table of contents
  1. first
  2. method 
  3. Results and Discussion 
  4. Summary and future prospects 
  5. References

first

With the spread of offshore wind power generation, the importance of highly accurate wind information is increasing. This study examines the effect of surface roughness length on wind speed estimation by the meso-meteorological model WRF (Weather Research and Forecasting model). Specifically, based on wind observation data at two locations, Mutsu Ogawara (Aomori Prefecture) and Hasaki Pier (Ibaraki Prefecture), the study examines the known problem of the tendency for wind speed to be overestimated during land breezes due to differences in roughness length.2)We evaluated the extent to which it would improve the 

Roughness length is an index that represents the friction characteristics of the earth's surface, and has a significant impact on the accuracy of surface wind speed.1)The WRF simulation used to develop this model used initial roughness lengths based on land use data from the USGS (United States Geological Survey). In this study, we used roughness lengths from the Japan Meteorological Agency's Non-Hydrostatic Model (NHM), with some modifications, to obtain more accurate results. 

method 

In this study, we compared Doppler lidar observation data with WRF simulations and performed a sensitivity analysis using two roughness length settings (USGS and JMA). Table 2 shows a summary of the wind observations. 

Site Mutsu Ogawara Hasaki Pier 
Measurement model DIABREZZA Windcube V1 
During the trial period 2017 / 06-2018 / 05 2015 / 10-2016 / 09 
Observation altitude 58–254m (5m intervals) 47-207m (20m intervals) 
Table 1. Wind observation overview 

This time, we performed a WRF simulation with a 500m resolution computational domain, referring to NeoWins etc. Figure 1 shows the WRF computational domain. 

Figure 1 (a) Mutsu Ogawara and (b) Hasaki Pier 
Observation sites, azimuth sector definitions, and WRF computation domain■ 
*The red dots indicate observation points, and the red dotted lines indicate the boundaries of azimuth sectors. 

For the WRF simulation, the following two cases were compared. 

  • WRF-CTRL: Case using roughness length based on 24 types of land use from USGS 
  • WRF-JMA: Based on WRF-CTRL, with only roughness length based on NHM settings (some modifications made). 

In this study, we selected a case where atmospheric stability was neutral and examined the error trend of WRF wind speed due to dynamic factors (roughness length). 

Results and Discussion 

Table 2 shows the average error of WRF wind speeds at Mutsu Ogawara and Hasaki Pier during atmospheric neutral and land breeze conditions. The conventional settings (WRF-CTRL) showed a tendency to overestimate wind speeds during land breeze conditions. On the other hand, by using the WRF-JMA settings, the overestimation during land breeze conditions was significantly improved.

Site (above ground level) WRF-CTRL WRF-JMA 
Mutsu Ogawara (99m) 27.0% + 10.3 % 
Hasaki Pier (95m) + 13.9 % + 3.0 % 
Table 2. Obtained by referring to the observed wind speeds at Mutsu Ogawara and Hasaki Pier 
Average error of WRF wind speed during atmospheric neutral and land breeze conditions

Figure 2 shows the distribution of the relative difference in the average WRF wind speed at 100 m above sea level. Comparing the wind speeds of WRF-CTRL and WRF-JMA, it was suggested that the wind speed gradient in coastal areas was improved by changing the roughness length. This is thought to be due to the slowing down effect of the difference in roughness length.

Figure 2 (a) Mutsu Ogawara and (b) Hasaki Pier at an altitude of 100m above sea level 
Distribution of the relative difference in the mean WRF wind speed (WRF-JMA minus WRF-CTRL)■ 
*Ash areas are areas above 100m above sea level. 

Summary and future prospects 

In this study, we clarified that the tendency of wind speed estimation to be overestimated can be improved by changing the roughness length setting of WRF. However, further research is needed to improve the error. In the future, we plan to focus on the following points. 

  • Improved thermodynamic factors: Further improvements in accuracy using the latest land and sea surface temperature data.
  •  Consideration of zero-plane displacement: Reflection of detailed topographical information, such as coastal windbreaks.

References

  1. NEDO, Offshore Wind Condition Map NeoWins
    URL: https://appwdc1.infoc.nedo.go.jp/Nedo_Webgis/top.html(Accessed: 2024/12/20)
  2.  Misaki, T., Ohsawa, T., Konagaya, M., Shimada, S., Takeyama, Y., Nakamura, S., Accuracy comparison of coastal wind speeds between WRF simulations using different input datasets in Japan. Energies, 2019, 12(14), 2754.