-
Notifications
You must be signed in to change notification settings - Fork 0
Description
Overview
Replace phenomenological response function in validation.py with exact KRMHD dispersion relation to enable quantitative FDT validation at 10% accuracy level.
Current Status
PR #66 implemented exact plasma physics special functions (plasma dispersion function Z(ζ), modified Bessel functions I_m(b)) and correct power laws (m^(-3/2), m^(-1/2)). However, the kinetic_response_function() uses a phenomenological approximation:
response = kinetic_factor / (1.0 + abs(zeta)**2) # PhenomenologicalThis provides qualitatively correct behavior but lacks:
- Exact normalization from full dispersion relation
- Proper k⊥ρ_s dependence in resonance width
- FLR corrections to susceptibility
Required Implementation
Implement exact KRMHD dispersion relation from:
- Howes et al. (2006) ApJ 651:590, Equations 14-15
- Thesis Equation 3.37 (exact phase mixing spectrum)
The full dispersion relation has the form:
D(k,ω) = 1 - χ(k,ω) = 0
where χ(k,ω) includes:
- Plasma dispersion function Z(ζ) for Landau resonance ✅ (already implemented)
- Modified Bessel functions for FLR effects ✅ (already implemented)
- Proper k⊥ρ_s dependence in susceptibility
- Exact normalization factors
Algorithm
- Solve dispersion relation D(k,ω) = 0 for given k
- Compute linear susceptibility χ(k,ω) from solution
- Use |χ|² in spectrum formula instead of phenomenological factor
Success Criteria
- Quantitative agreement: Numerical FDT spectrum matches analytical within 10% for all m
- Physics validation:
- Correct Landau damping rate vs analytical theory
- Proper FLR corrections at k⊥ρ_s ~ 1
- Accurate critical moment m_crit vs k∥v_th/ν
- Benchmarking: Match thesis Figures 3.1, 3.3, B.1
References
Primary sources:
- Howes, G. G. et al. (2006) ApJ 651:590 - KRMHD dispersion relation (Eq. 14-15)
- Thesis Chapter 3 - Analytical FDT theory (Eq. 3.37, 3.58)
- Thesis Figs 3.1, 3.3, B.1 - Benchmark comparisons
Background:
- Schekochihin, A. A. et al. (2009) ApJS 182:310 - Kinetic cascades
- Adkins & Schekochihin (2017) arXiv:1709.03203 - Phase mixing power laws
Implementation Notes
Current infrastructure (ready to use):
- ✅ Plasma dispersion function:
plasma_dispersion_function(zeta) - ✅ Plasma dispersion derivative:
plasma_dispersion_derivative(zeta) - ✅ Modified Bessel ratios:
modified_bessel_ratio(m, x) - ✅ FLR correction factors:
flr_correction_factor(m, k_perp, rho_s) - ✅ 15 unit tests validating all special functions
New work required:
- Implement full KRMHD dispersion relation D(k,ω)
- Numerical solver for D(k,ω) = 0 (root finding)
- Extract linear susceptibility χ(k,ω) from solution
- Replace phenomenological response with exact χ
- Validate against thesis benchmarks
Complexity estimate:
- Medium-high (requires root finding of complex dispersion relation)
- Essential for production validation studies
- Well-defined physics problem with clear benchmarks
Testing Strategy
-
Unit tests: Verify dispersion relation solver
- Test k⊥→0 (fluid limit): should recover MHD dispersion
- Test k⊥ρ_s~1 (kinetic regime): check FLR corrections
- Test collisionless vs collisional limits
-
Integration tests: Compare with thesis figures
- Fig 3.1: Hermite spectrum vs m
- Fig 3.3: Phase mixing/unmixing regimes
- Fig B.1: Detailed benchmarks
-
Quantitative validation: Enforce 10% criterion
- For each k-mode, verify |numerical - analytical| / analytical < 0.1
- Test multiple k∥/k⊥ ratios
- Test multiple collision frequencies ν
Priority
Medium-High - Required for production validation studies
Current status: Qualitative FDT validation works (PR #66)
Target status: Quantitative FDT validation at 10% level
Related
- PR Implement exact FDT analytical expressions from thesis #66: Exact FDT expressions (plasma dispersion + Bessel functions) ✅ merged
- Issue Step 22: Kinetic Fluctuation-Dissipation Validation #27: Kinetic FDT validation (infrastructure) ✅ complete
- Thesis validation requirement: "10% agreement on normalized spectrum"
Notes
Per thesis: "The dotted lines in Fig. 3.3 are not fits to the numerical spectra, but are the exact expressions"
This is the final piece needed for production-quality FDT validation!