The digital terrain model in the computational modelling of the flow over the Perdigão site: the appropriate grid size

Abstract

Abstract. The digital terrain model (DTM), the representation of earth's surface at regularly spaced intervals, is the first input in the computational modelling of atmospheric flows. The ability of computational meshes based on high- (2 m; airborne laser scanning, ASL), medium- (10 m; military maps, Mil) and low-resolution (30 m; Shuttle Radar Topography Mission, SRTM) DTMs to replicate the Perdigão experiment site was appraised in two ways: by their ability to replicate the two main terrain attributes, elevation and slope, and by their effect on the wind flow computational results. The effect on the flow modelling was evaluated by comparing the wind speed, wind direction and turbulent kinetic energy using VENTOS®/2 at three locations, representative of the wind flow in the region. It was found that the SRTM was not an accurate representation of the Perdigão site. A 40 m mesh based on the highest-resolution data yielded an elevation error of less than 1.4 m and an RMSE of less than 2.5 m at five reference points compared to 5.0 m in the case of military maps and 7.6 m in the case of the SRTM. Mesh refinement beyond 40 m yielded no or insignificant changes on the flow field variables, wind speed, wind direction and turbulent kinetic energy. At least 40 m horizontal resolution – threshold resolution – based on topography available from aerial surveys is recommended in computational modelling of the flow over Perdigão.

Impact of the wind field at the complex terrain site Perdigão on the surface pressure fluctuations of a wind turbine

“…In order to obtain accurate simulation results, the terrain model on which the computational fluid dynamics (CFD) simulations are based must be correspondingly detailed. Therefore, Palma et al (2020) In many studies, simulations of the flow field in Perdigão have already been carried out to investigate the effects of orography, vegetation, thermal stratification as well as meteorological effects in general. Wagner et al (2019) performed nested LES with WRF for the Iberian Peninsula with a highest resolution of 200 m around Perdigão covering almost 50 days.…”

“…Characteristic eddy-structures behind the ridges were observed and with a finest mesh resolution of 80 m the mean wind speed was captured well. Steady-state RANS calculations were used by Palma et al (2020) to discuss the impact of the resolution of the terrain model as well as of the CFD mesh on the local flow. They found that the flow in the valley was most affected by the resolution and recommend a resolution below 40 m. Salim Dar et al (2019) performed LES of the double ridge in Perdigão with a resolution of 10 m including the wind turbine, being represented with an actuator-disc model.…”

Comment on wes-2021-101

“…1). The Perdigão terrain mesh in E-Wind was based on the high-resolution (2 m) map provided by Palma et al (2020). It was resampled to a resolution of 10 m to fit the mesh resolution and to avoid artefacts in the mesh.…”

Comment on wes-2021-101

“…In order to obtain accurate simulation results, the terrain model on which the computational fluid dynamics (CFD) simulations are based must be correspondingly detailed. Therefore, Palma et al (2020) created a detailed digital terrain model (DTM) with a resolution of 2 m, which includes orography and vegetation.…”

“…Basis of the Perdigão terrain mesh for FLOWer is the same highly resolved (2 m resolution) digital terrain model (DTM) of the site in Perdigão (Palma et al, 2020) as for E-Wind (without resampling). This DTM was shifted so that the tower base of the turbine is located at x = y = 0 and the x-axis was rotated to align with the horizontal wind direction W D = 230 • at the turbine position at hub height.…”