Deep Currents: A Residual-Cartography Audit of SPARC and RC100 Galaxy Dynamics

Deep Currents: A Residual-Cartography Audit of SPARC and RC100 Galaxy Dynamics presents a layer-by-layer residual analysis of galaxy rotation-curve data, focused on distinguishing robust empirical structure from observational, photometric, kinematic, and sample-selection systematics. The study uses public SPARC mass-model data, SPARC galaxy metadata, BTFR and central-density auxiliary tables, and RC100 high-redshift aperture-level measurements to test whether residuals around known galaxy-dynamics scaling relations contain a coherent, discovery-grade common signal. The analysis introduces a residual-cartography workflow: first defining calibrated empirical surfaces, then progressively removing mass/velocity-scale effects, radial-coverage effects, baryonic-structure trends, central-density correlations, distance and inclination systematics, photometric decomposition issues, gas/HI kinematic effects, low-velocity rotation-curve limitations, and model-floor fluctuations. Across 34 analysis layers, the initially visible residual structure is decomposed into identifiable families rather than converging to a single universal residual current. In the SPARC table-level analysis, the residual floor decreases substantially after accounting for known structure and quarantining problematic objects. A clean-control rerun yields a small robust residual floor of approximately 0.065 dex, with no strong surviving post-residual correlation. The most significant outliers are traced to local causes, including stellar mass-to-light tension, disk/bulge decomposition, edge-on deprojection, HI/gas asymmetry, shallow or non-flat rotation curves, distance uncertainty, inclination uncertainty, and low-velocity error floors. RC100 aperture-level tests likewise do not support a clean increasing acceleration-scale trend; apparent trends are better interpreted as structural, aperture, and high-redshift disk-evolution effects. The final result is a negative discovery in the strict sense: no discovery-grade common residual current remains in the public table-level SPARC/RC100 data. However, the work demonstrates that residual cartography is a useful methodological guardrail. It can separate true empirical regularities from layered astrophysical systematics and identify which galaxies require external follow-up through 3.6 μm imaging, HI cubes, position-velocity diagrams, side-to-side rotation curves, and independent distance or inclination priors. ZIP archive contents 1. Paper files 00_paper/deep_currents_residual_cartography.pdf 00_paper/deep_currents_residual_cartography.tex LaTeX auxiliary files: .aux, .log, .out paper figures: residual-floor progression closed-reef taxonomy model ladder by mask micro-reef residuals evidence-gate funnel 2. Source data Folder: 01_source_data/ MassModels_Lelli2016c.mrt.txt SPARC_Lelli2016c.mrt.txt BTFR_Lelli2016a.mrt.txt BTFR_Lelli2019.mrt.txt CDR_Lelli2016b.mrt.txt RAR.mrt RARbins.mrt Bulges.mrt.txt wise_ii.table1.txt Rotmod_LTG.zip Rotmod_ETG.zip SPARC audit paper PDF/TeX RC100 paper PDF extracted RC100 tables in CSV/XLSX RC100 figures and table-page images auxiliary analysis bundles and scripts 3. Layer archives Folder: 02_layer_archives/ Layer 1–7: initial residual-ocean / RC100 route analysis Layer 8: non-VES reframing Layer 9–17: outlier forensics, overshoot, M/L, morphology, decomposition, external proxy checks Layer 18–23: evidence protocol, image-profile proxy, clean-control rerun, final synthesis Layer 24–34: micro-reef dossiers, external gates, object-by-object closure, terminal closure 4. Extracted layer reports Folder: 03_layer_reports/ README.md layer*_report.md verdict matrices summary CSV files action matrices residual floor summaries gate decision matrices literature evidence matrices quarantine manifests terminal closure report 5. Scripts Folder: 04_scripts/ create_layer29.py create_layer30.py create_layer31.py create_layer32.py create_layer33_34.py sparc_audit_final.py sparc_regularities.py 6. Main package README Top-level file: README.md