Wind energy forecasting with the NCAR RTFDDA and ensemble RTFDDA systems

In collaboration with Xcel Energy, NCAR is developing both deterministic and probabilistic numerical weather prediction (NWP) modeling systems as part of an operational wind power forecasting systems for wind farms across Minnesota, Colorado, Texas and some other neighboring states. The deterministic NWP system is based on the NCAR RTFDDA (Real Time Four-Dimensional Data Assimilation and forecasting system) and runs with three nested-grid domains with 3.3 km grid spacing for the finest mesh domain. The probabilistic NWP system is based on a 30-member multi-model, multi-approach ensemble RTFDDA (E-RTFDDA) and runs with two nested-grid domains with a fine mesh at 10km grid spacing. Both RTFDDA and E-RTFDDA were developed at NCAR for ATEC (Army Test and Evaluation Command), in order to support the Army's real-time operations at their test ranges and field sites. The models have subsequently been applied for numerous other weather-critical applications in different regions over the globe. In this paper, we describe the model configurations and the research and developments with respect to the operational Xcel RTFDDA and E-RTFDDA modeling systems. Results from case studies, real-time parallel tests, and the operational production runs are compared with the corresponding wind farm observations to illustrate the performances and the characteristics of the model forecasts. Algorithms for post-processing of high-resolution RTFDDA and ensemble E-RTFDDA predictions are briefly introduced. Challenges and new R&D directions for improving the accuracy of the wind forecasts at wind turbine hub-heights and the wind shear across the turbine blade disk at wind farms will also be discussed.