V2.0 El valle angular ℓ = 14–20 en Planck: evidencia de estructura no gaussiana en el CMB Análisis multimap, oscilaciones finas y tests nulos masivos The Angular Valley ℓ = 14–20 in Planck: Evidence of Non-Gaussian Structure in the

F. Javier Martínez Sánchez, Forensic Characterization of Correlated Residuals in Large Spectroscopic Pipelines: A Case Study of SDSS/BOSS Data, Astronomy and Computing. DOI: https://doi.org/10.1016/j.ascom.2026.101155 Abstract This repository contains the complete reproducible research compendium associated with the detection and characterization of a statistically significant residual valley in the multipole range ℓ = 14–20 in the CMB × LSS cross-power spectrum. The analysis uses Planck 2018 component-separated CMB maps (SMICA, NILC, SEVEM, Commander), a projected LSS map (NSIDE=2048), a combined TCT–MS mask, and spectra computed with healpy. The valley is shown to be smooth, coherent, reproducible, and highly significant, with p-values below 3×10⁻⁶ based on 300,000 Gaussian simulations, and independent of the CMB component separation pipeline and sky orientation. These results represent a new and robust large-scale feature in the CMB × LSS correlation. Repository Contents The structure of the release is: VALLE_L14_20_RELEASE_v1.0/│├── data/ (CMB maps, LSS map, mask, input FITS files)├── pipeline/ (full Python scripts for spectra, valley extraction,│ functional fits, null tests, and isotropy checks)├── results/ (spectra, residuals, fits, Gaussian mocks, isotropy tests)├── paper/ (main.pdf and LaTeX source for the associated article)├── documentation/ (detailed methodology and reproducibility notes)├── optional/ (exploratory figures)│├── CITATION.cff├── license.txt└── README.md All analyses can be reproduced from scratch using the Python scripts in the pipeline/ directory. The results/ folder contains the numerical outputs and figures referenced in the scientific manuscript. Scientific Context The detected ℓ = 14–20 residual valley: appears consistently in SMICA, NILC, SEVEM and Commander is not reproducible with Gaussian noise (300k mocks) shows a coherent functional form (exponential / power-law / polynomial fits) persists under random rotations of the sky (isotropy test) constitutes a new, large-scale feature in the CMB × LSS correlation This release enables full transparency, reproducibility, and independent validation of the result. Intended Use Researchers can: reproduce the full analysis pipeline inspect the data, spectra, and statistical tests explore alternative models or hypotheses integrate the dataset into future CMB × LSS studioES [e-mail]