Predictor-weighting strategies for probabilistic wind power forecasting with an analog ensemble

Constantin Junk, Luca Delle Monache, Stefano Alessandrini, Guido Cervone, Lueder Von Bremen

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

Unlike deterministic forecasts, probabilistic predictions provide estimates of uncertainty, which is an additional value for decision-making. Previous studies have proposed the analog ensemble (AnEn), which is a technique to generate uncertainty information from a purely deterministic forecast. The objective of this study is to improve the AnEn performance for wind power forecasts by developing static and dynamic weighting strategies, which optimize the predictor combination with a brute-force continuous ranked probability score (CRPS) minimization and a principal component analysis (PCA) of the predictors. Predictors are taken from the high-resolution deterministic forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF), including forecasts of wind at several heights, geopotential height, pressure, and temperature, among others. The weighting strategies are compared at five wind farms in Europe and the U.S. situated in regions with different terrain complexity, both on and offshore, and significantly improve the deterministic and probabilistic AnEn forecast performance compared to the AnEn with 10-m wind speed and direction as predictors and compared to PCA-based approaches. The AnEn methodology also provides reliable estimation of the forecast uncertainty. The optimized predictor combinations are strongly dependent on terrain complexity, local wind regimes, and atmospheric stratification. Since the proposed predictor-weighting strategies can accomplish both the selection of relevant predictors as well as finding their optimal weights, the AnEn performance is improved by up to 20 % at on and offshore sites.

Original languageEnglish (US)
Pages (from-to)361-379
Number of pages19
JournalMeteorologische Zeitschrift
Volume24
Issue number4
DOIs
StatePublished - Jan 1 2015

Fingerprint

wind power
principal component analysis
wind farm
forecast
geopotential
wind direction
stratification
wind velocity
decision making
weather
methodology
prediction

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this

Junk, Constantin ; Monache, Luca Delle ; Alessandrini, Stefano ; Cervone, Guido ; Von Bremen, Lueder. / Predictor-weighting strategies for probabilistic wind power forecasting with an analog ensemble. In: Meteorologische Zeitschrift. 2015 ; Vol. 24, No. 4. pp. 361-379.
@article{74a36fd82a9c47c7b461aa3c8de523c1,
title = "Predictor-weighting strategies for probabilistic wind power forecasting with an analog ensemble",
abstract = "Unlike deterministic forecasts, probabilistic predictions provide estimates of uncertainty, which is an additional value for decision-making. Previous studies have proposed the analog ensemble (AnEn), which is a technique to generate uncertainty information from a purely deterministic forecast. The objective of this study is to improve the AnEn performance for wind power forecasts by developing static and dynamic weighting strategies, which optimize the predictor combination with a brute-force continuous ranked probability score (CRPS) minimization and a principal component analysis (PCA) of the predictors. Predictors are taken from the high-resolution deterministic forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF), including forecasts of wind at several heights, geopotential height, pressure, and temperature, among others. The weighting strategies are compared at five wind farms in Europe and the U.S. situated in regions with different terrain complexity, both on and offshore, and significantly improve the deterministic and probabilistic AnEn forecast performance compared to the AnEn with 10-m wind speed and direction as predictors and compared to PCA-based approaches. The AnEn methodology also provides reliable estimation of the forecast uncertainty. The optimized predictor combinations are strongly dependent on terrain complexity, local wind regimes, and atmospheric stratification. Since the proposed predictor-weighting strategies can accomplish both the selection of relevant predictors as well as finding their optimal weights, the AnEn performance is improved by up to 20 {\%} at on and offshore sites.",
author = "Constantin Junk and Monache, {Luca Delle} and Stefano Alessandrini and Guido Cervone and {Von Bremen}, Lueder",
year = "2015",
month = "1",
day = "1",
doi = "10.1127/metz/2015/0659",
language = "English (US)",
volume = "24",
pages = "361--379",
journal = "Meteorologische Zeitschrift",
issn = "0941-2948",
publisher = "Gebruder Borntraeger Verlagsbuchhandlung",
number = "4",

}

Predictor-weighting strategies for probabilistic wind power forecasting with an analog ensemble. / Junk, Constantin; Monache, Luca Delle; Alessandrini, Stefano; Cervone, Guido; Von Bremen, Lueder.

In: Meteorologische Zeitschrift, Vol. 24, No. 4, 01.01.2015, p. 361-379.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Predictor-weighting strategies for probabilistic wind power forecasting with an analog ensemble

AU - Junk, Constantin

AU - Monache, Luca Delle

AU - Alessandrini, Stefano

AU - Cervone, Guido

AU - Von Bremen, Lueder

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Unlike deterministic forecasts, probabilistic predictions provide estimates of uncertainty, which is an additional value for decision-making. Previous studies have proposed the analog ensemble (AnEn), which is a technique to generate uncertainty information from a purely deterministic forecast. The objective of this study is to improve the AnEn performance for wind power forecasts by developing static and dynamic weighting strategies, which optimize the predictor combination with a brute-force continuous ranked probability score (CRPS) minimization and a principal component analysis (PCA) of the predictors. Predictors are taken from the high-resolution deterministic forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF), including forecasts of wind at several heights, geopotential height, pressure, and temperature, among others. The weighting strategies are compared at five wind farms in Europe and the U.S. situated in regions with different terrain complexity, both on and offshore, and significantly improve the deterministic and probabilistic AnEn forecast performance compared to the AnEn with 10-m wind speed and direction as predictors and compared to PCA-based approaches. The AnEn methodology also provides reliable estimation of the forecast uncertainty. The optimized predictor combinations are strongly dependent on terrain complexity, local wind regimes, and atmospheric stratification. Since the proposed predictor-weighting strategies can accomplish both the selection of relevant predictors as well as finding their optimal weights, the AnEn performance is improved by up to 20 % at on and offshore sites.

AB - Unlike deterministic forecasts, probabilistic predictions provide estimates of uncertainty, which is an additional value for decision-making. Previous studies have proposed the analog ensemble (AnEn), which is a technique to generate uncertainty information from a purely deterministic forecast. The objective of this study is to improve the AnEn performance for wind power forecasts by developing static and dynamic weighting strategies, which optimize the predictor combination with a brute-force continuous ranked probability score (CRPS) minimization and a principal component analysis (PCA) of the predictors. Predictors are taken from the high-resolution deterministic forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF), including forecasts of wind at several heights, geopotential height, pressure, and temperature, among others. The weighting strategies are compared at five wind farms in Europe and the U.S. situated in regions with different terrain complexity, both on and offshore, and significantly improve the deterministic and probabilistic AnEn forecast performance compared to the AnEn with 10-m wind speed and direction as predictors and compared to PCA-based approaches. The AnEn methodology also provides reliable estimation of the forecast uncertainty. The optimized predictor combinations are strongly dependent on terrain complexity, local wind regimes, and atmospheric stratification. Since the proposed predictor-weighting strategies can accomplish both the selection of relevant predictors as well as finding their optimal weights, the AnEn performance is improved by up to 20 % at on and offshore sites.

UR - http://www.scopus.com/inward/record.url?scp=84937554047&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84937554047&partnerID=8YFLogxK

U2 - 10.1127/metz/2015/0659

DO - 10.1127/metz/2015/0659

M3 - Article

AN - SCOPUS:84937554047

VL - 24

SP - 361

EP - 379

JO - Meteorologische Zeitschrift

JF - Meteorologische Zeitschrift

SN - 0941-2948

IS - 4

ER -