Wind speed and power forecasting: Evaluating NCUM-G model performance

In response to the growing emphasis on transitioning to clean, sustainable energy and the increasing global energy demand, the utilization of renewable resources has notably expanded. Among these resources, wind stands as a significant contributor. Presently, India relies on wind farms for over 25% of its renewable energy production. Considering the cost of installation, maintenance, and commitment of the Government of India to become carbon neutral by 2070, it is important to harness more and more wind energy. This underscores the significance of precise wind energy forecasting, which is crucial for optimizing schedules and enhancing the efficiency of wind farm operations. Accurate forecasting stands as the linchpin for unlocking the full potential of this plentiful renewable resource. The present study evaluates the performance of the hourly wind speed forecast of the National Centre for Medium Range Weather Forecasting (NCMRWF) Global Unified Model (NCUM-G) during the last 4 years (2019-2022). Considering the wind sites in the top six wind-rich states of India, the performance of the NCUM-G is evaluated using different statistical error measures. The results suggest that model forecast accuracy in the first 24 hours is reasonably good and decreases appreciably at higher lead times. Accuracy in terms of a difference in the forecast and observed wind speed for thresholds of 0.75, 1.0, and 1.5 m/s are 79, 87, and 95%, respectively, whereas the mean absolute error is less than similar to 1 m/s. Furthermore, the study indicates an enhancement in model performance and improved skill in 2022 compared to 2019. The study also assesses the effectiveness of bias correction in wind speed forecasts, leading to improved wind power forecasting, particularly at sites with higher forecast errors.