottom of the corona, it should be possible to identify the network cell(s) associated with that feature's footpoint(s). If one hypothesizes that wave motion of the feature is being excited by network oscillations in the footpoint(s), then one may expect that the feature's spectrum will exhibit Doppler shifts with a particular time delay relationship to motions of the network at the footpoint(s).

Data Collection

There will be three phases to the sequence: a high cadence phase, in which oscillation data is collected, and pre- and post- observation phases, in which reference spectra and images are collected.

During the high cadence phase, we will record a large number of short duration (1 min) spectra of several features (initially, polar plumes) with SUMER and/or UVCS, and simultaneously record a dopplergram and/or line depth image of the feature's footpoints each minute with MDI. The region of interest (ROI) will be determined from synoptic EIT images shortly prior to the observing run. It is also desirable to obtain EIT images of the ROI at the start of, and preferably during, the high cadence phase of the observation. (It would be useful to attempt to use EIT images for a similar study in intensity).

A SUMER observing plan has been devised, that will generate repeated one-minute-cadence scans of a particular ROI, at low spatial resolution (six individual spectra per minute). Before and after the one-minute-cadence portion of the program, SUMER will generate high spatial resolution scans of the same ROI. The lines to be used include O V, O VI, and Mg IX.

A UVCS plan is ancipated, but has not yet been layed out in detail.

Data Reduction

By using the MDI network oscillation signal as a ``chopper'' input, we hope to be able to coherently add large numbers of short-duration coronal spectra, all from the same phase of a particular network cell's oscillation cycle.

Post facto, all of the spectra from a particular spatial position, and sharing a particular phase relationship with the MDI footpoint data, will be combined. For example, in the preliminary analysis, spectra that were taken during a minute in which the footpoint network cell's doppler shift was observed to shift from blue to red would be averaged together. Similarly, all spectra taken during a minute in which the footpoint network cell's doppler shift was observed to shift from red to blue would be averaged together. The measured quantity, network-phase-dependent doppler shift (or line broadening), is the difference between the wavelength (or line width) of each line in the two relative phases.

More detailed analysis will include a careful study of phase delays and (provided that a positive signal is identified) will provide a direct measurement of the alfvén speed in the feature of interest. Assuming a 1 minute cadence for the observations, a 5 minute oscillation period, and an 8 second exposure time for each spectrum, an eight hour run would yield an effective exposure time of (8 sec) (1/5 min) (480 min) = 768 sec in the coherently added spectra. Spatial binning along the slit length will yield up to an order of magnitude longer effective exposure times. The differential analysis technique will avoid the need for careful flat fielding or detector distortion mapping for this sensitive measurement.

SoHO Scheduling Considerations

It is desirable to obtain a reasonably long continuous observing period, both to reduce the frequency sidelobes of what is essentially a coronoseismology experiment, and to ensure that enough photons are collected to reduce shot noise to insignificance in the final synthetic spectra. At a height of 15 - 45 arc sec in O VI, SUMER count rates are on the order of 1 - 5 counts /pixel-second. It is anticipate that, at this accumulation rate, and with a 1/50 duty cycle for each raster, 6-8 hours will be required to obtain optimal results for the SUMER high cadence phase of the observations. Allowing some time before and after for the reference spectra yields a desired observing time of 8-10 hours. During the high cadence period, MDI should be downlinking high rate telemetry. It is suggested that an entire DSN ``long pass'' be dedicated to this measurement.

Additional Instruments

UVCS may participate as well; the decision is pending the results of a preliminary analysis of the JOP-039 data, which included high cadence spectra from UVCS.

A search for density signatures of waves in the EIT images has been considered and may be performed simultaneously with the SUMER study, pending a feasibility calculation.