Forbush Decrease Johan von Forstner - - PowerPoint PPT Presentation

SLIDE 1

Forbush Decrease

封莉 紫金山天文台 郭静楠、Johan von Forstner 德国基尔大学 顾斌、倪素兰 南京信息工程大学

SLIDE 2

Outline

• 1. Solar modulation of cosmic ray intensity
• 2. Forbush decrease

introduction to CME and its driven shock, and the related Forbush decrease Observational analyses FD simulations FD prediction

• 3. Sun shadow & CME shadow

magnetic field modeling from line-of-sight and vector photosphere magnetic field MHD simulation: e.g. ENLIL model, NSSC COIN-TVD model

SLIDE 3

Solar modulation

Solar cycle: 11/22 years Solar rotation: 27 days Earth rotation: diurnal

Phi: solar modulation parameter Phi (MV): Solar modulation parameter from ACE Cosmic Ray Isotope Spectrometer (CRIS) minimum maximum

SLIDE 4

“Recurrent” FD: recur with the solar rotation period and are associated with corotating interaction regions （27-day variation) “non- recurrent” FD: caused by the passage of transient solar wind structures (CME and its driven shock)

Forbush decrease

FD characteristics:

Rapid decrease of the GCR flux within a few hours to 1 or 2 days, Gradual recovery in the coming a few days or 10 days

SLIDE 5

Two-step Forbush Decrease

610 I.G. Richardson, H.V. Cane Fig ure1 Schematic of an interplanetary coronal mass ejection driving a shock ahead of it and the associated variations in the galactic cosmic ray intensity along trajectories that do (A) or do not (B) encounter the ICME (adapted from Cane, 2000 and Zurbuchen and Richardson, 2006).

melmair (1937) using ionization chambers. They were later shown, using neutron monitors (NMs), to originate in the interplanetary medium (Simpson, 1954) and to be of two types: “Recurrent”, which recur with the solar rotation period and are associated with corotating high-speed streams (for a review see Richardson, 2004 and references therein), and “non- recurrent”, caused by the passage of transient solar wind structures associated with coronal mass ejections at the Sun. The focus of this paper is the latter class of events, often termed “Forbush decreases” (FDs), though this term is also used by some researchers to refer to recurrent decreases. The properties and interplanetary drivers of FDs have been reviewed by Cane (2000). Figure 1 shows a schematic of a fast interplanetary coronal mass ejection (ICME), the man- ifestation in the solar wind of a coronal mass ejection at the Sun such as may be observed by coronagraphs, driving a shock ahead of it. Two processes may contribute to the decrease in GCR intensity (see, e.g ., Barnden, 1973a, 1973b; Wibberenz et al., 1998). The first is a decrease of the GCR radial diffusion coefficient in the turbulent “sheath” between the shock front and leading edge of the ICME. The resulting intensity-time profile is a linear decline during sheath passage followed by a recovery (cf., Figure 2 of Wibberenz et al., 1998). The second process arises from the at least partially closed magnetic configuration of ICMEs as evidenced for example by the presence of bi-directional suprathermal electron flows in many ICMEs suggesting that magnetic field lines are rooted at the Sun at both ends (see, e.g ., Gosling et al., 1987; Shodhan et al., 2000). GCRs enter the interior of the ICME as it moves away from the Sun by, for example, cross-field diffusion and gradient and curvature drifts (see, e.g ., Krittinatham and Ruffolo, 2009; Kubo and Shimazu, 2010) such

CME tracing shock tracing SOHO/LASCO SOHO/LASCO STEREO/HI1

SLIDE 6

ICME propagation from 1 AU to 1.5AU

Von Forstner et al., 2017, in prep

SLIDE 7

SLIDE 8

Zhao et al. 2016

FD for different CME events

SLIDE 9

Key parameters of FD : amplitude,recovery time as a function of median energy

2012-Mar-07 event

Recovery time: Median energy:

LHAASO: extend the FD study to higher energies

SLIDE 10

2012-Mar-07 2015-June-21

Normal case abnormal case

SLIDE 11

SLIDE 12

Earth Mars MSL main propagation direction Messenger Venus STA STB

2012-Mar-07 event Projections of the propagation of the 3D shock on the solar equatorial plane

Feng et al., 2012, 2013 ab, 2015ab, 2017

SLIDE 13

1D Simulation of FD

1D case

Ni et al., 2017, in prep

Diffusion coefficient and its scaleμ

convection diffusion Energy change

SLIDE 14

1D Simulation of FD

μis inversely proportional to V

SLIDE 15

FD prediction- shock precursor

Nagashima et al. 1992

Decrease of cosmic ray intensity in front the shock; have nearly the same rigidity spectrum as the FD.

SLIDE 16

Decrease of cosmic ray intensity in front the shock- shock precursor

796 R U F F O L O V ol . 515

F I

• G. 9.

È P hase- space di st ri but i

• n ofpart

i cl es asa f unct i

• n ofk nearan
• bl

i que shock w i t h f

• r q \ 1 (

sol i d l i nes)and q \ 1. 5 ( dashed t an h1\ 4 l i nes) at z\ 0. 05j ( upst ream , t hi ck l i nes) and z\ [ 0. 05j ( dow nst ream , t hi n l i nes) .

“ “ l

• ss-

coneÏ Ïe†ect ( B i eber & E venson 1997)f

• r t

he case of G al act i c cosm i c- ray ( G C R )depl et i

• n athi

gh k upst ream of an i nt erpl anet ary shock, w hi ch i s due t

• t

he pauci t y of G C R s com i ng f rom dow nst ream ( see ° 3. 2) .T he sam e e†ect

• ccurs here because t

he accel erat i

• n of part

i cl es com i ng f rom dow nst ream i sw eakert han f

• rpart

i cl es reÑect ed f rom upst ream . T he great est accel erat i

• n occurs f
• r part

i cl es reÑect ed w i t h t he great estchange i n pi t ch angl e ( see F i g.2) , i . e. , f

• r k sl

i ght l y bel

• w 0.
• 85. Si

nce st ronger accel erat i

• n

i m pl i es t hat f i s advect ed f rom l

• w er m om ent

a, and t he part i cl e spect rum i ncreases w i t h decreasi ng m om ent um i n t hi s case, t he st rongest accel erat i

• n corresponds t
• t

he great esti ncrease i n f . In t he dow nst ream regi

• n,part

i cl es are redi st ri but ed i n pi t ch angl e because of changes i n pi t ch angl e as part i cl es cross t he shock ;t he average Ñux al so i ncreases sl i ght l y due t

• accel

erat i

• n.Iti

s w ort h not i ng t hat f

• r a hi

ghl y obl i que shock, m ost part i cl es com i ng f rom upst ream are i n f act reÑect ed,i . e. ,w hen

• r i

n t he case ofa

• k o\ (

1 [ B 1/ B 2) 1@ 2, st rong,hi ghl y obl i que shock,f

• rpi

t ch angl es m ore t han 30¡ f rom t he m agnet i c Ðel d di rect i

• n. A not

her f eat ure of F i gures 8 and 9 i s t he sharp gradi ent i n fat k \ 0 f

• r t

he case of q \ 1. 5.F or t hi s f

• rm of t

he pi t ch angl e di †usi

• n

coef f i ci ent , r ( k)\ A ok o 0. 5( 1 [ k2) t ends t

• zero as k ] 0.

Si nce t he k Ñux, i s sl

• w l

y varyi ng i n a Sk\ [ ( r / 2) ( LF / Lk) , near- equi l i bri um si t uat i

• n, t

he vani shi ng di †usi

• n coef

f i

• ci

ent at k \ 0 i s abl e t

• sust

ai n an i nÐni t e gradi ent i n F at t hatval ue. W e bel i eve t hatt hi s behavi

• r offas a f

unct i

• n ofz and k

i s notan art i f actoft he assum pt i

• n ofa pow er-

l aw m om en- t um dependence because w hen an ext ra ani sot ropy w as art i Ðci al l y added, t he k and z dependence ( i ncl udi ng t he j um p at z\ 0)w as not si gni Ðcant l y a†ect ed ;t hi s w as al so t he case f

• r paral

l elshocks.O n t he ot her hand,com put ed val ues ofc are st rongl y a†ect ed by t he pow er- l aw assum p- t i

• n, so t

hat t hi s code i n i t s present f

• rm

i s essent i al l y unabl e t

• det

erm i ne c.T he error i n c w as m uch w eaker f

• r

v/ c\ 0. 5 t han f

• rv/

c\ 0. 1 ( because oft he l

• w er

rat i

• )

. u1/ v A s an exam pl e,f

• r q \ 1 t

he c val ues requi red f

• r a st

eady st at e w i t h v\ 0. 5c w ere 1. 965 and 1. 952 f

• r

and t an h1\ 1 4,respect i vel y,w hereas w i t h v\ 0. 1c t hey w ere 1. 985 and 1. 787,respect i vel y.O t herw i se,t he resul t s regardi ng f( k,z) f

• r v/

c\ 0. 1 w ere qual i t at i vel y si m i l ar t

• t

hose show n i n F i gures 7 t

• 9 f
• r v/

c\ 0. 5, w i t h m uch st ronger ani so- t ropi esand j um psi n att he shock. S fTk 3.

• 2. P recursorsofF orbush D ecreases

T o dem onst rat e t he versat i l i t y oft hi sm et hod,w e appl y i t t

• m odelF orbush decreasesofG al

act i c cosm i c rays( G C R s) as an i nt erpl anet ary shock passes t he E art h ( F orbush 1938; B erry & H ess 1942;F orbush & L ange 1942) , w hi ch rep- resent a t ransi ent phenom enon i nst ead of a st eady st at e. G round- based neut ron m oni t

• rs m easure secondary neu-

t rons f rom t he i m pact ofrel at i vi st i c,pri m ary charged par- t i cl es,m ai nl y prot

• ns,on t

he upper at m osphere.O w i ng t

• sel

ect i ve deÑect i

• n by t

he E art hÏ s m agnet i c Ðel d, neut ron m oni t

• r observat

i

• ns are sensi

t i ve t

• pri

m ary cosm i c rays f rom speci Ðc di rect i

• ns i

n space, and t he w orl dw i de net w ork ofneut ron m oni t

• rs provi

des det ai l ed i nf

• rm at

i

• n
• n t

hei r pi t ch angl e di st ri but i

• n,sensi

t i ve t

• vari

at i

• ns on

t he order of0. 1% .P recursors t

• F orbush decreases are of

pract i cal i nt erest as possi bl e predi ct

• rs of space w eat

her e†ect s on t he E art h, such as sat el l i t e f ai l ures, radi

• f

ade-

• ut

s,pow er out ages,et c. ,severalhours or even days bef

• re

t he passage of a m aj

• r i

nt erpl anet ary shock. Several anal yses of neut ron m oni t

• r observat

i

• ns have i

ndi cat ed t w o t ypes of precursors t

• F orbush decreases: (

1) an enhanced di urnal ani sot ropy of G C R s, w i t h an excess of part i cl es t ravel i ng t

• w ard t

he Sun al

• ng t

he i nt erpl anet ary m agnet i c Ðel d,and ( 2)a deÐci t ofG C R s i n a “ “ l

• ss cone,

Ï Ï i . e. , al

• ng a narrow range of pi

t ch angl es di rect ed nearl y al

• ng t

he i nt erpl anet ary m agnet i c Ðel d aw ay f rom t he Sun ( N agashi m a etal .1992;N agashi m a,F uj i m ot

• ,& M ori

shi t a 1994; Sakaki bara et al . 1995; B el

• v et al

. 1995; C ane, R i chardson, & von R osenvi nge 1996; B i eber & E venson 1997) . W e m odel t he F orbush decrease i n a rat her i deal i zed m anner, assum i ng t he conÐgurat i

• n of F i

gure 1 and negl ect i ng part i cl e dri f t s ( N i shi da 1983; K adokura & N i shi da 1986) , spat i al dependence of t he scat t eri ng m ean f ree pat h,shock curvat ure, t he Ðni t e spat i al ext ent of t he i nt erpl anet ary shock, adi abat i c f

• cusi

ng, and adi abat i c decel erat i

• n.N evert

hel ess,w e can expl ai n t he basi c f eat ures

• ft

he observed precursors,veri f yi ng t hei r i nt erpret at i

• n i

n t erm s of part i cl e t ransport i n t he vi ci ni t y of an obl i que shock. T he si m ul at i

• n condi

t i

• ns w ere i

nspi red by t he dram at i c C M E event of 1997 A pri l 7,w hi ch arri ved near E art h on A pri l 10È11,and f

• r w hi

ch a possi bl e l

• ss-

cone f eat ure i s i dent i Ðed by B i eber & E venson ( 1997) . In t hi s case, t he t ravelt i m e of3 days i ndi cat es a shock speed ofB 600 km s~ 1 or onl y 200 km s~ 1 f ast er t han t he ( t ypi cal )sol ar w i nd speed.A ssum i ng t hat t he shock norm al i s radi al ,w e t ake t he upst ream shock- Ðel d angl e t

• be t

he t ypi cal “ “ garden- hoseÏ Ïangl e of 45¡ and as bef

• re w e assum e

( t an h1\ 1) , km s~ 1. T hus w e Ðnd km s~ 1, us 1\ uA1\ 50 * un\ 133 and w hi ch i n t urn i m pl i es t hat t an h2\ 3. 20, B 2/ B 1\ 2. 37, part i cl es crossi ng t he shock f rom dow nst ream have pi t ch angl es al i gned w i t h t he m agnet i c Ðel d t

• w i

t hi n 40¡ ( k [ 0. 76) . W e used q \ 1. 5, w hi ch adequat el y descri bes i nt erpl anet ary scat t eri ng ( B i eber et al . 1986) . F or t he upst ream boundary condi t i

• n,w e speci

f y a const ant ( see F u ° 2. 5) ,and t he i ni t i alcondi t i

• n set

sF t

• t

hatconst anti n t he upst ream regi

• n and t
• zero i

n t he dow nst ream regi

• n.W e

used * k \ 2/ 45 ( 45 k- gri d poi nt s) and * z/ j \ 0. 05,w here j \ 0. 3 A U .W e consi dered v\ 0. 75c,correspondi ng t

• a

N o. 2, 1999 P A R T IC L E S N E A R A N O B L IQ U E SH O C K 795

F I

• G. 7.

È Spat i al dependence of t he pi t ch angl eÈaveraged phase- space di st ri but i

• n f

unct i

• n,

f

• r st

eady st at e part i cl e accel erat i

• n near a

S fTk, shock ( at z\ 0) f

• r vari
• us val

ues of t he t angent of t he angl e t an h1, bet w een t he m agnet i c Ðel d and t he shock norm al ,f

• rq \ 1 (

sol i d l i nes)and 1. 5 ( dot t ed l i nes) ,and f

• rv\ 0.

5c.F art herf rom t he shock, i sconst ant S fTk dow nst ream and exponent i al l y decayst

• w ard zero upst

ream .T he ordi nat e i snorm al i zed t

• t

he val ue f ardow nst ream .

m ul t i pl y F upon accel erat i

• n by a k-

dependent f act

• r t
• accountf
• r t

he hi gher ani sot ropy ofF att he l

• w erm om en-

t um f rom w hi ch a part i cl e w as accel erat ed, yi el di ng a si m i l ar F ( k,z)and c\ 2. 041.W e concl ude t hat t hi s expl a- nat i

• n can i

n f actaccountf

• r a syst

em at i c error i n c oft he

• bserved m agni

t ude and t hat c i s m ore sensi t i ve t

• t

he assum pt i

• n ofa pow er-

l aw dependence t han i st he di st ri bu- t i

• n ofpart

i cl esi n space orpi t ch angl e. T urni ng t

• obl

i que shocks, F i gure 7 show s t he spat i al dependence of f

• r

and 4 and f

• rq \ 1 and

S fTk t an h1\ 1 1. 5.In al l cases,t he di st ri but i

• n f

unct i

• n f

art her f rom t he shock i s consi st ent w i t h t he di †usi

• n approxi

m at i

• n,w i

t h const ant dow nst ream and exponent i al l y decayi ng SF Tk upst ream w i t h k \ u/ D .A conspi cuousf eat ure ofF i gure 7 i s t he j um p ( di scont i nui t y) i n at an obl i que shock ( t he S fTk Ðni t e sl

• pe i

sdue t

• t

he Ðni t e gri d spaci ng i n z) .T hi sf eat ure w as al so f

• und i

n si m ul at i

• ns by O st

row ski( 1991) ,G i esel er et al .( 1999) ,and T .N ai t

• (

1998,pri vat e com m uni cat i

• n)

. G i esel er etal .( 1999)presenta det ai l ed t heoret i caland com - put at i

• nalanal

ysi s oft hi s f eat ure,as w el las possi bl e obser- vat i

• nal si

gnat ures.W e Ðnd t hat t he j um p i s st ronger f

• r

m ore obl i que shocks and w eaker f

• r q \ 1.

5 t han f

• r q \ 1.

T he am pl i t ude oft he j um p i s on t he order ofa f ew percent f

• r such f

ast part i cl es ( v\ 0. 5c) ,and our si m ul at i

• ns i

ndi

• cat

e t hatt he j um p i s st ronger f

• r sl
• w er part

i cl es ( v\ 0. 1c) , i . e. ,a hi gheru/ v. A not her di †erence f rom t he case of a paral l el shock i s t hat f

• r obl

i que shocks, addi t i

• nal ei

genf unct i

• ns are

exci t ed i n f( k,z) near t he shock.( If one i s not suf f i ci ent l y caref ul i n t reat i ng t he boundary condi t i

• ns, as I w as not

duri ng t he i ni t i al st ages of t hi s w ork, addi t i

• nal ei

gen- f unct i

• ns are al

so exci t ed near t he boundari es; di s- cret i zat i

• n errors al

so yi el d spuri

• us ei

genf unct i

• ns near

t he shock,w hi ch becom e negl i gi bl e f

• r 95 k-

gri d poi nt s as used here. )F or al lst eady st at e si m ul at i

• ns,f(

k,z)w as con- si st entw i t h a sum ofseparabl e sol ut i

• nsofequat

i

• n (

3) .F or F i gure 8 show s t he dependence off t an h1\ 4 ( h1\ 75¡) ,

• n k and z w i

t hi n ^ 0. 8j oft he shock,and F i gure 9 show s f as a f unct i

• n ofk f
• r z\ ^ 0.

05j.( R ecal lt hatw e use k and p t

• ref

er t

• quant

i t i es i n t he l

• cal Ñui

d f ram e;t hus t hese pl

• t

s are f

• r a const

ant val ue of t he l

• cal p.A C om pt
• n-

G et t i ng t ransf

• rm at

i

• n t
• t

he shock f ram e w oul d have no not i ceabl e e†ecton our di st ri but i

• n pl
• t

s. )F or t an h1\ 1, t he resul t sw ere qual i t at i vel y si m i l ar butw i t h w eakerani so- t ropi es. In F i gure 9, w e see t hat upst ream di st ri but i

• ns (

t hi ck l i nes)i ncrease w i t h k up t

• k B 0.

7 ( w i t h a sl i ght l y st ronger ani sot ropy t han i n t he f ar upst ream regi

• n)

,and f

• r great

er k val ues,fdrops sharpl y.T he reason f

• r t

he sharp drop i s t hat gi ven our assum pt i

• n ofconservat

i

• n oft

he m agnet i c m om ent ,part i cl es w i t h

• r 0.

85 i n t hi s k [ ( 1 [ B 1/ B 2) 1@ 2, case,have com e f rom dow nst ream .A si m i l ar drop i n fhas been cal l ed a “ “ deÐci t coneÏ Ï ( N agashi m a et al . 1992) or

F I

• G. 8.

È P hase- space di st ri but i

• n ofpart

i cl es as a f unct i

• n ofk and z

( i n uni t s ofj)near an obl i que shock w i t h f

• r (

a)q \ 1 and ( b) t an h1\ 4 q \ 1. 5.N ot e t he changes i n t he pi t ch angl e di st ri but i

• n neart

he shock ( at z\ 0) .

upstream downstream

Z=0: oblique shock front position Z>0: shock upstream Z<0: shock downstream (region of FD)

Particle distribution as a function of distance to shock front and pitch angle

Loss cone or deficit cone: particles comes from the shock downstream

Ruffolo 1999

SLIDE 17

Shock Propagation Model Model the shock propagation after the shock has accelerated to a maximum value using a blast wave scenario.

Shock prediction

Venus Earth

VENUS 1AU

6Rs Feng et al. 2017, in prep

SLIDE 18

More work to be done …

Earth Mars MS L main propagation direction Messeng er Venus STA STB

• 1. Simulations along

the CME/shock flank

• 2. Compare the

simulations with the cosmic ray flux at Earth, Mars, etc.

SLIDE 19

More work to be done …

ARGO-SPT experiment: FD in different energy ranges Jia et al., 2005

• 1. More events

study

• 1. Simulations of

FD in different energy ranges

SLIDE 20

Sun shadow and CME shadow

Flux deficit Amenomori et al., 2013

SLIDE 21

Coronal magnetic field model

Extrapolation from line of sight magnetic field Zhao et al. 1995 Feng et al. 2013c Linear force free field extrapolation CSSS model

SLIDE 22

Coronal magnetic field model

Extrapolation from vector magnetic field

A&A 562, A105 (2014) Fig .3. Field lines of a) the potential field model and b) the NLFFF model around ARs 11429 and 11 430 overlaid on the AIA 171 Å image. Green and red lines represent open and closed magnetic field lines, respectively. (a) (b) Fig .4. Magnetic field line skeletons a) of the entire solar globe from the NLFFF model and b) image of the sun observed by SOHO/ LASCO C2 coronagraph at 16:33UT.

used to the standard correlation coef f i cient for scalar functions. The correlation was calculated (Schrijver et al. 2006) from Cvec =

i ui ·

ui

i |

ui|

2 i |

ui|

2 1/2 ,

(5) where ui and ui are the vectors at each grid point i. If the vector fields are identical, then Cvec = 1; if ui ⊥ui , then Cvec = 0. The degree of convergence towards a force-free and divergence-free model solution can be quantified by the integral measures of the Lorentz force and the divergence terms in the minimization func- tional in Eq. (4), which are computed over the entire solar globe. The Lf and Ld of Eq. (4) measure how well the force-free and divergence-free conditions are fulfilled, respectively. In Table 1, we provide some quantitative measures to rate the quality of

• ur reconstruction. Column 1 names the corresponding models.

Columns 2–3 show how well the force-balance and solenoidal

• Table1. Evaluation of the reconstruction quality for the potential field

and NLFFF models. Model Lf Ld Lphoto Cvec Potential 0. 000 0. 000 0. 001 1 NLFFF 0. 391 0. 697 0. 302 0. 893 No

• tes. We have used spherical grids of 300 ×450 ×900.

conditions are fulfilled for both models. Figure 5 shows how well the functional L converge to zero during iteration process. In the last column, the vector correlation shows that there is dis- agreement between the two model field solutions. The energy stored in the magnetic field as a result of a field line stressing into a nonpotential configuration has been identi- fied as the source of flare energy. Therefore, to understand the

A105, page 6 of 8

A&A 562, A105 (2014) Fig .3. Field lines of a) the potential field model and b) the NLFFF model around ARs 11429 and 11 430 overlaid on the AIA 171 Å image. Green and red lines represent open and closed magnetic field lines, respectively. (a) (b) Fig .4. Magnetic field line skeletons a) of the entire solar globe from the NLFFF model and b) image of the sun observed by SOHO/ LASCO C2 coronagraph at 16:33UT.

used to the standard correlation coef f i cient for scalar functions. The correlation was calculated (Schrijver et al. 2006) from Cvec =

i ui ·

ui

i |

ui|

2 i |

ui|

2 1/2 ,

(5) where ui and ui are the vectors at each grid point i. If the vector fields are identical, then Cvec = 1; if ui ⊥ui , then Cvec = 0. The degree of convergence towards a force-free and divergence-free model solution can be quantified by the integral measures of the Lorentz force and the divergence terms in the minimization func- tional in Eq. (4), which are computed over the entire solar globe. The Lf and Ld of Eq. (4) measure how well the force-free and divergence-free conditions are fulfilled, respectively. In Table 1, we provide some quantitative measures to rate the quality of

• ur reconstruction. Column 1 names the corresponding models.

Columns 2–3 show how well the force-balance and solenoidal

• Table1. Evaluation of the reconstruction quality for the potential field

and NLFFF models. Model Lf Ld Lphoto Cvec Potential 0. 000 0. 000 0. 001 1 NLFFF 0. 391 0. 697 0. 302 0. 893 No

• tes. We have used spherical grids of 300 ×450 ×900.

conditions are fulfilled for both models. Figure 5 shows how well the functional L converge to zero during iteration process. In the last column, the vector correlation shows that there is dis- agreement between the two model field solutions. The energy stored in the magnetic field as a result of a field line stressing into a nonpotential configuration has been identi- fied as the source of flare energy. Therefore, to understand the

A105, page 6 of 8

Tadesse et al. 2014

SLIDE 23

MHD simulations of interplanetary magnetic field

NSSC COIN-TVD model : Shen et al., 2013 Feng et al., 2017, in prep

SLIDE 24

CME shadow

LHAASO science document

SLIDE 25

MHD simulations of CME propagation

Shen et al., 2013, 2014

SLIDE 26

宇宙线传输

• Heliospheric transport of GCR is described by Parker’s

theory (Parker, 1965; Toptygin, 1985) Four basic processes：

• the diffusion of particles due to their scattering on

magnetic inhomogeneities,

• the convection of particles by out-blowing solar wind,
• adiabatic energy losses in expanding solar wind,
• drifts of particles in the magnetic field, including the

gradient-curvature drift in the regular heliospheric magnetic field, and the drift along the heliospheric current sheet, which is a thin magnetic interface between the two heliomagnetic hemispheres.

SLIDE 27

宇宙线的调制

Usoskin et al., LRSP 11 year cycle: solar activity 22 year cycle: Sharp and flat maxima. the increase of CR flux in 2009, when it was the highest ever recorded by NMs as caused by the favorable heliospheric conditions (unusually weak heliospheric magnetic field and the flat heliospheric current sheet) (McDonald et al., 2010).