California Freight Cleanup → Appendix

All 57 investigations behind the study.

43 of them have full deep-dive pages, organized by the focus area they primarily serve. The remaining 14 are supporting analyses — validation rungs, cross-checks, foundational methodology — that underpin findings cited elsewhere but don’t carry visitor narrative on their own. They’re listed at the bottom with one-line summaries.

Element 1 — Emissions Inventory

See Element 1 summary.

Element 2 — Gas-Fired EGU Health Impact

See Element 2 summary.

Element 3 — Quantitative Air Quality Model Uncertainty

See Element 3 summary. The multi-fidelity GP emulator stack — ISRM through L4 MFGP — underpins the defensible uncertainty bounds used throughout the cascade.

Element 4 — Co-Benefits / Disbenefits

See Element 4 summary.

Element 5 — DAC Distributional Analysis

See Element 5 summary.

Element 6 — Ratepayer Burden / Non-Energy Impact

See Element 6 summary. Heaviest criterion (Criterion 3, 25 pts). Sandwich-bracket CRF defensibility + D_all_in_4B zero-regret result.

Element 7 — Wildfire-PV Interaction

See Element 7 summary.

Element 8 — Sensor Placement Optimization

See Element 8 summary.

Cross-Cascade Methods

See Cross-Cascade Methods summary. Analytical infrastructure spanning multiple elements: portfolio frontier, POMDP decision framework, NSGA-II Pareto search, Sobol-informed research priority. Serves Criterion 1 Technical Merit.

Supporting analyses

14 investigations that underpin findings cited above but don’t carry visitor narrative directly. Source files live in the cascade repository — not deployed here. The path column below points to each investigation’s source files.

InvWhat it doesIn-repo
10 Ozone crossover analysis — terminal analysis of ozone net-positive crossover events; supports Investigation 4-1 framing. investigations/10_ozone-crossover/latest/
11 CRF-conditional policy — older CRF-policy interaction precursor to Investigation 6-6. investigations/11_crf-conditional-policy/latest/
13 DTE Stockton biomass anomaly — retire-vs-keep decision (3.7 deaths/yr at $0; included in A_free_lunch). investigations/13_biomass-anomaly/latest/
19 Indoor-air retrofit pathway — per-CRF (Di / Krewski) deaths per $B; load-bearing for Investigation 6-9 ISRM validation. investigations/19_indoor-air-coupling/latest/
20 Grid marginal EV — marginal-emissions analysis under different grid dispatch scenarios; inputs to Investigation 1-1. investigations/20_grid-marginal-ev/latest/
28 Data assimilation for PM2.5 — Bayesian fusion of monitoring + ISRM fields; foundation for Investigation 3-4 and Investigation 3-8. investigations/28_data-assimilation-pm25/latest/
30 Wildfire PCE-UQ — polynomial-chaos uncertainty quantification on wildfire emissions surrogate. investigations/30_wildfire-pce-uq/latest/
31 Unified Bayesian experimental design — cross-cutting BED methodology infrastructure. investigations/31_unified-bed/latest/
33 4D-Var assimilation — variational data assimilation infrastructure for atmospheric inversion. investigations/33_dvar-assimilation/latest/
34 PINN advection-diffusion-reaction surrogate — physics-informed neural network as alternative to MFGP. investigations/34_pinn-adr-surrogate/latest/
35 GP advection-diffusion-reaction surrogate — classical Gaussian-process surrogate variant. investigations/35_gp-adr-surrogate/latest/
36 EVPPI Strong-GAM — Strong et al. 2014 GAM-based EVPPI estimation; methodology infrastructure. investigations/36_evppi-strong-gam/latest/
37 Smoke-PV coupling — precursor pathway analysis for the Investigation 7-1 wildfire-PV finding. investigations/37_smoke-pv-coupling/latest/
57 MAIAC AOD daily L1 — third independent L1 source for daily L4 MFGP test (FAIL +132% RMSE; closes daily-L4 scope boundary). investigations/57_maiac-aod-daily-l1/latest/