Greetings, I have followed the suggestions for data display outlined in Mat's note of 20 April with some good luck. In this very long message I o Describe the new plots I have developed o Tell you where to find postscript files of the plots (I will FAX or FEDEX hard copies ONLY if you ask) o Provide the promised first cut at an outline for a paper. o And, the most important, writing assignments with deadlines for all! We have to get this completed quickly because Ann and I will be tied up with new launches later this year. Enjoy, Bill ps: I put a copy of this message in the file April_note.txt in the same directory with the postscript files. You can use anonymous ftp to get the files from sierra.space.lockheed.com in the directory /DATA/de/ICRH ######################################## o Most important item: Ann will take on the responsibility for writing the introduction and putting the pieces together into a finished product. She will be first author. I will take care of getting the figures into final form and take the lead in the data presentation section. Mats will provide a more complete outline of the discussion section and a few complete paragraphs. John will provide a few paragraphs for the discussion section Tom will keep us honest, help us write the abstract and provide the first rough draft of the conclusions section. All are encouraged to make comments/corrections/additions/ corrections/deletions (including to the above assignments). Have I left anybody or any important item out? Let's try to get our first inputs to Ann before the AGU week in May so she can get a first draft back to us before the IAGA meeting. ######################################## o Description the new plots I have developed (next section tells you how to get electronic versions of them. SPECTROGRAMS: c82047.ps, c82111.ps, c83064.ps (color) b82047.ps, b82111.ps, b83064.ps (black/white) There is now a color and/or gray scale spectrum for each interval with EICS H+, He+ and O+, 0-8 Hz PWI, and the high frequency PWI SFR data all on the same time scale. The two PWI spectrograms don't need much explanation except for the frequency range on the ordinate and power spectral density range covered by the full range of the color bar For the 0-8 Hz PWI the color bar goes from -4 to -9 to cover the range of log(10) PSD in units of (V/M)^2/Hz, in 24 frequency channels. the Frequency range is LINEAR. For the SFR/LFC plots the frequency range is 1 Hz to 410 kHz in 8+128=136 channels, and the color bar goes from -16 to -8 for log(10) of the psd in units of (V/M)^2/Hz. The 8 LFC bins are wider than the 128 SFR frequency bins so that the frequency scale is logarithmic from 1Hz to 410 kHz. The EICS spectrogram panels are a new format! Its similar to one of the formats Andrew Yau uses to display data from his Akebono instrument. You will remember that EICS takes 96 seconds to cover the complete mass/energy/pitch angle range for all three of the data intervals we are discussing. This is a long time compared to the SFR (32 seconds) and 0-8 Hz PWI data (8 seconds), so there are 6, 5, and 8 complete energy/pitch angle sweeps for the data obtained on 82047, 82111, and 83064 respectively. The data are presented in the proper 'block' of time, but the order in which the they were acquired is lost in this presentation. For each block of time 12 pitch angles (ordinate) and 15 energy bins covering the range 10 eV to 17 keV (along the time axis) are shown. The quantity encoded in the gray/color bar is number flux, not the velocity space density that we use for contour plots. The TIME INTERVALS covered on the spectrograms are of a different length for each of the three days. The plots have EXACTLY the same temporal and physical extent as the line plots for the same day described below. The start and stop times in seconds for each spectrogram are indicated at the top of the spectrogram in seconds of the day. I tried to match the SFR/LFC 32 second accumulation times will the 96 second EICS, but asked Ann for the wrong time intervals. Rather than bother Ann with busy work, I offset the SFR/LFC spectrograms to account for the different start times. The labels on the spectrograms (i.e. times, energies, frequencies, color scales are missing. I'll have our art department put those on after we all agree on the final figures. The published black and white spectrograms will have a gray scale reversed from that in the on-line version. It will go from white (lowest intensity) to black (highest intensity) so that it will look good in the journal. LINE PLOTS: l82047.ps, l82111.ps, l83064.ps avg_pwi_82047.ps I made several modifications to the line plots showing energies characteristic of EICS and calculated from the slope (alpha) and power spectrol density (psd) of the PWI data at the appropriate ion gyrofrequency. There was some ambiguity associated with negative slopes. The theory didn't address it and I was unclear. John pointed out that a flat slope (alpha=0) implies that most of the heating occurs at altitudes well below the observation and large values of alpha imply local heating. In the original version of the line plots I more or less randomly set alpha to either 0 or 100 at points were the slope on the PWI data was negative. The current plots were made with alpha set equal to 100 everywhere the PWI slope was negative. This results in significantly lower expected energies from the PWI calculation. I also made several physical changes to the line plots. There are now plots of alpha and the psd scaled from PWI data, so for each day and each mass there are 5 line plots: The characteristic transverse energy scaled from EICS I described in my earlier note. NOTE: that after looking at the plots It was clear that I misinterpreted the contours for the interval centered on 67718 seconds on 82047, so the characteristic energy and ratio plots have different values for this interval than those in the first batch I put on line for all three masses. The prior interval (67621) was rescaled for O+ energies only. The energy calculated from the slope (alpha) and psd at the appropriate ion gyrofrequency. The ratio of these two energies The slope (alpha) clearly indicating where I have set it to 100. The psd measurement. I looked at getting a characteristic parallel energy as John suggested but did not come up with anything I had confidence in. The times on the line plots are in decimal hours and EXACTLY match the center times of the EICS accumulation interval used. The offset in the center times of the EICS accumulation intervals and the data points in the line plots is intended. Remember that I had to restrict my self to scaling energies from EICS from 8 of the 16 spins in each accumulation interval to get sensible characteristic transverse energies. The bad news is that each line plot is on a separate page! I had to do this to get a reasonable aspect ratio and time scales that exactly match those on the spectrograms. You now have 15 lines plots (5 for each mass) associated with each spectrogram. If you get the electronic plot files, you will have 45 line plots and 3 spectrograms. I pasted together the 5 line plots for each mass for each interval and will put copies of them in the mail to you. If you want fax or FEDEX delivery ask and I'll get it out. The file: avg_pwi_82047.ps has 48 second average plots of the PWI psd over the 0-8 hz range. There are 8 plots on 2 sheets. You will remember that I could only use the EICS data from half (48 seconds) of the full eics instrument cycle. The center times 67620, 67716, 67812, and 67908 all correspond (within a few seconds) to the center times of the accumulation of the EICS data used to scale the characteristics. These plots are on the left side of each page. The plots on the right side are there, because I was too lazy to mess with the code and get rid of them. NOTE however that for the interval centered on 67716 that the dip near the He+ gyrofrequency,2.7 Hz (B=715 gamma)2.7 Hz is persistent. I did the 48 second average starting 8 seconds before and 8 seconds after and the dip at frequencies slightly less than 2.7 hz showed up in all three plots. This dip is what leads to a NEGATIVE slope near the He+ gyrofrequency for this interval which is set equal to zero in the plots (see below). There appears also to be a slight negative slope below 8 Hz. I took the slope for the H+ data (gyrofrequency=10.9 hz) from the 1Hz-400 kHz SFR/LFC plots Ann provided. The slope at ~11 Hz was positive. (see the outline of the paper below for a first order on how to handle the negative slope problem). ######################################## o Where to find postscript files of the new plots and other information on how to cut and paste 45 pages of line plots to get 15 line plots!. SPECTROGRAMS: c82047.ps, c82111.ps, c83064.ps (color) b82047.ps, b82111.ps, b83064.ps (black/white) LINE PLOTS: l82047.ps, l82111.ps, l83064.ps avg_pwi_82047.ps I have put several files on on sierra.space.lockheed.com in the directory /DATA/de/ICRH You can get them by using anonymous ftp (ftp to sierra.space.lockheed.com, username anonymous, any password, and cd /DATA/de/ICRH etc...) ftp sierra.space.lockheed.com Name (sierra:pete): anonymous 331 Guest login ok, send ident as password. Password: 230 Guest login ok, access restrictions apply. ftp> cd DATA/de/ICRH 250 CWD command successful. ftp> dir -rw-rw-rw- 1 ftp guest 11878 Apr 7 11:36 April_note.txt -rw-rw-rw- 1 webmaste nobody 274625 Apr 10 09:31 avg_pwi_82047.ps -rw-rw-rw- 1 webmaste nobody 84580 Apr 7 10:37 b82047.ps -rw-rw-rw- 1 webmaste nobody 78974 Apr 7 10:38 b82111.ps -rw-rw-rw- 1 webmaste nobody 103311 Apr 7 10:39 b83064.ps -rw-rw-rw- 1 webmaste nobody 84580 Apr 6 15:49 c82047.ps -rw-rw-rw- 1 webmaste nobody 78974 Apr 6 15:41 c82111.ps -rw-rw-rw- 1 webmaste nobody 103311 Apr 6 16:01 c83064.ps -rw-r--r-- 1 webmaste nobody 587 Apr 11 13:42 h82047.dat -rw-r--r-- 1 webmaste nobody 471 Apr 5 12:59 h82111.dat -rw-r--r-- 1 webmaste nobody 654 Apr 5 12:59 h83064.dat -rw-r--r-- 1 webmaste nobody 598 Apr 11 13:42 he82047.dat -rw-r--r-- 1 webmaste nobody 597 Apr 5 12:59 he82111.dat -rw-r--r-- 1 webmaste nobody 655 Apr 5 12:59 he83064.dat -rw-rw-rw- 1 webmaste nobody 583190 Apr 6 16:03 l82011.ps -rw-rw-rw- 1 webmaste nobody 735781 Apr 11 13:42 l82047.ps -rw-rw-rw- 1 webmaste nobody 627906 Apr 6 16:02 l83064.ps -rw-r--r-- 1 webmaste nobody 648 Apr 11 13:42 o82047.dat -rw-r--r-- 1 webmaste nobody 539 Apr 5 13:00 o82111.dat -rw-r--r-- 1 webmaste nobody 619 Apr 5 13:00 o83064.dat -rw-r--r-- 1 webmaste nobody 4858 Apr 5 12:58 theory.pro 226 ASCII Transfer complete. 1402 bytes received in 0.26 seconds (5.3 Kbytes/s) ftp> the *.dat files are the input files processed by the idl code theory.pro that produces the line plots. ######################################## o The promised first cut at an outline for a paper Working title: A multi-species test of ICRH heating at high altitudes Abstract: tbd Intro: Follows Ann's proposal with the additional information that we tried the Crew technique for He+ and found that there were inadequate counts in the EICS detector to get robust estimates of the He+ energy from moments of the particle distribution. DATA: We selected several events for hand analysis... Ie obtained characteristic energies (from EICS) and slopes (from PWI) by hand scaling appropriate plots. First interval: 83064. (March 5, 1983). EICS He+ conic finding code found a conic in the 96 second interval centered on 23:47:12 (23.79 on the plots) or 85632 seconds. where B=1668 nT, O+ gyrofreq =1.6 Hz and He+ freq= 6.4 Hz. The Kp index was was high, 5- Show spectrograms of all three masses and both frequency ranges and three line plots, one for each mass. Each shows the EICS scaled energy, the energy derived from the slope of the PSD at the gyrofrequency (alpha) and the value of the psd and uses equation 13 of Retter et al. [in Chang's conference proceeding, Vol 6 (85-87), page 97] to calculate a characteristic energy. Eta, the left hand polarization fraction was set to 1.0 in calculating the energies plotted. The third panel is the ratio of the two characteristic energies. for 83064 He+, O+, and H+ characteristic energies maximize in the fourth interval centered on 85619 seconds (23.78 hrs) which shows up nicely for all three masses in the color spectrograms and nicely for He+ and O+ in the b/w spectrograms. For all three masses there is a local minimum in the ratio of the two characteristic energies in the fourth interval. (I don't know what it means but it stands out in the data.) The psd at the He+ gyrofrequency (~6 Hz, it varies from 5.6 Hz in the first interval to 7.1 Hz in the last) peaks before and after 23.78 hrs so the characteristic energy calculated from the PWI data does not maximize at 23.78, rather it is high until the The psd at the H+ gyrofrequency (~20 Hz) peaks near 23.79 and remains intense until ~23.81. The psd at the O+ frequency (~1.5 Hz) is most intense in the two intervals before the interval with the maximum characteristic eics temperature. CONCLUSION for 83064: the crew approach to ICHR heating best fits the H+ data in this interval. For helium the peak energy predicted from the waves is anti-correlated with the particle observations. This could of-course be due to temporal aliasing of the EICS measurements or a systematic variation of eta (the left hand polarization of the waves) across the event. The crew approach to ICHR heating of O+ misses the peak, but this could be the result of ambiguities in the very low frequency (1.5 Hz) range from ... >>>>>>>>>>>>>>>> ...>>>>>><<<<<<<<<<<<<<<< Second interval 82047. CONIC finding code found the most intense He+ conic in the 96 second interval centered on 67920 seconds (18.87 hrs) or 18:52:00 It didn't find he+ conics in the other intervals shown in the spectrogram. SAME set of spectrograms for this case, all three masses and both frequency ranges. as well as same set of three line plots (one for each mass) of three panels each. Scaling the EICS characteristic energy disagrees slightly with the conic finding program! The max transverse He+ energy found in the previous sample centered at ~18.84. BE SURE you are looking at the current versions of the eics characteristic energies. The first set of these plots had the max energy scaled from EICS at 18.81. The changes that were made to the EICS energy are noted in the description of the line plots above and in the data files. Basically there was a strong perpendicular drift at 18.81 and 18.83 that made interpretation of the contours ambiguous. The He+ energy calculated from the slope and psd now maximizes at 18.84 also. There is a file avg_pwi_82047.ps (described above) that shows a dip near the He+ gyrofrequency centered on 67716 (18.81) that appears to be real; at least the dip remains when the averaging interval is shifted slightly. The slope of the PWI data for this interval is negative. The energy derived from PWI was calculated using alpha of 100. (See discussion about negative alphas' below). . For O+ the peak energy was scaled from EICS at the same time as the He+ peak (18.84). The PWI data however imply a lot of O+ heating over an extended region. This comes from the small values of alpha scaled from the PWI data. (The values of alpha for this interval are in the files he82047.dat, o82047.dat and h82047.dat). I scaled a slope of 0.7 at the O+ gyrofrequency (0.6 Hz) for the first two intervals here. This low value of the slope increases the predicted energy because it implies heating over a larger altitude range. The H+ EICS spectrogram shows a conic in the low energy channels centered on 18.81 that corresponds with a peak in the psd at the H+ gyrofrequency (10.5 Hz). It looks like we can argue that the crew method works really well for He+ and H+ for this interval and not that bad for O+ in picking the location of maximum transverse energy transfer. (defer discussion of efficiency to discussion section). >>>>>>>>>>>>>>>>>>>>>>>> Third interval 82111. Included to investigate negative slope problem. Lets present only the He+ eics spectrogram and no O+ and H+ line plots. (I make comments about them here, but suggest we leave them out of the paper). PWI shows line emission near ??? Hz prior to (start time) to about 0.14 hrs. This gives a slope at the He+ gyrofrequency (1.0 Hz) of -6.0 and -1.85 for the first two intervals (0.09 and 0.12 hrs). I used large alpha, which imply local heating, and didn't do too bad a job in matching the EICS energy profile if you allow a large uncertainty on the interval at 0.12 hrs. O+ and H+ peak energies scaled from EICS and PWI data both maximize well away from the He+ peak associated with the line emission. Note He+ line emission heating by the ICHR mechanism has been discussed at length by Anderson and Fuselier (JGR 99, 19413, 1994.) >>>>>>>>>>>>>>>>>>>>>> Discussion: Qualitative fit of peak energies determined by scaling energies and calculating crew theory assuming extended altitude range and scaling from psd and slope of psd at gyrofrequency. Quantitative fit of peak energies. Discussion of the magnitude of the correction factors in the line plots. For 82111, only He+ is heated, and here there are waves only around the He+ gyrofrequency (which was the point of including this event) and here the correction factor is above unity for the heating region, so we are doing fine on this one. Conclusions: tbd <<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>