Magnetosphere-Ionosphere-Thermosphere Coupling During Storms and Substorms Bill Lotko Bin Zhang Oliver Brambles Sheng Xi John Lyon Tian Luo Roger Varney Jeremy Ouellette Mike Wiltberger 2 3 4 CEDAR: Storms and Substorms w/o Borders 1
Why (or when do we need to) worry about the complications of SW-M-I-T coupling? M, I and T are especially interactive for strong SW driving Model predictions don t do well w/o coupling Utility depends on the fidelity of prediction: Space weather Understanding is demonstrated by prediction The coupled M-I-T system is equisitely complex and interesting 2
What might you learn from this tutorial? (or be reassured you that what you once thought was true is still true) Coupling agents Pathways (coupled) and feedback Electromagnetic Material Insights into M-I-T coupling Coupling during storms (with data-model comparisons) 3
Winter, high driving Agents of M-I-T Coupling 135 < IMF < 225 Winter, high driving 4
Agents of M-I-T Coupling Winter, high driving Conductance? 5
1. Electromagnetic Pathways of M-I-T Interaction Ionospheric Ohm s law, electrostatic condition, current continuity Current Find j source i j cos Σ i i Fejer 1953 Voltage Find source i Spatial distribution of Σ determines i for given j i and vice-versa 2. Material transport 6
O + acceleration Global M-I-T Interactions (active periods) O + O + loss tail reconnection magnetosphere inflation SW-MIT interaction plasma sheet ring current mantle cleft ion fountain lobe Feedback? Enhanced convection & entrainment of dayside O + into polar circulation auroral O + ion outflow Regions of enhanced Poynting fluxes, soft electron precipitation, structured FACs Liu et al. 05 Joule heating, upwelling, etc. Foster et al. 05 needs attention in system context 7
Empirical convection: Increasing strength of SW driving Weimer model Dungey CW rotation of the convection pattern Cycle when viewed from above NP Stronger Magnetopause driving: location: Convection v 2 is sw faster = 4B(r) and 2 /2 extends 0 (Better: to Shue lower et latitudes al., 1998) Excess flux circulation in dusk cell Harang region Weimer-2005 model 8
Empirical convection: Effect of IMF B y Weimer model Channelized fast flow NH cusp, PC pulled duskward No mirror symmetry with change in sign of B y (Heppner 1972) NH cusp, PC pulled dawnward Over flux circulation reverses: Now higher in dawn cell 9
Effect of season/dipole tilt Weimer model SUMMER: Rotation moderated, over flux circulation exacerbated Weimer-2005 model 10
Effect of EUV Hall conductance gradient BATSRUS global simulation Atkinson and Hutchison 1978 Ridley et al. 2004 CW rotation More flux circulates in dusk cell CONCLUSION Ionosphere polarizes so as to maintain J bˆ E 0 H H 11
Effect of EUV Hall conductance gradient LFM global simulation Perpendicular Velocity in Equatorial Plane Reconnection Rate Along the X-line x line mapped zero potential 1-HOUR AVERAGE STATES 0 2 hr 4 nt 4 nt 12
Effect of combined EUV and auroral Hall conductance gradient LFM global simulation 1-HOUR AVERAGE STATES 0 2 hr 4 nt 4 nt (Lotko et al. 2014) 13
Pre/Post-Midnight Asymmetry in Plasmasheet Fast Flows, Mass and Flux Transport WIND, 17 Perigee Passes, 1995-97 Raj et al. 2002 Events selected for V > 250 km s and > 0.5 (neutral sheet) 14
Asymmetries in poleward boundary intensifications and Alfvénic aurora Nishimura et al 2010 15
Storm and substorm processes 24 Aug 2005 CME Storm Initial phase: 06:00 09:00 UT B z small B y ~ 20 nt Kp ~ 3-6 The B y -dominant time period has been studied by Crowley et al. [2010] using TIME-GCM. Results show Joule heating is important in enhancing the F- region neutral density. Main phase: 09:00 16:00 B z 40 nt B y 40 nt Kp 9, Dst = -184 nt at 1200 UT 16
nt 50 0 Dst 22 23 24 25 26 27 28-50 -100-150 -200 10 9 8 7 6 5 4 3 2 1 0 Kp August 2005 August 2005 22 23 24 25 26 27 28 Weimer disclaimer: Model works best for B y and B z < 15 nt. 17
Coupled M-I-T (CMIT) model j cos Σ i i 18
19
Zhang et al. 2012 20
Change in Thermospheric Density due to Soft Electron Precipitation Difference between CMIT simulations w/ and w/o soft electron precipitation (BBE, cusp) Difference between CHAMP accelerometer measurements and MSIS90 model results Comparisons at 400 km altitude. CHAMP data are averages for 2002 for intervals of Kp = 0 2. CMIT results are a 1-hour averages for V sw = 400 km/s, n sw = 5 cm -3, IMF B z = 5 nt, F 10.7 = 150. 21
Standard CMIT simulation for the storm Neutral mass density 10-12 g/m 3 Zhang et al. 2014 CMIT tracks CHAMP reasonably well for weak driving (0600 0900 UT) CMIT overestimates ( x2) the CHAMP mass density during the main phase (0900 1600 UT) Question: What s missing in the model during the main-phase simulation? Plasmaspheric effects? Ionospheric outflows?? 22
spheric plumes in test MFLFM simulations. Figure.7 shows the formation of a plasmaspheri sunward surge and plume in an MFLFM simulation with a stationary Gallagher plasmaspher model superposed on a magnetospheric configuration driven by IMF B Effects of plasmaspheric plumes on dayside reconnection z at +5 nt. Th superposed configuration was then spun-up with corotation for 4 hours with the same IMF. Th IMF B z was then dropped to -5 nt for 3 hours and subsequently raised to +5 nt for 3 hours. Th development Plasma of a of sunward plasmaspheric surge (panel origin b) and is observed a rotated in plasmaspheric the dayside plume reconnection at the end of the 3 hour northward region [Borovsky interval (panel and Denton, d) is evident 2006; in Walsh the simulation. et al. 2014] In order to study long-lastin plumes, refilling of the plasmasphere will be enabled initially through simple empirica volumetric To refilling what extent rates. does the plasmasphere influence dayside reconnection? plume 300, 30, 3 cm -3 contours after 4-hr spin up at + 5 nt 1 hr after IMF flips 5 nt end of IMF at 5 nt 1.5 hr after IMF flips + 5 nt The dayside reconnection rate is smaller in a multi-fluid global magnetosphere simulation when plasmaspheric H + is included. Using the causally-driven mass-loading models, event simulations will be performed t Does plasmaspheric H investigate the effects of mass + influence the stormtime F-region neutral density? loading for realistic solar wind conditions in order to compare t published results in literature. 23
Brambles et al. 2011 24
Effects of ionospheric O + outflow on stormtime substorms Observations and modeling studies show that outflows of ionospheric O + are important in stormtime solar wind-magnetosphere-ionosphere coupling, especially during CME-driven storms exhibiting sawtooth oscillations. Note: Simulated onsets (with outflow) occur but are delayed 1.5 hr relative to observed onsets. Do O + outflows influence the stormtime F-region neutral density? 25
Controlled Simulation Experiments CMIT with: Gallagher et al. [1988] statistical H + plasmasphere initialized at 0:00 UT 24 Aug 2005 but not sustained. Two types of O + outflow Fixed outflow flux: No causal regulation 26
Simulated F-region Neutral Density Compared to CHAMP 10-12 g/m 3 Better agreement when auroral O + is included in CMIT CHAMP Baseline Gallagher plasmasphere Auroral O + CMIT PW O + Zhang et al. 2014 27
Orbit-Averaged Neutral Density Compared to CHAMP Baseline Gallagher plasmasphere Auroral O + PW O + Auroral O + x2 CHAMP CMIT 10-12 g/m 3 Zhang et al. 2014 28
Effects of O + on M-I Coupling Plasmaspheric H + : Little effect on CPCP, fieldaligned current Polar wind O + : Reduces CPCP Auroral O + outflow: Reduces CPCP, increases ring current intensity (but not enough and not sustained in these simulations) kv nt MA GW 8000 Baseline Gallagher plasmasphere Auroral O + PW O + Auroral O + x2 Actual Dst DST Baseline Gallagher plasmasphere Auroral O + PW O + Hemispheric power is similar in all four runs between 10-11 UT but with different polar cap distributions. 4000 Zhang et al. 2014 29
Polar Wind O+ Effects cont d Region-2 currents are larger when auroral O+ outflow is included higher integrated current Less Joule heating in polar cap with more R1-R2 current closure PJ, mw/m2 Neutral temperature and density at 400 km altitude are lower when auroral O+ outflow is included N, 10-12 kg/m3 461 kv i, kv 512 kv kv 512 AURORAL O+ 12.0 MA 12.1 MA 14.2 MA 6750 GW 6470 GW 6600 GW J, A/m2 CPCP is smaller when O+ outflow is included in the simulation 453 kv BASE Zhang et al. 2014 30
Key Points: Auroral precipitation Increases meridional gradient in E-region conductivity Ionosphere polarizes at the gradient Exacerbates dawn-dusk asymmetry in ionospheric convection plasmasheet fast flows Why does the M-I system maintain nearly divergence-free Hall currents? 31
Key Points: Soft electron precipitation Produced by direct-entry (cusp) and conversion of Alfvén wave power to fieldaligned electrons (cusp and nightside convection throat) Enhances conductivity in the bottomside F-region Joule heating is enhanced there neutral mass density is elevated at CHAMP altitude (but it increases too much) 32
Key Points: O + ionospheric outflows Lowers reconnection rate (dayside and nightside Lower CPCP Slower convection Less Joule heating, esp. in polar cap Auroral outflows have greatest impact Do ionospheric outflows directly affect the neutral gas and vice-versa? 33