Case G: validation
Contents
- Measurement locations
- Explore the points
- OpenFOAM 12
- Measured against modelled
- Vertical profiles
- Field maps
- OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration open-field
- OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration pine
- OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration conifer
- OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration screen
- OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration half-belt-conifer
- Not run
- Notes on every run
- Leaderboard
Engine status: OpenFOAM 12 scored · OpenLB pending · FluidX3D pending · ML surrogate pending
Nomenclature: symbols, columns and scores
Geometry, reference scales and release
- \(H\), \(b\), \(D\) (m) — height of the building, cube, tower or tree belt; width of the building; edge of the cubes of Case C: the lengths the cases are drawn and normalised by
- \(U_H\) (m/s) — reference speed: the mean approach-flow speed at the height \(H\) (each Data page gives the value used)
- \(U_R\), \(H_R\), \(Z_R\) (m/s, m) — reference speed and height of a workbook: Case E, \(U_R\) at \(Z_R\) = 250 m full scale (1 m in the model); Case M, \(U_R\) and \(H_R\) at the reference point of the tunnel and of each field run
- \(U_{\mathrm{ref}}\), \(L_{\mathrm{ref}}\) (m/s, m) — reference speed and length of a case’s normalisation: \(U_H\) and \(H\) (\(U_R\) and \(H_R\) in Case M); for the velocity ratios the approach speed at 15.9 m (Case E) or at 500 m (Case F)
- \(U_b\) (m/s) — Case G: mean approach-flow speed at 9 m, the normalisation of its approach-flow table
- \(q\), \(c_{\mathrm{gas}}\) (m³/s, ppm) — tracer release of the experiment: volume flow of the released gas, and its tracer content (\(10^6\) ppm when pure); \(Q\) is the source strength given to a solve
- \(R_b\) (dimensionless) — bulk Richardson number of the heated-floor cases J and L (negative: unstable stratification)
- \(\mathrm{Re_H}\), \(\mathrm{Re_D}\) (dimensionless) — Reynolds numbers \(U_H H/\nu\) and \(U_H D/\nu\) (\(\nu\): kinematic viscosity of air)
- \(u_\ast\), \(z_0\), \(\kappa\) (m/s, mm, dimensionless) — friction velocity and roughness length of the log law fitted to the approach flow, and the von Kármán constant (0.41)
- \(\mathrm{WD}\) (column
wind_direction_deg; °) — wind direction (each frame says how it is counted); \(\mathrm{WD_R}\) is the mean direction of a field run at its reference point (Case M) - \(x\), \(y\), \(z\) (columns
x,y,z; m) — coordinates in the case’s frame (\(+x\), \(+y\) and \(+z\) are the positive directions; each Data page says which way \(+x\) points), \(z\) up from the floor of the tunnel or from the ground - \(\mathrm{d}x\) (m) — lattice spacing of the OpenLB and FluidX3D runs, in metres or as a fraction of \(H\) (\(H/64\))
- \(L_r\) (m) — reattachment length behind the building, read from a wake line at ground level; reported as \(L_r/H\)
Velocity and turbulence
- \(U\), \(V\), \(W\) (columns
U,V,W; m/s) — mean velocity along \(+x\), \(+y\), \(+z\); \(V\) is positive toward \(+y\) in every case - \(\lvert\mathbf{U}\rvert\) (m/s) — magnitude of the mean velocity vector, \(\sqrt{U^2+V^2+W^2}\)
- \(V_{\mathrm{s}}\) (column
Vs; m/s) — mean omnidirectional scalar speed of a thermistor (never a vector magnitude) - \(\overline{u'u'}\), \(\overline{v'v'}\), \(\overline{w'w'}\) (columns
uu,vv,ww; m²/s²) — velocity variances (overbar: time average, prime: fluctuation about it) - \(\overline{u'v'}\), \(\overline{u'w'}\), \(\overline{v'w'}\) (columns
uv,uw,vw; m²/s²) — Reynolds shear stresses - \(k = \tfrac{1}{2}\left(\overline{u'u'}+\overline{v'v'}+\overline{w'w'}\right)\) (column
k; m²/s²) — turbulent kinetic energy - \(U/U_{\mathrm{ref}}\) (column
velocity_ratio; dimensionless) — mean speed over the case’s reference speed (\(V_{\mathrm{s}}/U_{\mathrm{ref}}\) for a thermistor) - \(\sigma_{U/U_{\mathrm{ref}}}\) (column
velocity_ratio_sd; dimensionless) — standard deviation ofvelocity_ratioover the observations behind a field value (Case F) - \(k/U_{\mathrm{ref}}^2\) (column
k_ratio; dimensionless) — turbulent kinetic energy over the square of the reference speed, where only normalised values are published (Case G)
Concentration and temperature
- \(C\) (column
C; ppm) — mean tracer-gas concentration as published (volume fraction \(\times 10^6\) of the released gas diluted in air) - \(c_{\mathrm{rms}}\) (column
c_rms; ppm) — root mean square of the concentration fluctuation - \(c_0\) (ppm) — reference concentration, defined for each case on its Data page: \(c_0 = \dfrac{c_{\mathrm{gas}}\thinspace q}{U_H H^2}\) (\(U_R H_R^2\) in the denominator in Case M)
- \(c^\ast\) (column
concentration_ratio; dimensionless) — dimensionless concentration \(c^\ast = C/c_0\); the value of a solve is \(c\thinspace U_H H^2/Q\) for its mean concentration \(c\) and its source strength \(Q\) - \(\mathrm{Sc_t}\) (dimensionless) — turbulent Schmidt number of a solve (0.3, 0.5 and 0.7)
- \(T\), \(T_{\mathrm{rms}}\) (columns
T,T_rms; °C, K) — mean air temperature as published (Cases J and L) and the root mean square of its fluctuation - \(\overline{u'c'}\), \(\overline{v'c'}\), \(\overline{w'c'}\) (columns
uc,vc,wc; ppm·m/s) — turbulent concentration fluxes - \(\overline{u'\theta'}\), \(\overline{v'\theta'}\), \(\overline{w'\theta'}\) (columns
uT,vT,wT; K·m/s) — turbulent heat fluxes (\(\theta'\): the temperature fluctuation) - \(\theta = \dfrac{T_f - T}{T_f - T_H}\) (dimensionless) — floor-referenced temperature the solves of Cases J and L are scored on: 0 at the floor, 1 in the approach air at \(H\); \(T_f\) is the floor temperature and \(T_H\) the temperature of the approach flow at \(H\)
Scores
- \(n\) — number of pairs of a measured value \(o_i\) and a modelled value \(p_i\) at one point
- \(R\) — Pearson correlation coefficient of the pairs
- \(R^2\) — square of \(R\): share of the measured variance a linear fit to the model explains
- \(\mathrm{FAC2}\) — fraction of pairs with \(0.5 \le p_i/o_i \le 2\)
- \(\mathrm{FB}\) — fractional bias \(2(\overline{o}-\overline{p})/(\overline{o}+\overline{p})\), positive when the model is low
- \(\mathrm{HR}\) (hit rate) — fraction of pairs with \(\lvert p_i-o_i\rvert/\lvert o_i\rvert \le \delta_{\mathrm{rel}}\) or \(\lvert p_i-o_i\rvert \le \delta_{\mathrm{abs}}\), where \(\delta_{\mathrm{rel}} = 0.25\) and \(\delta_{\mathrm{abs}} = 0.05\thinspace U_H\) for a velocity component
- \(\mathrm{NMSE} = \dfrac{\overline{(o-p)^2}}{\overline{o}\thinspace \overline{p}}\) — normalised mean square error
- \(\mathrm{NAD} = \dfrac{\overline{\lvert o-p\rvert}}{\overline{o}+\overline{p}}\) — normalised absolute difference
- \(\mathrm{MG} = \exp\overline{(\ln o-\ln p)}\) — geometric mean bias
- \(\mathrm{VG} = \exp\overline{(\ln o-\ln p)^2}\) — geometric variance
- The ratio of \(\mathrm{FAC2}\) keeps its sign, so a model of the other sign than the measurement is a miss; \(\mathrm{FB}\), \(\mathrm{NMSE}\) and \(\mathrm{NAD}\) divide by the means, so they are undefined (shown as —) for a signed component unless the measured and the modelled mean are both positive
- VDI 3783-9 calls the hit rate \(q\) and the tolerances \(D\) and \(W\); on this site those letters keep their other meanings (\(q\) the tracer release rate, \(D\) the block size of Case C, \(W\) the vertical velocity)
- The detection limit of a concentration is one value per case, \(10^{-3}\) of the largest measured value. It sets \(\delta_{\mathrm{abs}}\) of the hit rate and the both-below rule of \(\mathrm{FAC2}\), and floors the values in \(\mathrm{MG}\) and \(\mathrm{VG}\).
- The scores \(\mathrm{FB}\), \(\mathrm{FAC2}\), \(\mathrm{NMSE}\), \(\mathrm{NAD}\), \(\mathrm{MG}\) and \(\mathrm{VG}\) follow Chang & Hanna (2004); the hit rate follows VDI 3783-9.
Quantities scored
- \(U\), \(V\) and \(W\) (mean velocity components) over \(U_H\)
- \(k\) (turbulent kinetic energy) over \(U_H^2\)
- the speed \(\lvert\mathbf{U}\rvert\) and \(\sqrt{\lvert\mathbf{U}\rvert^2+2k}\), against a thermistor’s \(V_{\mathrm{s}}\)
- \(c^\ast = C/c_0\) (dimensionless concentration) at turbulent Schmidt numbers \(\mathrm{Sc_t}\) of 0.3, 0.5 and 0.7
- \(\theta\) (floor-referenced temperature)
LAD and Cd are not in the dataset (only the shading rate, 81 %). The drag values are crowns of the Eddy3D vegetation library.
Measurement locations
Explore the points
OpenFOAM 12
Status: scored
Eddy3D production runs solved on 2026-10-18, realizable k-ε at numerics level 3 and 3000 iterations, on one mesh. The solves ran to the iteration limit with the SIMPLE residual control unmet; the initial residual of epsilon at the last iteration was \(2.7 \times 10^{-2}\) to \(3.6 \times 10^{-2}\).
- A field observation at full scale. The measurement is \(U/U_H\) and \(k/U_H^2\) at 28 points, \(U_H\) the wind at the tree height (7 m) during each period.
- Eddy3D models trees as a Darcy-Forchheimer momentum sink only. The leaf area density (LAD) and drag coefficient (Cd) of the belt are not in the dataset (only its shading rate, 81 %); the drag values are crowns of the Eddy3D vegetation library.
- Eddy3D solved a sweep: no belt, and three crowns from its vegetation library of rising Cd·LAD (0.30, 0.625 and 0.875 1/m), on a two-dimensional slab, and the conifer crown on the real 74 m belt as a half domain. Each arm is a separate run.
- The model is divided by its own \(U(7\text{ m})\) at \(x = -35\thinspace\mathrm{m}\), because the tabulated inlet has a smooth floor and the approach does not stay the imposed profile; scores over the imposed \(U_H\) = 5.299 m/s are in the files too.
Measured against modelled
| Arm | Cd·LAD [1/m] | \(U\) R | \(U\) FAC2 | \(U\) FB | \(U\) hit rate | mean \(U/U_H\) | RMSE of \(U\) |
|---|---|---|---|---|---|---|---|
| no belt (control) | 0 | +0.300 | 0.464 | −0.659 | 0.000 | 0.948 | 0.474 |
| pine | 0.30 | +0.503 | 1.000 | −0.425 | 0.036 | 0.735 | 0.262 |
| conifer | 0.625 | +0.489 | 1.000 | −0.248 | 0.286 | 0.613 | 0.145 |
| screen | 0.875 | +0.459 | 1.000 | −0.148 | 0.857 | 0.554 | 0.097 |
| conifer, real belt (half domain) | 0.625 | +0.468 | 1.000 | −0.202 | 0.607 | 0.585 | 0.120 |
The field’s mean \(U/U_H\) over the 28 points is 0.478. The model is \(\lvert\mathbf{U}\rvert\) over its own \(U(7\text{ m})\); the same scores over the imposed \(U_H\), and for \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}\), are in the files. The same arms for \(k\):
| Arm | \(k\) R | \(k\) FAC2 | \(k\) FB | mean \(k/U_H^2\) |
|---|---|---|---|---|
| no belt (control) | −0.087 | 0.714 | +0.300 | 0.023 |
| pine | +0.213 | 0.714 | +0.138 | 0.027 |
| conifer | +0.278 | 0.750 | −0.090 | 0.034 |
| screen | +0.234 | 0.679 | −0.206 | 0.038 |
| conifer, real belt (half domain) | +0.228 | 0.750 | −0.081 | 0.034 |
The field’s mean \(k/U_H^2\) is 0.031.
- Drag. The mean \(U/U_H\) falls from 0.948 with no belt to 0.735, 0.613 and 0.554 as Cd·LAD rises, and the RMSE of \(U\) from 0.474 to 0.097. \(U\) R stays between +0.46 and +0.50 in every arm with trees (+0.30 with none). The densest crown has mean \(U/U_H\) 0.554 against 0.478 measured.
- Factor of two. In every arm with trees, all 28 points are within a factor of two (FAC2 1.000); with no belt FAC2 is 0.464. For \(U\), FB is negative in every arm (the model is high).
- Stations. The field’s station-mean speed is lowest at \(x/H = 1\) (0.435) and recovers to 0.51 at \(4\thinspace H\); the conifer arm has its minimum at the same station (0.586) and recovers to 0.661 at \(5\thinspace H\). The field’s mean \(k/U_H^2\) rises from 0.018 at \(1\thinspace H\) to 0.053 at \(5\thinspace H\); in every arm it falls (conifer 0.036 to 0.030). Eddy3D’s trees have no canopy \(k\) or ε source terms.
- Two dimensions against the real belt. The real belt on a half domain is slower than the infinite slab by 0.028 of \(U/U(7\text{ m})\) on average (at most 0.042 over the 28 points; R between the two 0.99). Between the pine and the conifer crown of the sweep the difference is 0.12.
- Flow structure. In every arm with trees: a velocity deficit directly behind the belt against the open field, recovery with distance, a speed-up through the 1.2 m gap, and more drag slowing the lee more and pushing more wind through the gap.
Interpretation: Case G drag sweep
- Drag changes the level of the mean speed and leaves the pattern (R) where it was.
- The arm nearest the data is the densest one and the edge of the sweep, so the sweep does not bracket the field.
- The streamwise shape is present only in part. For \(k\) the trend is the other way, and the modelled \(k\) behind the belt is what the inflow and the shear layer over the belt produce.
- The two-dimensional idealisation costs 0.028 of \(U/U(7\text{ m})\), against 0.12 between the pine and the conifer crown.
Vertical profiles
The field and the model at four of the masts, from the belt to \(5\thinspace H\) behind it.
Field maps
The solved velocity on the measured planes, every point drawn on it as a dot coloured by its measured value on the same scale: a dot that stands out from the colour around it is a point the model misses.
OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration open-field
| Score | R | \(R^2\) | FAC2 | FB | Hit rate |
|---|---|---|---|---|---|
| \(\lvert\mathbf{U}\rvert/U(7\text{ m})\) (control: no belt) | +0.300 | 0.090 | 0.464 | -0.659 | 0.000 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U(7\text{ m})\) (control: no belt) | +0.278 | 0.077 | 0.357 | -0.681 | 0.000 |
| \(k/U(7\text{ m})^2\) (control: no belt) | -0.087 | 0.008 | 0.714 | +0.300 | 0.250 |
| \(\lvert\mathbf{U}\rvert/U_H\), imposed (control: no belt) | +0.300 | 0.090 | 0.571 | -0.645 | 0.000 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U_H\), imposed (control: no belt) | +0.278 | 0.077 | 0.464 | -0.667 | 0.000 |
Every score is over the same 28 points.
- Reproduce: Eddy3D fixture Eddy3DCli.Tests/Validation/TestAijCaseGContainerE2E (arm: no belt, full scale; OpenFOAM 12 realizable k-ε, numerics level 3, 3000 iterations, 90,576 cells, solved 2026-10-18); figures validation/plot_cases_gk_openfoam.py
OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration pine
| Score | R | \(R^2\) | FAC2 | FB | Hit rate |
|---|---|---|---|---|---|
| \(\lvert\mathbf{U}\rvert/U(7\text{ m})\) | +0.503 | 0.253 | 1.000 | -0.425 | 0.036 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U(7\text{ m})\) | +0.491 | 0.241 | 1.000 | -0.470 | 0.000 |
| \(k/U(7\text{ m})^2\) | +0.213 | 0.046 | 0.714 | +0.138 | 0.143 |
| \(\lvert\mathbf{U}\rvert/U_H\), imposed | +0.503 | 0.253 | 1.000 | -0.407 | 0.036 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U_H\), imposed | +0.491 | 0.241 | 1.000 | -0.452 | 0.036 |
Every score is over the same 28 points.
- Reproduce: Eddy3D fixture Eddy3DCli.Tests/Validation/TestAijCaseGContainerE2E (arm: pine, Cd·LAD 0.30 1/m, full scale; OpenFOAM 12 realizable k-ε, numerics level 3, 3000 iterations, 90,576 cells, solved 2026-10-18); figures validation/plot_cases_gk_openfoam.py
OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration conifer
| Score | R | \(R^2\) | FAC2 | FB | Hit rate |
|---|---|---|---|---|---|
| \(\lvert\mathbf{U}\rvert/U(7\text{ m})\) | +0.489 | 0.239 | 1.000 | -0.248 | 0.286 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U(7\text{ m})\) | +0.504 | 0.254 | 1.000 | -0.329 | 0.071 |
| \(k/U(7\text{ m})^2\) | +0.278 | 0.077 | 0.750 | -0.090 | 0.179 |
| \(\lvert\mathbf{U}\rvert/U_H\), imposed | +0.489 | 0.239 | 1.000 | -0.227 | 0.464 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U_H\), imposed | +0.504 | 0.254 | 1.000 | -0.309 | 0.143 |
Every score is over the same 28 points.
- Reproduce: Eddy3D fixture Eddy3DCli.Tests/Validation/TestAijCaseGContainerE2E (arm: conifer, Cd·LAD 0.625 1/m, full scale; OpenFOAM 12 realizable k-ε, numerics level 3, 3000 iterations, 90,576 cells, solved 2026-10-18); figures validation/plot_cases_gk_openfoam.py
OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration screen
| Score | R | \(R^2\) | FAC2 | FB | Hit rate |
|---|---|---|---|---|---|
| \(\lvert\mathbf{U}\rvert/U(7\text{ m})\) | +0.459 | 0.211 | 1.000 | -0.148 | 0.857 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U(7\text{ m})\) | +0.472 | 0.223 | 1.000 | -0.258 | 0.250 |
| \(k/U(7\text{ m})^2\) | +0.234 | 0.055 | 0.679 | -0.206 | 0.214 |
| \(\lvert\mathbf{U}\rvert/U_H\), imposed | +0.459 | 0.211 | 1.000 | -0.126 | 0.893 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U_H\), imposed | +0.472 | 0.223 | 1.000 | -0.237 | 0.321 |
Every score is over the same 28 points.
- Reproduce: Eddy3D fixture Eddy3DCli.Tests/Validation/TestAijCaseGContainerE2E (arm: screen, Cd·LAD 0.875 1/m, full scale; OpenFOAM 12 realizable k-ε, numerics level 3, 3000 iterations, 90,576 cells, solved 2026-10-18); figures validation/plot_cases_gk_openfoam.py
OpenFOAM 12, realizable k-epsilon (Eddy3D Outdoor), configuration half-belt-conifer
| Score | R | \(R^2\) | FAC2 | FB | Hit rate |
|---|---|---|---|---|---|
| \(\lvert\mathbf{U}\rvert/U(7\text{ m})\) | +0.468 | 0.219 | 1.000 | -0.202 | 0.607 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U(7\text{ m})\) | +0.469 | 0.220 | 1.000 | -0.290 | 0.179 |
| \(k/U(7\text{ m})^2\) | +0.228 | 0.052 | 0.750 | -0.081 | 0.179 |
| \(\lvert\mathbf{U}\rvert/U_H\), imposed | +0.468 | 0.219 | 1.000 | -0.178 | 0.750 |
| \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}/U_H\), imposed | +0.469 | 0.220 | 1.000 | -0.266 | 0.286 |
Every score is over the same 28 points.
- Reproduce: Eddy3D fixture Eddy3DCli.Tests/Validation/TestAijCaseGContainerE2E (arm: the real 74 m belt as a half domain, conifer, Cd·LAD 0.625 1/m, full scale; OpenFOAM 12 realizable k-ε, numerics level 3, 3000 iterations, 625,252 cells, solved 2026-10-18); figures validation/plot_cases_gk_openfoam.py
Not run
- A second mesh or iteration count.
- LAD or Cd other than the three library crowns. The wind direction relative to the belt normal is not published.
- OpenLB, FluidX3D and the ML surrogate.
Interpretation: Limits of these runs
- With one mesh and the residual control unmet, none of these numbers is mesh- or iteration-independent.
- The two-dimensional slab is an assumption, and the half belt is the only check on it.
Notes on every run
- Quantity: \(U/U_H\) and \(k/U_H^2\) at the 28 points, \(U_H\) the wind at the tree height (7 m): the measured mean scalar speed against the modelled \(\lvert\mathbf{U}\rvert\) and \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}\), both over the model's own \(U(7\text{ m})\) at \(x = -35\thinspace\mathrm{m}\), and over the imposed \(U_H\) = 5.299 m/s; LAD and Cd are not in the dataset
- Normalisation: the field's \(U/U_H\) and \(k/U_H^2\) (\(U_H\) = the wind at the tree height, 7 m, during each period) against the model's \(\lvert\mathbf{U}\rvert\) and \(k\) over the model's own \(U(7\text{ m})\) at \(x = -35\thinspace\mathrm{m}\) ('own approach') and, in the second set of scores, over the imposed \(U_H\) = 5.299 m/s ('imposed'); the model is also scored as \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}\), because the field is the mean of an instantaneous speed; hit rate \(\delta_{\mathrm{rel}} = 0.25\), \(\delta_{\mathrm{abs}} = 0.05\)
- Per-point values:
/api/cases/G/pairs/openfoam12-realizableKE.csv
Notes:
- The three crowns are entries of the Eddy3D vegetation library (Cd·LAD 0.30, 0.625 and 0.875 1/m, loss coefficients 1.2, 2.5 and 3.5).
- Eddy3D models trees as a Darcy-Forchheimer momentum sink only: there are no canopy \(k\) or epsilon source terms.
- The slab arms idealise the belt as two-dimensional (the belt through both slip planes, infinite width; the data paper's Figure 7 draws a two-dimensional tree); the half-belt arm is the real 74 m belt on a half domain with its centre plane on the symmetry plane. Both are scored against the same 28 values.
- The tabulated inlet carries the workbook's inflow over a smooth floor (the site's roughness is not in the model), so the approach flow does not stay the imposed profile: in the open-field arm it is 1.19 of it at 1.5 m and 0.995 at 7 m at \(x = -10\thinspace\mathrm{m}\), and the model's \(U(7\text{ m})\) at \(x = -35\thinspace\mathrm{m}\) is 0.976 to 0.984 of the imposed \(U_H\). The headline scores divide the model by that own \(U(7\text{ m})\).
- The field is the mean of the instantaneous speed of a 3-D sonic anemometer, which exceeds \(\lvert\mathbf{U}\rvert\) in a turbulent wake: the model is scored as \(\lvert\mathbf{U}\rvert\) and as \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}\), the bracket of Cases C to E.
- The approach profile and the wake were measured on different days (28 and 26 December 2000), and the approach \(k\) does not match the workbook's own inflow note (\(k/U_b^2\) = 0.20 at 9 m against the note's constant 3.02 m²/s², 0.096 \(U_b^2\)). The files do not explain the mismatch.
- Every arm ran to the 3000-iteration limit with the residualControl unmet; the epsilon initial residual at the last iteration was 2.7e-2 to 3.6e-2 in the slab arms and 3.2e-2 in the half belt.
- Every arm with trees shows a velocity deficit directly behind the belt against the open field, recovery with distance, a speed-up through the 1.2 m gap, and more drag slowing the lee more and pushing more wind through the gap.
Interpretation: Case G drag sweep
The arm nearest the 28 values says which library crown the belt resembles. The modelled \(k\) behind the belt follows from the inflow and the shear layer over the belt.
How the scores are computed
- Scores recomputed from the pairs by validation/dispersion_metrics.py; the SHA-256 of every file the snapshot was read from is in its results file, under
sources.
Leaderboard
| Rank | Run | Source | Hit rate, \(k\) | Hit rate, \(\sqrt{\lvert\mathbf{U}\rvert^2 + 2k}\) | Hit rate, \(\lvert\mathbf{U}\rvert\) | Hit rate | FAC2 | R | \(R^2\) |
|---|---|---|---|---|---|---|---|---|---|
| 1 | OpenFOAM 12 · screen | Eddy3D | 0.214 | 0.250 | 0.857 | 0.440 | 0.893 | +0.388 | 0.163 |
| 2 | OpenFOAM 12 · half-belt-conifer | Eddy3D | 0.179 | 0.179 | 0.607 | 0.321 | 0.917 | +0.388 | 0.164 |
| 3 | OpenFOAM 12 · conifer | Eddy3D | 0.179 | 0.071 | 0.286 | 0.179 | 0.917 | +0.423 | 0.190 |
| 4 | OpenFOAM 12 · open-field | Eddy3D | 0.250 | 0.000 | 0.000 | 0.083 | 0.512 | +0.164 | 0.058 |
| 5 | OpenFOAM 12 · pine | Eddy3D | 0.143 | 0.000 | 0.036 | 0.060 | 0.905 | +0.403 | 0.180 |
| 6 | OpenFOAM simpleFoam (Eddy3D Outdoor case writer) (eddy3d-of12-nutcap-lifted) | community | 0.071 | 0.036 | 0.071 | 0.060 | 0.762 | +0.332 | 0.110 |
All runs scored by validation/score_submissions.py. Ranked by mean hit rate (VDI 3783) over the tasks; bold: best per column. Not ranked: the ML surrogate (no point ids) and negative controls. Community rows are self-reported (Validate your own solver).