Saarblitz, Wind farm in HDR,, creative commons by-nc-sa 2.0

Furry Hill

Managed by

Data Provider: 

Ian Harman (CSIRO)

Data accesibility: 

The test case is offered to participants of the IEA Task 31 Wakebench who agree with the terms and conditions described below.

Site Description: 

The experiment is described in Finnigan and Brunet (1995). It is a logical follow-up of the homogeneous waving wheat experiment of Brunet et al. (1994) to study the effects of terrain-induced heterogeneity in canopy flows. To this end, the waving wheat model is placed on top of a two-dimensional hill.

The model canopy is made of nylon stalks 47 mm high and 0.25 mm wide in a square grid of side 5 mm and frontal area index 0.47. The model has aeroelastic similarity with respect to a wheat field (Finnigan and Mulhearn, 1978). The canopy area density is A = 10 m-1. The canopy drag coefficient Cd depends on the height. An average value of 0.68 can be assumed over the depth of the canopy, where drag = CdAU2.

A 50 mm fence at the beginning of the test section generates a deep turbulent boundary layer which develops over a 3 m long rough gravel surface, which raises progressively in order to match the estimated zero-plane displacement of the canopy (34 mm). The canopy then extends over 5.25 m in the streamwise direction occupying the full with of the tunnel. Downstream of the canopy, a raised gravel floor covers the rest of the working section. Three meters from the upwind edge of the canopy, a 2D-ridge with its axis perpendicular to the wind direction was inserted beneath the canopy. The ridge has a Witch of Agnesi profile: Zh(X)=h/(1+(X/L)2), with X=0 at the crest of the ridge, a ridge height of h = 150 mm and a length scale of L = 420 mm. The profile shape was truncated at X = ±925 mm where it smoothly meets the flat floor. The steepness h/L = 0.36 is sufficiently high to produce flow separation in the lee of the hill.

Figure 1: Sketch of the Furry Hill experiment from Finnigan and Brunet (1995). Dimmensions in meters.


The experiment was conducted in the Pye Laboratory wind tunnel (Wooding, 1968), CSRIO Black Mountain Laboratories, Camberra. The test section is 11 m long, 1.8 m wide and 0.65 m high. This facility enables the establishment of a zero pressure gradient boundary-layer (to within ±1 Pa/m locally and better over the full length of the tunnel). This was ensured before the hill model was installed.

Velocity component observations were taken using coplanar triple hot-wires mounted on a traversing system. Static pressure was also measured at the surface below the canopy by 33 pressure taps and on the tunnel roof by 20 taps. The static pressure in the mid-air was measured with a static probe and by subtracting from the dynamic pressure obtained from the hot wires. Both methods agree within 10% except in the separation region where the Pitot tube was not used.   

Measurements inside the canopy were made using a rectangular enclosure to protect the hotwire from damaging by flailing stalks.

Measurement Campaign: 

The dataset consists of vertical profiles of mean conditions and higher-order moments (up to third-order) of the flow over the waving wheat surface at 15 positions over the hill and 21 vertical levels.  

The tunnel was operated at a free-stream velocity of 12 m/s, when the stalks behave as though they were somewhat stiffer than wheat, more like a forest canopy.

It is important to notice that x-wires will rectify reversing flows so they will fail at reproducing the reverse flow in the separation region behind the hill. The turbulence moments will also be unreliable there and should only be taken as a general indication of the level of turbulence activity. Hence, flow data below 200 mm from X = 283 mm onwards should be viewed as contaminated due to recirculation/rectification issues. Nevertheless, flow visualizations allowed for a determination of the boundary of the separation region.


The test case is suitable for the analysis of the interaction of canopy flow with hilly terrain. For a better assessment of model performance it is expected that the user simulates the WavingWheat test case first.    

The Furry Hill test case has been used for validation in various references. Wilson et al. (1998) mention how to simulate the vertical gradient of Reynolds stress above the canopy by introducing an effective pressure gradient defined from the measured shear stress profile. They used a first-order mixing-length model and found similar performance than second-order models. They also studied the effect of using a z-dependent drag profile to characterize the canopy and found very small differences compared to the solution with an average drag coefficient. Ross and Vosper (2005) also used a mixing-length model with success in the flow over the Furry Hill. They showed how flow separation was enhanced by the presence of the canopy. Sogachev and Panferov (2006) used a RANS k-w turbulent model, leading to good agreement in the mean velocity, shear stress and turbulent dissipation profiles.

The main deviations of RANS models appear in the large overestimation of turbulent and momentum fluxes on the lee side of the hill due to the inherent limitations of these models in wake regions. Dupont et al. (2008) used LES over the Furry Hill in order to capture the transient turbulent characteristics of the flow. In effect, the LES approach shows the intermittency characteristics of the hill wake and the dominant sweep motions in momentum transfer at the canopy top all along the hill.


Brunet Y., Finnigan J.J., Raupach M.R., 1994, A Wind Tunnel Study of Air Flow in Waving Wheat: Single-Point Velocity Statistics,Boundary-Layer Meteorol.70: 95-132

Dupont S., Brunet Y., Finnigan J.J., 2008, Large-eddy simulation of turbulent flow over a forested hill: Validation and coherent structure identification,Q. J. R. Meteorol. Soc.134: 1911-1929

Finnigan J.J., Brunet Y., 1995, Turbulent airflow in forests, pp 3-40 in Wind and Trees, eds. Coutts and Grace, Cambridge University Press, 485 pp

Finnigan J.J., Mulhearn P.J., 1978, Modelling Waving Crops in a Wind Tunnel,Boundary-Layer Meteorol.14: 253-277

Ross A.N., Vosper S.B., 2005, Neutral turbulent flow over forested hills,Q. J. R. Meteorol. Soc.131: 1841-1862

Wilson J.D., Finnigan J.J., Raupach M.R., 1998, A first-order closure for disturbed plant-canopy flows, and its application to winds in a canopy on a ridge,Q. J. R. Meteorol. Soc.124: 705-732

Wooding, 1968, A low-speed wind tunnel for model studies in micrometeorology, II The Pye Laboratory Wind Tunnel, CSIRO Div. Plant Industry Tech 25: 25-39, available from CSIRO Marine and Atmospheric Research, FC Pye Laboratory, Camberra, ACT, Australia


Interested participants will have to subscribe to the attached data licensing agreement agreed between CENER and CSIRO. Please send two signed copies to: Javier Sanz Rodrigo, Calle Ciudad de la Innovación 7, 31621-Sarriguren, Spain



Loading widget...
No files to show