Tjalfe Gorm Poulsen

Evaporation from both cracked and uncracked soil surfaces under moist conditions was measured as a function of near‐surface wind speed, with wind speeds ranging between 0 and 5 ms ⁻¹ . Crack aperture width and distance between neighbouring cracks ranged from 2 to 20 mm and 40 to 120 mm, respectively, while crack depth was kept constant at 50 mm. Experiments were carried out in a wind tunnel with a 1 m × 1 m × 2.5 m test section, using large‐scale (50 cm diameter, 15 cm depth) soil samples. Sixty‐six combination of crack properties and wind speed were considered. Resulting evaporation data were used to assess the shape of the relationships between evaporation rate, wind speed conditions and crack properties. A semiempirical model for estimating evaporation from cracked soil, as a function of wind speed, crack width and crack distance, given evaporation as a function of wind speed from uncracked soil, was developed. Results showed that evaporation from cracked soil under moist conditions increased by 60%–65% compared to uncracked soil, for the wind speed and crack properties given above. Evaporation from crack apertures was up to eight times higher than from an uncracked surface under the same wind conditions. Results further indicated that for constant wind speed and crack distance, there exists a crack width at which evaporation from the soil is at maximum. The semiempirical model based on exponential and power function relationships, was able to very closely match the experimental data and is further expected to be able to approximate evaporation under natural conditions. Highlights: Near‐surface wind speed is strongly affected by presence of cracks Cracking in moist soils can increase evaporation up to 65% over uncracked soil There exists an intermediate crack width at which evaporation is at a maximum Evaporation may be estimated based on wind speed and crack properties

Local (cm) scale water loss distribution and evaporation in soil with flat and furrowed surfaces as a function of near‐surface wind speed and wind direction was investigated. Four levels of average main wind speed (10 cm above furrow ridge summits, 0, 1, 3 and 5 m·s ⁻¹ ), three main wind directions (compared to the direction of the furrows, 0°, 45° and 90°) and two levels of relative air humidity (70% and 86%) were considered. Experiments were conducted under controlled (wind tunnel) conditions, using 50 cm diameter and 15 cm height soil samples. For furrowed samples, 6 cm high ridges with 12 cm distance between ridge tops were used. Resulting wind speed profiles above ridge tops were relatively similar (for the same main wind speed), regardless of wind direction, but differed greatly in the valleys between ridges depending on especially wind direction. Strong indication of standing eddies on the lee side of furrow ridges for wind directions ≥45° was found. Local evaporative water loss spatial distribution in furrowed soil, was strongly dependent on wind direction, while local evaporation rates depended strongly on main wind speed and relative humidity level. Significantly higher evaporative water losses were observed at the wind‐exposed ridge slopes (for wind directions ≥45°) compared to lee slopes. Highlights: Local (cm) scale water loss in furrowed soil under windy conditions was investigated. Wind speed profiles between furrow ridges were strongly dependent on wind direction. Local water loss was significantly higher on windward slopes in furrowed soil. Local water loss spatial distribution was strongly dependent on wind direction.

Below‐surface horizontal pore gas velocity profiles in response to above‐surface wind conditions were measured for 16 combinations of wind speed (1.4–5.7 ms ⁻¹ ) and wind gust frequency (0–1 Hz) in four porous media with different particle size (0.6–10 mm), yielding 64 combinations of wind condition and porous medium using a recently developed experimental gas tracer tracking method. Experiments were carried out under controlled conditions in a wind tunnel. Gusty wind conditions were created using a wind‐driven propeller operating at variable rotation frequencies. A recently presented empirical model for predicting horizontal pore gas velocity in porous media in response to above‐surface wind conditions was validated against the experimental data. The model uses one empirical parameter ( B ) together with the wind speed at the surface as input parameters. Further analyses were carried out to identify relations between B , above‐surface wind conditions and porous medium properties. Results indicate that it is possible to accurately estimate B and thus predict below‐surface pore gas velocity profiles based on porous medium properties and surface wind speed. B was found inversely proportional to above‐surface wind speed, porous medium particle diameter and porous medium gas permeability, all in a linear manner. Highlights: Horizontal pore gas velocity profiles measured for 64 porous medium–wind speed combinations Recent two‐parameter empirical model for pore gas velocity evaluated using the experimental data Model accurately approximated experimental pore gas velocity data. Model parameters could be predicted from medium gas permeability and above‐surface wind speed

Below-surface horizontal pore air velocity profiles in porous media in response to above-surface wind conditions were measured for 16 combinations of wind speed (1.4 – 5.7 ms-1) and main wind gust frequency (0 – 1 Hz) in four granular porous media (particle size: 0.6 – 10 mm) yielding 64 combinations of wind condition and porous medium. Measurements were carried out under controlled (wind tunnel) conditions using a recently developed experimental gas tracer tracking method. Gusty wind conditions were induced by a wind driven propeller capable of variable rotation frequencies.A recent empirical model for predicting wind-induced horizontal pore velocity in porous media (using one empirical parameter, B together with surface wind speed as input) was validated against the experimental data and found to provide accurate approximations. Further investigations revealed that B could be accurately predicted from porous medium gas permeability. Overall the results indicate, that it is possible to accurately predict below-surface wind-induced pore velocity profiles based on the above-surface wind speed profile and the porous medium gas permeability. 

The impacts of five factors (wind speed, crack width, crack distance, crack depth and crack orientation (compared to wind direction)), on evaporation from cracked soil under moist soil conditions were investigated experimentally. Experiments were carried out under laboratory conditions in a wind tunnel with a 1 m × 1 m × 2.5 m test section using large‐scale (50 cm diameter, 25 cm depth) samples of natural soils wetted to a soil water content of 0.2 g g ⁻¹ . To reduce the number of necessary experiments, a 2 ^k factorial experimental design was used. Crack widths, crack distances, crack depths and crack orientations of 5–20 mm, 40–100 mm, 50–200 mm and 0–90°, respectively, were used. Effects of each factor on evaporation rate, individually and by interaction between multiple factors, were subsequently calculated. Results showed that individual effects of all five factors on evaporation rate were significant at the 95% level. Wind speed had the largest impact on evaporation rate, followed by crack width, crack distance, crack depth and crack orientation in that order. Overall, the four crack properties together explained 47% of the variations in evaporation rate. Highlights: Evaporation from cracked, moist soil under windy conditions was investigated. Crack orientation significantly affects wind conditions close to the surface. Both wind speed and crack properties significantly affected evaporation rate. Crack properties together explained 47% of variations in evaporation rate.

Above‐surface mean wind speed profiles and corresponding below‐surface horizontal pore air velocity profiles were measured for two porous media with different particle size and gas permeability under six different wind speed conditions. Experiments followed a 2 ^k factorial design to minimize the number of required experiments. Below‐surface pore air velocity profiles were measured using a newly developed gas tracer tracking method applied at different elevations below the surface. Results suggested that pore air velocity decreases exponentially with decreasing elevation. A simple empirical exponential model was therefore proposed to approximate the pore air velocity–elevation relationship. The pore‐velocity model was chosen such that it, together with an existing, well‐known, semi‐empirical wind‐speed profile model commonly used for describing wind speed in the atmospheric boundary layer, is able to approximate the entire air velocity profile above and below the surface. The pore air velocity profile is described by two empirical parameters that depend on the air velocity at the surface and the shape of the pore velocity profile. The links between the velocity profile shape (via the two empirical parameters), porous medium gas permeability and above‐surface wind speed conditions were analyzed based on the experimental data. Results indicate that below‐surface horizontal pore velocity profiles are strongly related to both above‐surface wind conditions and porous medium characteristics. Highlights: Horizontal air velocity profiles above and below porous medium surfaces were measured A novel approach for approximating wind‐induced, horizontal pore velocity was proposed Medium and wind speed characteristics strongly affected pore velocity profiles Key parameters were: gas permeability > mean wind speed > wind gust frequency

Summary: A simple method for experimental determination of horizontal wind‐induced, near‐surface pore gas velocities in porous media is presented. This method uses traditional tracer gas tracking methodology, but is designed for applications where mass loss of tracer gas from the experimental domain occurs at an unknown rate (as is the case in near‐surface, wind‐exposed porous media), making traditional inverse transport modelling inapplicable. The method was applied to a dry, granular porous medium consisting of crushed basalt (2–4 mm grain size), using CO ₂ as the tracer gas. Experiments were conducted under controlled (wind tunnel) conditions, at average near‐surface wind speeds ranging from 0 to 5.6 ms ⁻¹ , for both gusty and non‐gusty winds. Velocity profiles of near‐surface horizontal pore gas in the top 15 cm of the porous medium were determined for 10 different combinations of average wind speed and wind gust frequency. Pore gas velocities of up to 0.5 cms ⁻¹ were observed that depended on depth and near‐surface wind conditions. Results indicate that horizontal gas flow can occur in wind‐exposed porous media under both non‐gusty and gusty wind conditions, and that wind gustiness has a significant effect on the velocity profiles of horizontal pore gas inside porous media. Results show further that pore gas velocities were almost linearly proportional to near‐surface average wind speed, regardless of porous medium depth. Highlights: Developed a novel method for measuring wind‐induced pore gas velocity in porous media. Near‐surface wind action induced horizontal pore gas velocities. Pore gas velocity profiles depended strongly on wind gustiness (gust frequency). Pore gas velocities were directly proportional to average near‐surface wind speed.