경험적 난류 스펙트럼 기반의 LDAPS 풍속 예보 시간 해상도 개선 연구
Temporal Resolution Enhancement of LDAPS Wind Forecasts Using Empirical Reconstruction of Turbulence Spectra
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
This study proposes an empirical turbulence spectrum-based approach to improve the temporal resolution of Local Data Assimilation and Prediction System (LDAPS) wind forecasts. While LDAPS forecasts provide valuable large-scale wind predictions, their inherent temporal resolution of one hour limits accurate modeling of short-term wind variability, which is critical for wind power applications. To overcome this, the measured power spectral density (PSD) from 14 days of lidar observations was averaged and used to complement the high-frequency components of the LDAPS forecasts, resulting in high-resolution wind speed time series with a one-minute resolution. Comparative analyses with Kaimal and von Kármán spectral models demonstrated that the empirical PSD model effectively suppresses unrealistic high-frequency variability and aligns more closely with the measured wind speed fluctuations. This approach significantly reduced the mean absolute error (MAE) and root mean square error (RMSE) when compared to traditional models, underscoring its potential as a reliable tool for high-resolution wind forecasting in operational contexts.
Abstract
This study proposes an empirical turbulence spectrum-based approach to improve the temporal resolution of Local Data Assimilation and Prediction System (LDAPS) wind forecasts. While LDAPS forecasts provide valuable large-scale wind predictions, their inherent temporal resolution of one hour limits accurate modeling of short-term wind variability, which is critical for wind power applications. To overcome this, the measured power spectral density (PSD) from 14 days of lidar observations was averaged and used to complement the high-frequency components of the LDAPS forecasts, resulting in high-resolution wind speed time series with a one-minute resolution. Comparative analyses with Kaimal and von Kármán spectral models demonstrated that the empirical PSD model effectively suppresses unrealistic high-frequency variability and aligns more closely with the measured wind speed fluctuations. This approach significantly reduced the mean absolute error (MAE) and root mean square error (RMSE) when compared to traditional models, underscoring its potential as a reliable tool for high-resolution wind forecasting in operational contexts.