Using the concept of anomaly integration and historical climate data, we have developed a novel operational framework to implement deterministic numerical weather prediction within 15 days. Real‐case validation shows pronounced improvements in the forecasts of global geopotential heights in 20 out of 30 cases with the Community Atmosphere Model version 3.0. Seven other cases are marginally improved, and only three are deteriorated, in which all are ameliorated within the first‐week period. The average of the 30 cases shows an obvious increase of anomaly correlation coefficient (ACC) and a decrease of root mean square error (RMSE) of the geopotential height over global, hemispherical, and tropical zones. Significant amelioration on tropical circulation is displayed within the first‐week prediction. The forecasting skill is extended by 0.6 day in terms of days of the ACC greater than 0.6 for 500 hPa 30 case averaged geopotential height on global scale. The 30 case mean ACC and RMSE of 500 hPa temperature show the increment of 0.2 and −1.6 K, respectively, in the first‐week prediction. In the case of January 2008, much more reasonable horizontal distribution and vertical structure are achieved in bias‐corrected model geopotential height, temperature, relative humidity, and horizontal wind components in comparison to reanalysis data. In spite of a need for additional storage of historical modeling data, the new method does not increase computational costs and therefore is suitable for routine application.
A numerical model for shallow-water equations has been built and tested on the Yin–Yang overset spherical grid. A high-order multimoment finite-volume method is used for the spatial discretization in which two kinds of so-called moments of the physical field [i.e., the volume integrated average (VIA) and the point value (PV)] are treated as the model variables and updated separately in time. In the present model, the PV is computed by the semi-implicit semi-Lagrangian formulation, whereas the VIA is predicted in time via a flux-based finite-volume method and is numerically conserved on each component grid. The concept of including an extra moment (i.e., the volume-integrated value) to enforce the numerical conservativeness provides a general methodology and applies to the existing semi-implicit semi-Lagrangian formulations. Based on both VIA and PV, the high-order interpolation reconstruction can only be done over a single grid cell, which then minimizes the overlapping zone between the Yin and Yang components and effectively reduces the numerical errors introduced in the interpolation required to communicate the data between the two components. The present model completely gets around the singularity and grid convergence in the polar regions of the conventional longitude–latitude grid. Being an issue demanding further investigation, the high-order interpolation across the overlapping region of the Yin–Yang grid in the current model does not rigorously guarantee the numerical conservativeness. Nevertheless, these numerical tests show that the global conservation error in the present model is negligibly small. The model has competitive accuracy and efficiency.
The rapid development of wind energy in Japan and the associated high uncertainties and fluctuations in power generation present a big challenge for both wind power generators and electric grids. Accurate and reliable wind power predictions are necessary to optimize the integration of wind power into existing electrical systems. In this study, a hybrid forecasting system of wind power generation was developed by integrating the Kalman filter (KF) with the high resolution Weather Research and Forecasting (WRF) model as well as an empirical formula of wind power output (power curve). The system has been validated with observations including wind speed and power output over a six-month period for 15 turbine sites at a wind farm in Awaji-island, Japan. The results show that the tuned WRF model is able to provide hub-height wind speed prediction for the target area with reliability to some extent. The predicted wind field can be substantially improved by the Kalman filter as a post-processing procedure. The 15-turbine averaged improvements of mean error, root mean square error, and correlation coefficient are 97%, 22%, and 10%, respectively. Meanwhile, the Kalman filter also demonstrates a promising capability of reducing the uncertainties in the power curve model. Systematic validations regarding both wind speed and power output were carried out against the observations for the target wind farm, which show that the hybrid power forecasting system presented in this paper can be an effective and practical tool for short-term predictions of wind speed and power output in Japan area.