Electron temperature from highn Balmer series at JET

Electron temperature from high-n Balmer series at JET: A question of “n” A. Meigs, M. O’Mullane, and E. Delabie (discussions of future modelling EDGE 2 D/Eirene with Paula Belo and Mathias Groth) From Stark widths one can get estimate of electron density (ADAS 2013, PSI 2012, …) Presumably, given this knowledge of ne, then the line intensities or line ratio’s should give an estimate of Te (and possibly nn/ni) ADSA 2014/Ameigs 29 -09 -14

Questions: • What are the best set(s) of n-levels to measure? • Is the “high-resolution” spectra (left) capable of giving Te estimate? (or even shifted to cover 9 -2 to 14 -2) • Can we do better for Te by using a survey spectra? • Are other effects (MAR) possibly important? How can we tell? ADSA 2014/Ameigs 29 -09 -14

KT 3 D Overview: Lines of sight Passive Divertor Spectroscopy • • • Mirror-link largest wavelength range of any JET diagnostic (200 -1200 nm) D influx and recycling profiles Recombination monitoring ne and Te in recombining divertor Impurity influx and profiles (Be, C, W, O, N, Ar…) ADAS 2014/Ameigs 29 -09 -14

Single Shell Model of Balmer Line Intensities Intensity of a Balmer transition u-2: Only 3 -body recombination and electron impact excitation are considered (for now) Where: Too many unknown’s u = upper n-level for the transition Needs full modelling codes to get at all parameters ne = electron density Can we reduce this set using line ratios ni = ion density nn = neutral density Pu. R (ne, Te) = the recombination Photon Emission Coefficient from ADAS for a transition u-2 Pu. E (ne, Te) = the excitation Photon Emission Coefficient from ADAS DL = the integration path length (assuming single shell model) Wrapper to ADF 15 using file '/home/cxs/adas/adf 15/pec 12#h_balmer#h 0. dat' so that recombination only and excitation only components could be retrieved. ADSA 2014/Ameigs 29 -09 -14

Can we do better with line ratio’s (and which ones)? Where the second formula we have created a new variable a = nn/ni Assuming the single shell model is ok, then the line ratio removes: DL, linear ne (only ne now in PEC’s) ADSA 2014/Ameigs 29 -09 -14

Balmer Line Ratios: Further factoring makes apparent a “correction” term to the pure recombination ratio: So the relative size of this correction term depends on both the relative ratio of excitation to recombination for the u-state and the r-state. For a given a, a u-state dominated by recombination and a r-state driven more by excitation would provide the largest correction factor. ADSA 2014/Ameigs 29 -09 -14

ADAS Balmer Line Ratios: 12/6 5. 7 x ADSA 2014/Ameigs 29 -09 -14

ADAS Balmer Line Ratios: 9/6 2. 9 x ADSA 2014/Ameigs 29 -09 -14

Examples of Balmer Line Ratios: 12/9 1. 6 x ADSA 2014/Ameigs 29 -09 -14

Balmer Line Ratios: Measured in “Survey Spectra” JET pulse: 80821 Ohmic Vertical outer strike point Gas ramp to density limit disruption. 13 -2 ADSA 2014/Ameigs 9 -2 8 -2 7 -2 6 -2 29 -09 -14

80821 Electron density from line shape ADSA 2014/Ameigs 29 -09 -14

80821 Intensity of the 10 -2 Line and ratio 10 -2/6 -2 Intensity 10 -2 transition Intensity Ratio 10 -2 to 6 -2 Some ratios above ADAS recombination only results ADSA 2014/Ameigs 29 -09 -14

80821 Intensity of the 8 -2 Line and ratio 8 -2/6 -2 Intensity 8 -2 transition Intensity Ratio 8 -2 to 6 -2 Possible reasons for too large ratios: 1) For highest n’s (14 -2, 13 -2, maybe 12 -2) near the continuum merging/limit: incomplete baseline subtraction 2) For both high-n and lower-n: other population mechanisms MAR? More ratio’s above ADAS prediction ADSA 2014/Ameigs 29 -09 -14

80821 Te from fit to line ratios: 7 -12 over 6 Naively perform fit of ratio’s as function of n considering as reference the 6 -2 transition All fits suffer from low-n ratio discrepancy forcing high-n under-shoot of Te t= 59. 42 s Is the low-n discrepancy indicative of the contribution of excitation? Certainly, the excitation term would correct the ADAS values towards the n=7 and n-8 data. ADSA 2014/Ameigs 29 -09 -14

80821 Te from fit to line ratios: 7 -12 over 6 t= 60. 22 s ADSA 2014/Ameigs 29 -09 -14

80821 Te from fit to line ratios: restrict Nupper 9 -12 t= 60. 22 s Still a small problem with n=9 ADSA 2014/Ameigs 29 -09 -14

ADAS Balmer Ratio: 7/6 ADSA 2014/Ameigs 29 -09 -14

80821 Look at individual ratios Infer Te thru interpolation: given a ratio, interpolate ADAS ratio to obtain Te (still using ne from line shape) Ensure no extrapolation beyond the Te grid passed to ADAS (Tmax = 10. 0 e. V) if result > max Te of grid, then Te = -1. 0 Not (yet) ensuring that the ratio is in “useful range” or even above ADAS limit (extrapolation limit will get probably all of the above limit values) **After doing this I think the limit should have been ~3. 0 e. V (but need to look at the individual ratio curves to find max) ADSA 2014/Ameigs 29 -09 -14

80821 Look at individual ratios Zoomed away from x-point Electron density 10/6 Te 10/6 ADSA 2014/Ameigs 29 -09 -14

80821 Look at 9 -12 fit vs individual ratios Comparison of Te from individual ratios 10/6, 11/6, 9/6, 12/6 and the fitted 9 -12 N-fit error bars are 2 -s ADSA 2014/Ameigs 29 -09 -14

High resolution setting: 81469 L-mode Density Limit 1. 46 x ADSA 2014/Ameigs 29 -09 -14

High res. setting: 81469 L-mode DL, Horizontal target ne 11 -2 Intensity 12 -2 Intensity Ratio 11/10 Ratio 12/10 Restricted (for now) to 11 -2 and 12 -2 as 13 -2 and 14 -2 may not be “reliable” (later, this is questioned) ADSA 2014/Ameigs 29 -09 -14

High resolution setting: 81469 L-mode Density Limit Gibberish…. . Te from 11/10 Te from 12/10 Why? Dn too small so dynamic range of ratio too restricted? Does horizontal target delocalize the measurement too much? Are MAR effects more important in this higher density plasma compared to the Ohmic case? ADSA 2014/Ameigs 29 -09 -14

81469 n-fit 11/10 and 12/10 only vs infered 11/10, 12/10 Large errors indicate one or other ratio at or near limits ADSA 2014/Ameigs Te max= 4 e. V for inferred results 29 -09 -14

High resolution setting: 81469 L-mode Density Limit The sky blue areas are above the ADAS limit Ratio 12/10 ADSA 2014/Ameigs 29 -09 -14

81469: extend to 13 -2 and 14 -2 over 10 -2… For fun, I thought to include 13 -2 and 14 -2 ratios…. . Note: • 13 -2 may experience more of the continuum merging baseline which is not accounted for in current fits. • 14 -2 will have more of this baseline, plus it is under the Be III triplet! (how much to trust the intensity) Peak near strike point! ADSA 2014/Ameigs 29 -09 -14

81469 profiles versus time Nice correlation with ne versus time (but is it “just” a correlation) Are the fits believable yet? NO, I think not. ADSA 2014/Ameigs 29 -09 -14

Conclusions? And further work Conclusions: • Survey spectra may allow some determination of electron temperature and possibly identify when excitation comes in; plus maybe nn/ni estimation Further work: 1. Check if 80821, a clean Ohmic plasma, has problems with 11/10 and 12/10 ratio’s in the survey spectra 2. Look at the 9 -14 spectra (started but mixed results and way too many slides as it is) 3. Most measurements since ~Feb use the 9 -14 spectral setting and where repeats are possible the 6 -14 survey spectra is taken Some H-mode plasmas fitted for ne but none looked at for line ratio’s 4. EDGE 2 D/Eirene (plasma modelling codes) have implemented simulation of the KT 3 diagnostics lines of sight and have imported ADAS results to calculate the LOS-integrals for the high-n Balmer series (need to make sure the ADAS dataset is correct one) Look at LOS-integrals and ratios from modelling codes compared to experimental results. 5. Nail down some constraints on the n-fit so that acceptable ranges are not exceeded and so that periods where the data ratio’s exceed the ADAS ratio’s are identified more easily 6. Look into the state of the art on MAR and see how to include it in the CR modelling of the Balmer populations. 7. For survey try to fit getting nn/ni from the 7/6 and 8/6 ADSA 2014/Ameigs 29 -09 -14

Look at L-mode Vertical target: new range 9 -14 85106 outer vertical target L-mode density limit ADSA 2014/Ameigs 29 -09 -14

Recombination only“Intensity” vs n for several Te Assuming recombination only Setting ni= ne ADSA 2014/Ameigs 29 -09 -14

9 -14 for two densities ADSA 2014/Ameigs 29 -09 -14

Wrapper to ADF 15 for high-n Balmer lines FUNCTION agm_readadf 15_balmer, nupper, $ dens = dens, $ ; ; input, if not present then ADAS grid returned te = te, $ ; ; input, if not present then ADAS grid returned exciteonly = exciteonly, $ recombonly = recombonly, $ all = all, $ ; ; if want 2 D grid on Te and Ne values-- this is a ; ; keyword to read_adf 15. If not set then 1 D results ; ; numbering the smaller of Te and dens arrays verbose = verbose filename = '/home/cxs/adas/adf 15/pec 12#h_balmer#h 0. dat' ADSA 2014/Ameigs 29 -09 -14