Advanced Meteorological Modeling Across Scales

Background

Within the application of numerical weather prediction (NWP) models for the assessment of global wind resources, two main challenges arise:
  1. The atmosphere needs to be properly represented consistently over a broad range of scales (from synotical to meso-micro scale) and this is expected to be achieved by modeling the system in its entity.
  2. If surface characteristics become important in influencing the wind patterns, high resolution is required for modeling the area of interest.

Current modeling approaches split efforts using nested simulations to tackle those issues: global models with lower resolution drive separate models with local refinements of higher resolution. At the same time, newer models are being developed aiming at tackling both requirements at the same time.

Research Questions

The first focus point of this PhD project is the investigation of wind statistics across scales, that means the analysis and interpretation of data over a variety of scales. The extent to which traditional  modeling methods provide sufficient precision in those statistics with regard to wind to power calculation systems is not well known. Areas characterized by complex terrain or organized atmospheric features (for example gravity waves or convection with open cell structures) are especially challenging. Within the scope of traditional NWP models (for example the Weather Research and Forecasting model (WRF) [1]), uniform grid sizes over the whole model domain are required with the consequence of computational challenges when large areas of interest are demanded or in case of high spatial and temporal resolution.  This conditions are faced for example in investigations of offshore wind farms and their wake effect, which will be the second focus point of the PhD project. Reducing the above mentioned issues of computational challenges, new generations of NWP models feature flexible mesh design (Illustration 1). This makes it possible to focus the computational power to areas where they actually required the most. One model of this category is the Model for Prediction Across Scales (MPAS, [2]) which will be used to investigate how this new progresses in model design can be applied with regard to challenges within the area of wind resource assessment.

Methods

In a first step, the traditional standard wind resource assessment approach by means of WRF is used to determine wind statistics across scales. Subsequently, the same studies are carried out with the new model environment MPAS to determine to what extent the new system can make contributions, can support or provide enhancements. Main areas of investigations are larger scale organized atmospheric structures and wind farm wakes and all analysises are validated with measurements.

Greater Framework

This project is contributing to OffshoreWake [3], which is an ongoing ForskEL project where both subjects are of relevance. The introduction of MPAS into the area of wind energy applications will be the first of its type.

Contact

Marc Imberger
PhD student
DTU Wind Energy