EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Quantitative Assessment of Uncertainty Reduction as a Function of Measurement Campaign Length Using Linear and Machine-Learning MCP Models

Alejandro Abascal Mendez, Ana Del Castillo Martín, Olga Álvarez Pérez-Aradros, Paulo Henrique Figueiredo Vaz, Ana Patricia Talayero Navales, Roberto Lázaro Gastón, Andrés Llombart Estopiñán

Inventions · 02.03.2026 · 11 (2) · 23

DOI: 10.3390/inventions11020023 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Inventions, online erschienen am 02.03.2026. Verfasst von Alejandro Abascal Mendez und 6 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study evaluates the impact of measurement campaign duration on wind resource characterization using three MCP (Measure–Correlate–Predict) models: Total Least Squares (TLS), Multiple Linear Regression (LR), and Quantile Gradient Boosting (GB). The analysis is based on data from 30 meteorological masts (nine primary and twenty-one secondary masts) installed worldwide across different terrains, with up to twenty-seven months of concurrent wind measurements between primary and secondary masts. Fixed campaign durations of 3, 4, 5, 6, 9, and 12 months were simulated using moving intervals to quantify the effect of measurement length on mean wind speed estimation. This working framework also serves to represent conditions typical of campaigns where LIDAR systems are used to complement meteorological mast deployments, as LIDAR units generally operate for shorter periods due to frequent relocation as part of broader measurement strategies. Wind speed estimation was assessed through metrics such as Mean Absolute Error (MAE), relative uncertainty, and monthly uncertainty reduction, taking into account terrain complexity and correlation coefficient (R2) between masts. Results indicate that extending the measurement period improves the accuracy and consistency of wind speed estimates, with significant reductions in uncertainty observed after six months. Across sites, the average monthly uncertainty reduction ranges from 0.13% to 0.41% of the mean wind speed per additional month of measurements, depending on terrain complexity and inter-mast correlation. Linear models (TLS and LR) consistently show better performance in terms of error and uncertainty reduction compared to GB. Based on an extensive and diverse MCP dataset covering multiple terrains and locations, this study provides empirically derived monthly uncertainty-reduction benchmarks for campaign-length optimisation under different site conditions, contributing to more reliable wind resource assessments and, consequently, energy yield estimates.

Zitieren

APA

Mendez, A. A., Martín, A. D. C., Pérez-Aradros, O. Á., Vaz, P. H. F., Navales, A. P. T., Gastón, R. L., & Estopiñán, A. L. (2026). A Quantitative Assessment of Uncertainty Reduction as a Function of Measurement Campaign Length Using Linear and Machine-Learning MCP Models. Inventions, 11(2), 23. https://doi.org/10.3390/inventions11020023

Vancouver

Mendez AA, Martín ADC, Pérez-Aradros OÁ, Vaz PHF, Navales APT, Gastón RL, et al. A Quantitative Assessment of Uncertainty Reduction as a Function of Measurement Campaign Length Using Linear and Machine-Learning MCP Models. Inventions. 2026;11(2):23. doi:10.3390/inventions11020023

BibTeX
@article{Mendez2026,
  author = {Mendez, Alejandro Abascal and Martín, Ana Del Castillo and Pérez-Aradros, Olga Álvarez and Vaz, Paulo Henrique Figueiredo and Navales, Ana Patricia Talayero and Gastón, Roberto Lázaro and Estopiñán, Andrés Llombart},
  title = {A Quantitative Assessment of Uncertainty Reduction as a Function of Measurement Campaign Length Using Linear and Machine-Learning MCP Models},
  journal = {Inventions},
  year = {2026},
  volume = {11},
  number = {2},
  pages = {23},
  doi = {10.3390/inventions11020023}
}
Kostenlos starten →