On the Offshore Advection of Boundary-Layer Structures and the Influence on Offshore Wind Conditions

Abstract

The coastal discontinuity imposes strong signals to the atmospheric conditions over the sea that are important for wind-energy potential. Here, we provide a comprehensive investigation of the influence of the land–sea transition on wind conditions in the Baltic Sea using data from an offshore meteorological tower, data from a wind farm, and mesoscale model simulations. Results show a strong induced stable stratification when warm inland air flows over a colder sea. This stratification demonstrates a strong diurnal pattern and is most pronounced in spring when the land–sea temperature difference is greatest. The strength of the induced stratification is proportional to this parameter and inversely proportional to fetch. Extended periods of stable stratification lead to increased influence of inertial oscillations and increased frequency of low-level jets. Furthermore, heterogeneity in land-surface roughness along the coastline is found to produce pronounced horizontal streaks of reduced wind speeds that under stable stratification are advected several tens of kilometres over the sea. The intensity and length of the streaks dampen as atmospheric stability decreases. Increasing sea surface roughness leads to a deformation of these streaks with increasing fetch. Slight changes in wind direction shift the path of these advective streaks, which when passing through an offshore wind farm are found to produce large fluctuations in wind power. Implications of these coastline effects on the accurate modelling and forecasting of offshore wind conditions, as well as damage risk to the turbine, are discussed.

Appendix: On the Influence of the PBL Scheme on Wind Conditions in Stable Stratification

To investigate the influence of the PBL scheme on the streaks described in Sect. 3.2.1, the simulations for 23 July 2012 were repeated with four different PBL schemes. Figure 17 shows hub-height fields of wind speed as in Fig. 11d obtained from the WRF model for four different PBL schemes: Mellor–Yamada–Nakanishi–Niino (MYNN) (a) (Nakanishi and Niino 2004), Yonsei-University (YSU) (c) (Hong et al. 2006) and Asymetric Convective Model version 2 (ACM) (d) (Pleim 2007). In addition we also implemented a recent suggested modification of the Mellor–Yamada–Janjić (MYJv2) scheme (Foreman and Emeis 2012) for more accurate wind-energy forecasts into the WRF model (b). All other parametrization schemes were defined as described in Sect. 2.2.

Fig. 17

Hub-height fields ($z$ = 67 m) of wind speed for four different PBL schemes [a Mellor–Yamada–Nakanishi–Niino, b Mellor–Yamada–Janjić modified after Foreman and Emeis (2012), c Yonsei University, d Asymmetric Convective Model version 2] obtained from WRF model simulations during stable offshore conditions on 23 July 2012 at 1200 UTC (hourly averages). The grey shaded triangle marks the position and extent of the wind farm

The results show that streaks are detectable in all of the PBL schemes chosen. The intensity of the streaks is stronger and the length of the streaks greater with the YSU and ACM parametrizations and slightly weaker in the MYNN and MYJv2 schemes. Location and origin of the streaks (patches of high roughness on the Darß-Zingst peninsula) are the same for all parametrizations chosen. The wind field simulated with the MYNN (Fig. 17a) scheme shows a stronger deflection of the flow towards the right at the coastal passage and a weakened afterwards, while the deviations are comparable for all other schemes. Generally the wind speeds tend to be higher in the MYNN and MYJv2 schemes, while they are slightly lower in the ACM and YSU schemes. In summary, it can be stated that there is no large influence of the PBL scheme on the occurrence of roughness-induced streaks of low wind speeds.