Summary of EFI splinter Delft (NL) Summary of EFI splinter Raffaella D’Amicis 7th Data Quality Workshop 27 October 2017

Contents We had a very rich and fruitful EFI splinter session EFI science Thermal Ion Imagers: anomaly, operations and datasets, validation Langmuir Probes: status, operations and datasets Validation: faceplate, ISR, radio occultation Summary of open points and recommendations

EFI science EFI keeps delivering scientifically valuable measurements Several published and submitted papers in 2017 Published: Birkeland Current Boundary Flows, Archer et al., JGR 630 nm auroral emission height, Gillies et al., JGR EFI Instrument and performance, Knudsen et al., JGR Electron heating in pulsating aurora, Liang et al., JGR High-latitude Poynting flux, Park et al., JGR Multiple arc current systems, Wu et al., JGR Submitted: Distinguishing BCBFs and SAIDs, Archer & Knudsen, JGR “STEVE” arcs– MacDonald et al., Science Advances Alfvén waves in a dual-arc system, Miles et al., GRL Alfvén waves in MI Coupling, Pakhotin et al., JGR Langmuir probe cal/val, Lomidze et al., Radio Science Birkeland Current Boundary Flows vs Sub-Auroral Ion Drifts Archer and Knudsen, JGR, submitted

EFI science EFI keeps delivering scientifically valuable measurements Several published and submitted papers in 2017 Published: Birkeland Current Boundary Flows, Archer et al., JGR 630 nm auroral emission height, Gillies et al., JGR EFI Instrument and performance, Knudsen et al., JGR Electron heating in pulsating aurora, Liang et al., JGR High-latitude Poynting flux, Park et al., JGR Multiple arc current systems, Wu et al., JGR Submitted: Distinguishing BCBFs and SAIDs, Archer & Knudsen, JGR “STEVE” arcs– MacDonald et al., Science Advances Alfvén waves in a dual-arc system, Miles et al., GRL Alfvén waves in MI Coupling, Pakhotin et al., JGR Langmuir probe cal/val, Lomidze et al., Radio Science Correlated E and dB from FACs and Alfvén waves Pakhotin et al., JGR, submitted

TII imaging anomaly Swarm A and B EFI TII’s continue to provide good data with reduced operations (now 3 orbits per day) Good Poor Numerous tests have been conducted to investigate H2O hypothesis, and to mitigate the problem, including scrubbing, heater tests and high-voltage supply commissioning. Tests are still on-going and are designed to improve image anomaly although we are aware that moisture might never be fully removed within operating lifetime. Good vs Poor images showing “ring” and “wing” anomalies (Courtesy of J. Burchill and Torr Scientific)

EFI-TII operations Anomaly workaround: Special operations: Reduced number of daily operational orbits: 6 in 2016 and 3 in 2017 on Alpha (since July 2017) and Bravo 3 then 1 (since June 2017) on Charlie Daily scrubbing of Inner Dome implemented since Spring 2017 Full scrubbing of Phosphor Plate, MCP and Inner Dome sequentially (TBC) Special operations: Commissioning of MCP from 2.2 kV to 2.4 kV (to allow improved scrubbing) on all spacecraft Commissioning of Phosphor Plate from 5 kV to 7 kV (to allow improved scrubbing) on Charlie, and on Alpha and Bravo next month Bravo yaw-manoeuvred 15°during 1 week in March to test heating the TII sensor Image performance tested under various Inner Dome voltage (60 to 100 V) and Phosphor Screen voltage (3 to 5 kV)

EFI science operations 03/2017 SW-B Heating test with yaw bias 08/2017 Sw-A/B MCP commissioning to 2400V 04/2017 Scrubbing included in R&C and OPF (Courtesy of Serenella Di Betta) 08/2017 Investigation on spikes in phosphor plate voltage. 08/2017 SW-C Phosphor Plate commissioning to 7000V 11/2016 Sw-A/B MCP commissioning to 2250V 05/2017 TPF commanding as part of the nominal Ramp-Up event in R&C 05/2015 EFI Active only for few orbits per day Regular activity 09/2017 SW-C both LP in high gain during the TII Ready state 09/2016 Sw-C EFI Heating test 09/2017 HK sampling rate to 1Hz around the Ramp-Up to support detection of spikes in the voltages – automatized via R&C 08/2016 Sw-C MCP commissioning to 2250V 04/2014 End of official commissioning for EFI 10/2015 Introduction of OPF to handle EFI state transitions 06/2017 Sw-A TII inner dome ring test 07/2016 Different TII Image rate for different EFI state 07/2017 Sw-C MCP commissioning to 2400V 11/2017 SW-A/B Phosphor Plate commissioning to 7000V 22/11/2013 Launch Tests to investigate the anomaly of the TII image 05/2016 Start of scrubbings 2014 2015 2016 2017

TII datasets 2 Hz Cross-track flow datasets 16 Hz (Courtesy of J. Burchill)

EFI cross-track flow data validation Why cross-track flow? All the measurements are based on the so-called signal angle, so they are species-independent. Within this context the Origin location (x-not, y-not) is the only parameters we need to find out in the signal angle. Hypothesis of stationarity works. Day-side equatorial plasma flows show evidence for westward ion drift of order 40 m/s  Consistent with statistical diurnal measurements by ISR Known ion drift associated with plasma corotation and satellite yaw enable calibration of TII H detector Vertical detector signal modulation is not sufficient for its calibration  Is it possible to modulate satellite pitch angle on a regular basis? (from A. Kouznetsov’s talk)

Preparing TII measurements L1b Prototype is complex to calibrate Derive cross-track flows from flow angles, assuming 7.6 km/s ram speeds as is done for some products for DE2, DMSP Use reference frame co-rotating with Earth, remove flows arising from yaw and pitch motions Calibrate cross-track horizontal flow against (co-rotation + yaw) signal Exclude noisy data Send latest measurements to ESA Expert judgment needed

Validation of TII E field data The optimization procedure consists in minimising the difference between electric conductances computed with SECS and a proxy derived from Robinson formula: Provides better results when the ionospheric current (and its magnetic effect at Swarm) is stronger. Needs to be tuned to the latitudinal range of the ionospheric current. When swA and swB are longitudinally ‘conjugate’ and the event is roughly 1D, both the 2D and 1D optimization provide roughly similar results (as they should). For 1D events, often observed in the auroral region: Optimization of Ect-V looks promising, though in general will depend on the geometry. Once again, the optimization should be tuned to the latitude range of Jion. The tilt, q, of the 1D aurora, with respect to the E-V direction, matters. Move the parametrization, (a, b) and (c, d), to the TII system, and perhaps apply it to v instead of E. Parametrize also the possible tilt, q, of the 1D aurora. Analysis of more events, ideally assisted by optical data. (courtesy of O. Marghitu)

Equatorial plasma depletions as observed by Swarm Correlation E and B field Study of the direction of electromagnetic energy flux associated with equatorial plasma depletions. Asymmetry in Poynting flux to be caused probably by an asymmetry between the hemispheres on the ionospheric conductivity (Rodriguez-Zuluaga et al 2017)

Recommendation Different TII measurements (datasets) under study. Since for example in 2016 only few EPD had EFI TIICT (19%), a recommendation to take into account the occurrence of equatorial plasma depletions.

A comparison of Swarm cross-track ion-drifts and SuperDARN line-of-sight velocities Since Swarm EFI gives better quality data for the cross-track drift component, a comparison is performed with paths that are almost perpendicular to one or several SuperDARN radar beams. Basically agreement consistent with the analysis by Fiori et al. (2016) based on Su perDARN statistical model. Advantage of combining data as for example in convection reversal boundary. (Courtesy of A. Koustov) Year 2016

LP overview and status Longterm behaviour, no signs of degradation or erosion Solar activity presently helps against the 0-tracking failures which now minimized Signs of effects from different surfaces, TiN (Titanium Nitride) vs gold-plated “Samegain” of probes provides opportunities for investigations, perhaps new data products

LP datasets Provisional LP data (/Advanced folder) Preliminary data provided by IRF whose main products are electron density and temperature and s/c potential. Coverage: all s/c, BOM – 17/07/2015 Operational LP data (/Level1b folder) Data from PLASMA operational chain produced at ESA PDGS. Coverage: 18/07/2015 - present The two datasets are interoperable. The LP provisional dataset is still kept for providing users data before 18/07/2015. This dataset will be dismissed as soon as a complete reprocessing of operational plasma data will be executed. 2 Hz Langmuir Probe Extended Dataset: electron temperature and spacecraft potential derived from the two probes separately (not blended). To add also the currrents, complex admittances and bias 1/128 Hz Sweep mode dataset: to add also currents 16 Hz Faceplate currents plasma density: improved version recently published with coordinates correction, “re-sizing” of FP area. Coverage 02/10/2014 – 08/07/2017 for all s/c.

LP data validation Several detailed studies on LP parameters validation were performed.

LP validation of Ne: comparing LP and faceplate The comparison with the empirical ISO-standard model IRI-2016 suggests that Swarm Ne values are too low by about 10%-15% The comparison of Faceplate ion density estimations with the LP measurements seems to pinpoint as well on the effect of variable effective (or reduced) ion masses The assumption of a pure O+ (amu = 16) ion plasma is certainly wrong, in particular for nighttime conditions, and should be reconsidered; estimates of real ion composition (meff) appears to be feasible and worth pursuing meff = ( ∑i=1,n pi / mi ) -1 with ∑i=1,n pi = 1 50% [O+] & 50% [H+] => meff = 1.882 95% [O+] & 5% [N+] => meff = 15.886 (courtesy of M. Förster)

Mass composition and ion flow estimates Swarm faceplate current and Langmuir probe ion admittance measurements can be combined to solve for any two of ‘ion ram speed’, ‘ion density’, and ‘effective ion mass’ under the OML Langmuir probe approximation (assuming vi,ram = 7.6 km/s and mi = 16 a.m.u). Resulting ion ram speed estimates show evidence of geophysical flow variability. Comparison with TII flows provides an initial validation of the LP-Faceplate technique. Effective ion mass estimates at low-latitude are of right order of magnitude. Various error sources are likely present, including effect of minor ions on the density (>10%) and flow estimates (>1 km/s).

Validation of LP electron temperature on the basis of the IRI model Operational dataset (but can be applied also to the other datasets), all s/c Differences between data and IRI depend on local time and latitude (probably Ne dependent) Corrected Te values using spherical harmonics up to grade 6 (available on request emailing to vtr@ufa.cas.cz) Comparisons with Jicamarca data, Swarm results (Stolle et al., JGR 2011) and numerical simulation using the FLIP model show a substantial improvement with respect to original Te OPER data (some differences during daytime at high Ne density region still remain) (courtesy of V. Truhlik)

LP validation (Ne and Te) with ISR and others Independent measurements (ISRs, Ionosondes and Radio occultation) indicate that on average, Swarm LP plasma frequencies (densities) are systematically underestimated by ~10 % (18 %) for densities > 105 cm-3 Electron temperatures are systematically overestimated by ~350 K (18 %) for high-gain LPs and ~700 K (40 %) for the low-gain LPs. The correlations of both density and temperature from Swarm with data from other measurements are usually very high (0.83-0.99). The ISR-based corrections (initial comparison and calibration) of Swarm LP densities and temperatures removes systematic biases. The other measurements are used for validation. (courtesy of L. Lomidze)

ISRs Ionosondes initial initial COSMIC initial (courtesy of L. Lomidze) ISRs Corrected initial Ionosondes initial ISR-corrected COSMIC initial ISR-corrected

Te validation Low-gain LP vs high-gain LP vs IRI The data from the high-gain LPs are more accurate and reliable. Zonally and seasonally averaged electron temperatures agree significantly better with the corresponding IRI-2016 values, globally after the ISR-based corrections. (courtesy of L. Lomidze) 25°≤|Mlat|≤50° initial ISR-corrected low-gain high-gain IRI-2016

Summary of open points and recommendations Continue operations (scrubbing, commissioning, etc) on TII to improve image anomaly, keeping S/C C for tests LP probes both in high gain now as a test on S/C C but then as a normal mode of operation also on S/C A and B Validation still on going with several interesting results; open point discrepancies with different datasets still under study; study on the estimate of real ion composition; After the full reprocessing of L1b operational plasma data, dismission of preliminary LP dataset Extended 2 Hz LP dataset to be updated asap with other parameters (currents, complex admittances and bias) and extended in time Also sweep mode to be updated with currents but with a lower priority Continue updating TII cross track ion flow dataset, extending also to Swarm C and introducing likely also ion temperature in the next future