EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Image‐Based Multivariate Curve Resolution and Scalar Field Analysis of ERA5 Precipitation Over the Iberian Peninsula

Stefan Platikanov, Antoni Ginebreda, Romà Tauler

Journal of Chemometrics · 2026-11 · 40 (11)

DOI: 10.1002/cem.70191 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Chemometrics, online erschienen am 05.11.2026. Verfasst von Stefan Platikanov und 2 weiteren Autor:innen. Die Arbeit nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Ministerio de Ciencia e Innovación.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Gridded climate reanalysis datasets can be viewed as image cubes, yet chemometric decomposition methods remain underutilized in climate data analysis. This study applied pixel‐based Multivariate Curve Resolution–Alternating Least Squares (MCR‐ALS) with a trilinearity constraint to ERA5 precipitation data over the Iberian Peninsula (1940–2024). The 33 × 57 × 85 (latitude × longitude × year) data cube was unfolded pixel‐wise into an augmented two‐dimensional matrix, and the trilinearity constraint factorized the spatial profiles into separate latitude and longitude components, yielding an essentially unique decomposition and coherent two‐dimensional reconstructed precipitation maps. Grid‐scale artifacts in the resolved maps were subsequently mitigated by a discrete scalar‐field spectral filtering framework. Five precipitation zones were resolved, jointly explaining 95.6% of the total data variance: a Southwestern France–Western Languedoc Atlantic transition, the Central Pyrenees orographic zone, the Galician Atlantic façade, southern Portugal/Extremadura/Andalusia, and the central‐eastern Mediterranean façade. These zones closely matched those previously identified from ground‐based station data by principal component analysis, supporting the robustness of the pixel‐based gridded decomposition. The proposed framework, combining chemometric image‐based decomposition with a scalar‐field spectral treatment, provides an effective approach to extract interpretable spatiotemporal patterns from long‐term climate reanalysis datasets and is directly extensible to other climate variables.

Förderung

  • Ministerio de Ciencia e Innovación

Zitieren

APA

Platikanov, S., Ginebreda, A., & Tauler, R. (2026). Image‐Based Multivariate Curve Resolution and Scalar Field Analysis of ERA5 Precipitation Over the Iberian Peninsula. Journal of Chemometrics, 40(11). https://doi.org/10.1002/cem.70191

Vancouver

Platikanov S, Ginebreda A, Tauler R. Image‐Based Multivariate Curve Resolution and Scalar Field Analysis of ERA5 Precipitation Over the Iberian Peninsula. Journal of Chemometrics. 2026;40(11). doi:10.1002/cem.70191

BibTeX
@article{Platikanov2026,
  author = {Platikanov, Stefan and Ginebreda, Antoni and Tauler, Romà},
  title = {Image‐Based Multivariate Curve Resolution and Scalar Field Analysis of ERA5 Precipitation Over the Iberian Peninsula},
  journal = {Journal of Chemometrics},
  year = {2026},
  volume = {40},
  number = {11},
  doi = {10.1002/cem.70191}
}
Kostenlos starten →