Astronomy 253 — Problem Set #1 Solutions Problem #1: Geomagnetic Fields In regions not too far from the Earth (r . 10RE , where RE is the Earth radius), the geomagnetic field can be approximated as a dipole field with magnetic moment ME ≈ 8.0 × 1025 erg G−1 (in cgs units). In spherical polar coordinates (anti-)aligned with the Earth’s magnetic dipole moment, the three field components are B=− ME (2 sin λ, cos λ, 0) , r3 (1) where λ ≡ π/2 − θ is the latitude. (1). Show that the equation for a field line is r = r0 cos2 λ, and compute how field strength varies with λ along a field line. (5 pts). Solution: At location (r, θ), field line shape should satisfy dr Br dr d sin θ = = 2 cot θ ⇒ =2 ⇒ r = r0 sin2 θ = r0 cos2 λ . rdθ Bθ r sin θ (2) We can now use r0 to label individual field lines. The three field components along a field line is now given by ME B=− 3 (2 sin λ, cos λ, 0) . (3) r0 cos6 λ Thus the field strength is B(r0 , λ) = ME p 1 + 3 sin2 λ . r03 cos6 λ (4) (2). The Earth’s outer Van Allen Radiation Belt is filled with energetic particles (both electrons and ions, 0.1 − 10 MeV) that are trapped, bouncing back and forth between mirror points. The belt extends from about 3RE to 9RE , reaching its greatest intensity at about 4 − 5RE . For this problem, we consider 3.0 × 104 km (∼ 4.7RE ) as a characteristic radius of the belt at equatorial plane. [2a]. Provide a rough estimate of maximum proton/electron energy (in units of eV) above which guiding-center motion is no longer a good approximation. (5 pts) Solution: Particle gyroradius must be much smaller than the scale at which magnetic field varies (∼ RE ). Magnetic field strength at the equatorial plane is the weakest, corresponding to largest gyro-radius. At R0 = 3 × 104 km, field strength is B(R0 ) ≈ ME ≈ 3 × 10−3 G . R03 (5) Electrons are trans-relativistic, with gyro-radius rL,e = 5.2 × 10−3 RE 1 E MeV B mG −1 . (6) Protons are non-relativistic, with gyro-radius r rL,i = 0.23RE B −1 E . MeV mG (7) Thus, electron gyroradius is much smaller than RE unless their energy reaches GeV level. Proton gyroradius becomes comparable to RE at a few tens of MeV. [2b]. Particles would be lost due to collisions with neutrals if they manage to enter the Earth’s atmosphere (thickness RE ). Compute the equatorial loss cone αleq : the critical particle pitch angle in the equatorial plane below which particles would be lost. (5 pts) 2 /B = const. Together with energy conSolution: Conservation of magnetic moment demands v⊥ 2 servation, we have sin α/B = const. Given equatorial pitch angle αeq , mirror reflection occurs at latitude λm , which satisifies sin2 αeq 1 = . (8) B(R0 , λ = 0) B(R0 , λm ) For critical pitch angle (loss cone), we further require RE = R0 cos2 λm . (9) Altogether, and using the expression of B from problem (1), we obtain cos6 λm (RE /R0 )3 =p sin2 αleq = p ≈ 5.2 × 10−3 . 2 4 − 3(RE /R0 ) 1 + 3 sin λm (10) This pitch angle translates to about 4.1◦ , independent of particle energy. [2c]. For trapped particles, provide a rough estimate of the bounce period τb between mirror reflection points. (5 pts) Solution: The bounce period can be expressed as Z λm dλ ds τb = 4 , vk dλ 0 where ds is the arc length along the field line, and for dipole field we have p p ds/dλ = r 1 + 4 tan2 λ = R0 cos2 λ 1 + 4 tan2 λ . (11) (12) Parallel velocity is given by vk = v p 1 − (v⊥ /v)2 = v(1 − B/Bm )1/2 , (13) where Bm = B(R0 , λm ) if magnetic field strength at reflection point. Substituting it back, we obtain √ Z 4R0 λm cos2 λ 1 + 4 tan2 λdλ (14) τb = v 0 (1 − B/Bm )1/2 2 While this integral can not be evaluated analytically, numerical integrations show that it is of order unity and varies by within a factor of 2 for λm between 0 and 90◦ (from ∼ 0.7 to ∼ 1.4). Therefore, a reasonable approximation for τb is simply τb ≈ 4R0 /v. For non-relativistic ions, we have τb,i 4R0 ∼ ∼ 8.7 s vi E MeV −1/2 . (15) For trans-relativistic electrons, we have τb,e ∼ 4R0 ∼ 0.4 s . c (16) (3). Trapped particles will also undergo curvature and grad-B drifts (ignore the E × B drift due to the Earth’s rotation since it is negligible). Derive the combined curvature and grad-B drift velocities for electrons and ions (assuming they are non-relativistic). What is the direction of the drift? How does drift velocity depend on latitude? What are the drift timescales at the equatorial plane for electrons and ions? (10 pts) Hint: the curvature drift term in Equation (2.24) in Boyd & Sanderson is not rigorous, but is applicable when field configuration is current-free (∇ × B ≈ 0) as in a vacuum dipole field. Start from the standard expression for curvature drift given in the lecture, v c = b × vk2 (b · ∇)b/Ω, and use the following identities for unit vector b: (b · ∇)b = −b × (∇ × b) , ∇ × B = (∇B) × b + B(∇ × b) , (17) to show (b · ∇)b = ∇⊥ B/B in the current-free limit, and then combine it with the grad-B drift term. Solution: The combined grad-B and curvature drift velocities are given by 2 v⊥ µ 1 1 2 2 ∇B + vk (b · ∇)b = b × ∇B + vk (b · ∇)b . Vd = b× Ω m Ω 2B (18) We have the following identities for unit vector b: (b · ∇)b = −b × (∇ × b) , ∇ × B = (∇B) × b + B(∇ × b) . (19) Since the (vacuum) dipole field is current-free (we assume the field is unaffected by the ring current, which is generally a good approximation), or ∇ × B ≈ 0, thus (b · ∇)b ≈ b 1 ∇⊥ B × [(∇B) × b] = [∇B − b(b · ∇B)] = . B B B (20) Substituting back to (18), we obtain Vd= 2 /2 + v 2 v⊥ k (B × ∇B) = mv 2 c (1 + cos2 α)(B × ∇B) 2qB 3 ΩB 2 2Ek c = (1 − B/2Bm )(B × ∇B) qB 3 3 (21) Note that because of different sign of charge, electrons and ions drift towards opposite directions. Computing total field strength and its gradient is straightforward ME p 3M E 2 B= 3 1 + 3 cos θ, sin θ cos θ, 0 . (22) 1 + 3 cos2 θ , ∇B = √ r r4 1 + 3 cos2 θ The resulting drift velocity is vd = − 6Ek cr2 B sin θ(1 + cos2 θ) 6Ek cr02 B cos λ5 (1 + sin2 λ) (1 − ) eφ = − (1 − ) eφ . 2 2 qME 2Bm (1 + 3 cos θ) qME 2Bm (1 + 3 sin2 λ)2 (23) Therefore, particles drift in the azimuthal direction. The ions drift westward (negative drift velocity), while electrons drift eastward. Such drift motion creates a net current known as the ring current. The drift velocity depends sensitively on latitude (due to the cos5 λ factor) and hence the ring current is concentrated in the equator. At the equator, we have v eq d 6Ek cr02 B 6Ek c r0 2 Ek 2 −1 = (1 − )≈ ≈ 4.9 × 10 km s qME 2Bm qBr 4RE MeV (24) For both electrons and ions, the drift timescale is teq d ∼ 0.36 hours 4RE r0 Ek MeV −1 (25) In reality, particles only spend a fraction of time near the equatorial plane where the drift is the fastest, and (25) underestimates the drift timescale by a factor of ∼ 3. Therefore, it typically takes an hour for these energetic particles to complete one cycle of drift around the Earth. While relativistic correction is needed for electrons, the derivations above suffices for our purposes. Problem #2: Conservation of Energy The goal of this problem is to show that ideal MHD conserves energy. We will do this by finding equations governing the evolution of magnetic, internal, and kinetic energy in conservative form: ∂ (stuff) + ∇ · (flux of stuff) = (sources − sinks) ∂t (26) For reference, the equations of MHD that you may need are: ∂B = −c∇ × E ∂t c J = ∇×B 4π ∂V J×B ρ + ρ (V · ∇) V = − ∇p ∂t c V×B E+ = 0 c ∇·B = 0 ∂p + V · ∇p + γp∇ · V = 0 ∂t 4 (27) (28) (29) (30) (31) (32) 2 (a) The magnetic energy density is given by B 8π . Start from Faraday’s law and derive an equation for the time evolution of magnetic energy 2 density in terms of the Poynting flux, electric field, and ∂B ∂ B current density. Hint: 4π · ∂t = ∂t B 8π . You will need vector identities and Ampere’s law. (8 pts) Start from Faraday’s law and use the hint: ∂B = −c∇ × E ∂t B ∂B B · = · (−c∇ × E) 4π ∂t 4π ∂ B2 c = − B · (∇ × E) ∂t 8π 4π Now use ∇ · (A × B) ≡ B · (∇ × A) − A · (∇ × B) to get c ∂ B2 = − [∇ · (E × B) + E · (∇ × B)] ∂t 8π 4π Then use Ampere’s law for the last term on the RHS and rearrange to get ∂ B2 cE × B +∇· = −E · J ∂t 8π 4π p (b) The internal energy density is given by γ−1 . Start from the energy equation and derive an equation for the time evolution of the thermal energy density. You will need the vector identity f ∇ · A ≡ −A · ∇f + ∇ · (f A), but no other equations from MHD. In the conservation equation, the divergence should be taken of not just the internal energy flux but also contain the contribution from work done on/by the plasma through pressure gradients forces. (8 pts) The energy equation is ∂p + V · ∇p + γp∇ · V = 0 ∂t Use the vector identity from the problem ∂p + V · ∇p − γV · ∇p + γ∇ · (pV) = 0 ∂t ∂p + γ∇ · (pV) − (γ − 1)V · ∇p = 0 ∂t Divide by (γ − 1) and rearrange: ∂ p p +∇· V + pV = V · ∇p ∂t γ − 1 γ−1 2 (c) The kinetic energy density is given by ρV2 . Start from the momentum equation and derive an equation for the time evolution of kinetic energy density (recalling the hint from part a). You will need the continuity equation (multiplied by some factor) to get the additional terms from 5 evaluating ∂ ∂t ρV 2 2 with the product rule. You will need the ideal Ohm’s law to put it terms of a 2 Poynting flux, that A · (A · ∇A) = A · ∇ A2 , and the vector identity A · (B × C) = −(B × A) · C. (8 pts) The momentum equation is ρ ∂V + ρV · ∇V = ∂t J×B − ∇p c Take V· this equation. ∂V J×B V· ρ + ρV · ∇V = V · − ∇p ∂t c 2 V (−V × B) · J ∂ V2 + ρV · ∇ = − V · ∇p ρ ∂t 2 2 c 2 V ∂ V2 + ρV · ∇ = E · J − V · ∇p ρ ∂t 2 2 Let’s take the time derivative of ρV 2 /2 to see what cross terms we need: V 2 ∂ρ ∂ V2 ∂ ρV 2 = +ρ ∂t 2 2 ∂t ∂t 2 This suggests that we will need to multiply the continuity equation by V 2 /2. Add the continuity equation times V 2 /2: 2 V V 2 ∂ρ ∂ V2 ρ + ρV · ∇ + + ∇ · (ρV) = E · J − V · ∇p ∂t 2 2 2 ∂t 2 V ∂ V2 V 2 ∂ρ V2 + ρV · ∇ ρ + + ∇ · (ρV) = E · J − V · ∇p ∂t 2 2 ∂t 2 2 {z } {z } | | ∂ ∂t ρV 2 2 ∇· ρV 2 V 2 Combine the first and second terms, and the third and fourth terms on the LHS with the product rule: 2 ∂ ρV 2 ρV +∇· V = E · J − V · ∇p ∂t 2 2 (d) From parts (a)–(c), you should have three equations which add up to 2 ∂ B 2 ρV 2 p ρV p cE × B + + +∇· V+ V + pV + = 0, ∂t 8π 2 γ−1 2 γ−1 4π (33) indicating that total energy is conserved. For each of the equations derived in parts (a)–(c), there should be source and sink terms that show up on the right hand side that cancel when you add the equations. What do each of these source and sink terms physically represent? (6 pts) 6 • The term E · J represents the transfer between magnetic energy and kinetic energy. It is the work done by electromagnetic forces on the plasma. (Note that this is true in ideal MHD. When resistivity is considered, the additional E · J leads to Joule heating where E = J /σ, and σ is electric conductivity.) • The term V · ∇p represents the transfer between kinetic energy and internal energy. It is the work done by pressure gradients to accelerate or decelerate the plasma. 7