Sheltering effect of vegetation against soil erosion and snow transport

Sheltering effect of vegetation against soil erosion and snow transport (τ - surface τ ) / 0 τ 0-0.6-0.4-0. 0 0. 0.4 0.6 0.8. plant 0 - - - 0 4 6 8 0 x/d INTERIM REPORT No. Benjamin Walter WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland July 00

Table of contents: Overview... 3 Project description... 4. Problem statement and approach... 4. Goals and investigations... 4 3 Experimental setup and instrumentation... 4 4 Wind tunnel studies...5 4. Calibration of Irwin sensors... 5 4. Single obstacle shear stress measurements... 6 5 Past and Present milestones... 7 6 Next steps... 8 7 Summary and Conclusions... 8 References... 8 Attachment : Investment Proposal... 9 Attachment : Poster: ICAR Conference 00...

Overview Title of project: Funded by: Sheltering effect of vegetation against soil erosion and snow transport Vontobel-Stiftung Tödistrasse 7 CH-800 Zürich Project duration: st November 008 3 st October 0 Institute: PhD student: Area of research: Main supervisor: Co-supervisors: Participants: WSL Institute for Snow and Avalanche Research SLF, Davos Benjamin Walter Research Unit: Snow and Permafrost Team: Snow Cover and Micrometeorology Flüelastrasse CH-760 Davos Dorf Tel: +4 / 8 / 47 0 85 Fax: +4 / 8 / 47 0 0 Email: walter@slf.ch Experimental Fluid Dynamics, Meteorology, Physics, Basic and Applied Research Prof. Dr. Fernando Porte-Agel, EPFL, Lausanne Dr. Michael Lehning Dr. Christof Gromke Dr. Katherine Leonard Dr. Frank Graf Dr. Andrew Clifton, Katrin Burri (PhD-Student), Benjamin Eggert (Master Thesis student) Page 3/

Project description. Problem statement and approach A lack of knowledge related to the sheltering effect of vegetation against soil erosion and snow transport exists despite the fact that desertification is a major concern in many countries. Plants influence the erosion, transport and redeposition of soil and snow (Fig. ) by the wind mainly through momentum absorption, local stress concentration, trapping particles in motion and reducing the area of sediment exposed to the wind. Wind tunnel studies using real plants are suggested to investigate these mechanisms. In contrast to field experiments, wind tunnel studies offer the advantage of analysing the mechanisms separately. Prior wind tunnel studies have not used real plants. Real plants have highly irregular structures that can be extremely flexible and porous. They align with the flow at higher wind speeds, resulting in considerable changes to the drag and flow regimes relative to rigid imitations of comparable size. Umean(z) E R O S I O N T R A N S P O R T rolling/creeping saltation suspension REDEPOSITION τs τ Fig. : Turbulent boundary layer over vegetated surface. U mean is the mean velocity, τ is the total shear stress and τ s is the shear stress acting on the surface.. Goals and investigations The overall goal of the project is to understand and quantify the basic mechanisms governing the sheltering effect of vegetation against soil erosion and snow transport. The ability of plants to absorb momentum from the wind results in reduced shear forces acting on the ground. These shear forces are responsible for the onset of soil and snow erosion. The ratio of the total shear stress τ above the plants and the reduced shear stress τ s acting on the ground quantifies the sheltering effect of vegetation (Wolfe, 993; Crawley, 003) (Fig. ). This shear stress partitioning is the main objective of the current research project. The final goal of the study is the development of a model which predicts the shear stress acting on the ground for a given vegetation cover density and wind velocity. Such information can be used as an input for sediment and snow transport models. 3 Experimental setup and instrumentation The setup for the experiments has been built and has already been used for investigating the shear stress distribution around a single real plant using Irwin sensors (Irwin, 980; Wu, 993) (Fig., Fig. 3 and Fig. 7). Experiments with four different planting densities (0, 5.5, 4.5 and 50 plants/m ) will be performed as soon as the self designed multi-channel pressure scanner will be available (Fig. 4 and Fig. 5) (Attachment ). This pressure scanner is necessary since the currently used pressure measurement system does not permit simultaneous measurements at multiple locations. Our custom made multi-channel pressure Page 4/

scanner instead enables simultaneous, highly time-resolved pressure measurements at up to 3 locations. One pressure transducer will thus be available for each Irwin sensor (Fig. 4). The measurement data will contain information about spatial and temporal correlations of the airflow and shear stresses around plants. The pressure scanner additionally allows for a high effective time resolution (nominal sampling frequency 50 Hz per channel). The highly reduced measurement time allows for much more detailed investigations since significantly more setups can be measured, e.g. shear stress distributions at different wind speeds. sensor hole sensor tube Fig. : View inside the wind tunnel with experimental setup (shown here: non perforated boards for sensor calibration). Fig. 3: Irwin sensor flush mounted in a sand rough floor for calibration. Fig. 4: Prototype pressure transducer of the new multichannel pressure scanner connected to an Irwin sensor. Fig. 5: View inside the pressure transducer. 4 Wind tunnel studies 4. Calibration of Irwin sensors Irwin sensors (Irwin, 980) are used to measure the surface shear stress at its position (Fig. 3). The sensor requires no alignment with the flow because of its axial symmetry. The pressure difference p measured between the sensor tube and the sensor hole (Fig. 3) in a turbulent boundary layer can be calibrated on surface shear stress velocities u τ (Fig. 6). The surface shear stress τ surface is given by: τ surface = ρ air u τ. The shear stress velocities u τ were measured using -dimensional hotwire anemometry (Bruun, 995). The calibration data shows excellent correlation to Irwin s original calibration on a smooth floor (Fig. 6a) (Irwin, 980). Further analysis of the calibration procedure showed, that it is beneficial to take the original calibration curve of Irwin for all sensors on smooth floors. It is not necessary to calibrate every single of the 3 Irwin sensors on shear stress velocities since Page 5/

the influence of slight geometrical variations of the sensor dimensions have minor influence on the sensor performance. Quick tests of the sensors functionality and accuracy are sufficient. The setup further allows for measuring shear stress distributions around real plants on sand rough floors where the sensor performance is slightly different compared to the smooth floor case (Fig. 3). For those experiments, a calibration of the Irwin sensors for rough floors has been determined (Fig. 6b). Measurements of the shear stress distribution around a single cylinder for validating the experimental setup and the functionality of the Irwin sensors show good agreement with literature (Sutton, 008). 0 4 data h =,7 / 5 / 7 mm fitted curve Irwin 980 0 4 data h =.7mm fitted curve Irwin 980 u τ h/ν 0 u τ h/ν 0 0 0 0 0 4 0 6 ph /ρν Fig. 6a: Irwin sensor calibration with shear stress velocity u τ on a smooth wooden floor. h is the sensor tube height, ν the kinematic viscosity and ρ the air density. 0 0 0 0 4 0 6 Ph /ρν Fig. 6b: Sensor calibration with shear stress velocity u τ on a sand rough floor (Fig. 3). 4. Single obstacle shear stress measurements Measurements of the shear stress distribution around a single living grass plant (Lolium Perenne) and a single solid cylinder of comparable size have been performed (Fig. 7 and Fig. 8). The basal to frontal area index of the plant and the cylinder as well as the Reynolds number of the two experimental setups have been checked for similarity and show good agreement. Distinctive differences between the shear stress pattern around the plant and the cylinder can be attributed to the influence of the plant s porosity and flexibility. The sheltered zone behind the plant is narrower in cross-stream and longer in streamwise direction than that of the cylinder. For the plant, the lowest shear stresses in the sheltered zone are 50% lower than the mean surface shear stress τ 0 in the undisturbed flow. The sheltering was higher behind the cylinder with shear stress values reduced by 70% relative to background (τ 0 ). Speed-up zones on the sides of the roughness elements experienced peak shear stress values 60% above background for the plant and 30% higher for the cylinder. While the sheltering effect of the plant is smaller in size and magnitude than that of the cylinder, the peak shear stresses in the lateral speed up zones are significantly lower. Since the onset of soil erosion occurs when a critical threshold shear stress is experienced, the lower peak shear stress means that plants provide better protection against soil erosion than rigid elements. This result suggests that parameterizations of flow over vegetated surfaces based on Page 6/

measurements of rigid elements may be incorrect. These results were presented at the 7 th International Conference on Aeolian Research (ICAR) 00 in Santa Rosa, Argentina (Attachment ). Further work will investigate sheltering and shear stress concentrations as a function of cylinder / plant density using real canopies instead of single objects. (τ surface - τ 0 ) / τ 0-0.6-0.4-0. 0 0. 0.4 0.6 0.8. plant y/d 0 - - - 0 4 6 8 0 x/d (τ surface - τ 0 ) / τ 0-0.6-0.4-0. 0 0. 0.4 0.6 0.8. cylinder y/d 0 - Fig. 7: Experimental setup for single plant / cylinder experiments. - - 0 4 6 8 0 x/d Fig. 8: Normalized surface shear stress around a single flexible live plant and a single rigid cylinder. Measurement positions are marked. 5 Past and Present milestones Aug. 09 Sept. 09: Production of the experimental setup (Irwin sensors, perforated boards, plant equipment). Sept. 09 Oct. 09: Oral presentation at the nd International Conference Wind Effects on Trees, October 3 th 6 th 009, Freiburg, Germany Nov. 09 Feb 0: Feb. 0 Mar. 0: Testing and calibration of Irwin sensors on smooth and on rough surfaces. Measurements of shear stress distribution around a single flexible live plant and a single rigid cylinder. Mar. 0 Jul. 0: Writing proposal (Attachment ), planning and construction of the multi-channel pressure scanner. Jun 0 Jul. 0: Poster presentation at the 7 th International Conference on Aeolian Research, July 5 th 9 th 00, Santa Rosa, Argentina (Attachment ). Page 7/

6 Next steps Jul. 0 Aug. 0: Aug. 0 Nov. 0: Oct. 0 Jan : Dec. 0 Mar. : Commissioning of the multi-channel pressure scanner. Measurements of shear stress distribution in different dense plant canopies at various wind velocities. Publication of results in international journals. Further experiments on the influence of flexibility and porosity of roughness elements on shear stress distribution and air flow. 7 Summary and Conclusions The project is proceeding well and the first results after the development of the experiment setup are very promising. The calibration of the Irwin sensors on smooth wooden floors shows excellent correlation to the original calibration presented in Irwin (980). Measurements to validate the experimental setup and the functionality of the Irwin sensors show good agreement with literature. First results on the measured shear stress distribution around a single plant show distinct differences to well-investigated shear stress distributions around rigid cylinders (Attachment ). The project is a few months behind the original schedule due to difficulties with the calibration procedure of the Irwin sensors and the need for a new multi channel pressure scanner (Attachment ). However, the new multi-channel pressure scanner will allow for much more detailed investigations on the sheltering effect of plants against soil erosion and snow transport. References Bruun, H. H., 995: Hot-Wire Anemometry, principles and signal analysis, Oxford Science Publications, pp. 507 Crawley, D. M., Nickling, W. G., 003: Drag partition for regularly-arrayed rough surfaces, Boundary-Layer Meteorology, 07, 445-468 Irwin, H.P.A.H., 980: A simple omnidirectional sensor for wind-tunnel studies of pedestrian-level winds, Journal of Wind Engineering and Industrial Aerodynamics, 7, 9-39 Sutton, S.L.F., McKenna-Neuman, C., 008: Variation in bed level shear stress on surface sheltered by nonerodible roughness elements, Journal of Geophysical Research, 3 Wolfe, S. A., Nickling, W. G., 993: The protective role of sparse vegetation in wind erosion, Progress in Physical Geography, 7, 50-68 Wu, H., Stathopoulos, T., 993: Further experiments on Irwin s wind sensor, Journal of Wind Engineering and Industrial Aerodynamics, 53, 44-45 Page 8/

Attachment : Investment Proposal Presented to: Eidg. Forschungsanstalt für Wald, Schnee und Landschaft WSL Handed in: March 00 Accepted: April 00 Proposal: Upgrade of pressure measurement system and supporting hardware for the SLF wind tunnel Summary: This is an application for financial support of a multi-channel pressure scanner including a PC for data acquisition to be employed in the SLF wind tunnel. The device will significantly expand the facilities existing capabilities and will first be used for measuring shear stress distributions around live plants. The current pressure measurement system does not permit simultaneous measurements at multiple locations. Multi-channel pressure scanners instead enable simultaneous, highly time-resolved pressure measurements at up to 3 locations. Further, we want to apply for one additional computer including data acquisition devices to update the wind tunnel infrastructure. The old data acquisition does not fulfil the requirements any more. The pressure scanner and new data acquisition will also allow future projects (e.g. soil erosion Tibet) to run much more efficiently. A summary of the parts and prices can be found in Table, quotes and the PC migration schedule are attached. Table : Summary of parts and prices Item No. Cost / Item [CHF] Total Cost [CHF] Pressure Scanner 7.750,00 7.750,00 host computer.9,30.9,30 div. cables 5 50,00 50,00 Additional Computer.9,30.9,30 NI PCI-6034E + cable.449,00 79,00.449,00 358,00 NI PCI-GPIB + cable 999,00 999,00 NI PCI-843/4 + cable 949,00 69,00 949,00 69,00 div. cables 5 50,00 50,00 taxes (7.6%) shipping (estimated.5%).6,96 57,69 Total: 37.653,5 Justification: Pressure scanning is a crucial measurement technique in wind tunnel investigations. At the SLF boundary layer wind tunnel, pressure scanning is currently used for measuring shear stresses, pressure gradients along the wind tunnel and flow velocities. Project / Measurement description: Within the project Sheltering effect of vegetation against soil erosion and snow transport, shear stresses acting on the ground surface around live plants are measured using Irwin sensors (Figure ). This shear stress distribution is of great interest as it determines the onset and magnitude of differential soil erosion on rough or vegetated surfaces. The shear stress distribution around live plants has never been measured systematically before. Such information gives very useful input for sediment, e.g. soil, sand or snow transport models. Page 9/

Current situation: Up to 3 Irwin sensors (already built by the SLF workshop) are placed in vegetation canopies of different densities to measure the shear stress distribution on the ground surface. This is achieved by measuring the pressure difference between port No. and port No. of each Irwin sensor (Figure ). Currently, a pressure measurement system consisting of a single differential pressure transducer (model: MKS Baratron 0 AD) with a nominal sampling frequency of 0 Hz is used. The existing device enables the measurement of only one Irwin sensor at a time. The data evaluation is consequently restricted to time-averaged statistics and does not allow for spatial and temporal correlation analyses. Switching between Irwin sensors includes manual tube changing which is time-consuming. Wind Figure : Flush mounted Irwin sensor. The pressure difference between the sensor tube protruding into the wind (port No. ) and the sensor hole measuring pressures on the ground surface (port No. ) is calibrated to shear stresses. Proposed Solution: The multi-channel pressure scanner applied for can measure up to 3 differential pressures simultaneously. One pressure transducer will thus be available for each Irwin sensor. The measurement data will contain the desired and very useful information about spatial and temporal correlations of the airflow and shear stresses around plants. The pressure scanner additionally allows for a high effective time resolution (nominal sampling frequency 500 Hz per channel). The highly reduced measurement time allows for much more detailed investigations since significantly more setups can be measured, e.g. shear stress distributions at different wind speeds. Further applications: The multi-channel pressure scanner will also be applied for measuring streamwise pressure gradients and flow velocities. Static pressure differences between 4 pressure taps distributed along the wind tunnel to one single reference port at the inlet of the tunnel can be measured to identify the pressure gradient along the tunnel. With the multi-channel pressure scanner, all 4 static pressure differences can be measured simultaneously. A pressure rake consisting of pressure taps is used to determine vertical mean velocity profiles and their low frequency fluctuations. With the multi-channel pressure scanner, all pressure taps can be measured simultaneously with a high effective time resolution allowing for analysis of spatial and temporal correlations of the air flow. In both cases, the required measurement time will be reduced significantly. Instrumentation: The new pressure scanner requires an up-to-date host computer for the operation of the new system. Two linked computers are currently used for measuring various instruments at different frequencies. These computers, installed in 00/003, do not meet the requirements of the proposed multi-channel pressure scanner. A renewal of both out-of-date computers is required in conjunction with the purchase of a new pressure scanner due to the linking between the two machines. Additionally, analogue-digital converters, data acquisition devices, connectors and cables are needed for the pressure scanner and new computers. A summary of the parts and prices can be found in Table, quotes are attached. Page 0/

Attachment : Poster: ICAR Conference 00 Surface Shear Stresses Around a Single Flexible Live Plant Compared to a Rigid Cylinder Benjamin Walter (walter@slf.ch), C. Gromke, K. Leonard, A. Clifton, M. Lehning WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland Introduction - Erosion, transport and redeposition (Fig. ) of soil and snow particles contribute to land degradation and desertification in arid and semiarid regions. Umean(z) E R O SIO N T R A N S P O R T rolling/creeping Experimental Setup - The SLF boundary layer Wind Tunnel: open circuit tunnel, 4m long, m cross sectional area, operates in suction mode, velocities from 0-0m/s. - Measurements of shear stress distributions around a single rigid cylinder and a single flexible live plant were done using Irwin sensors (Fig. ). Fig. : Measurement of shear stress distribution. - Irwin sensor (Fig. 3a): requires no alignment with the flow because of its axial symmetry. Can be used for measuring surface shear stress velocities u τ (surface shear stress: τ ρ u ). - Pressure difference p between sensor tube and sensor hole (Fig. 3a) in a turbulent boundary layer were calibrated against surface shear stress velocities u τ (Fig. 3b). - Shear stress velocities u τ for sensor calibration on smooth wooden floor were measured using - dimensional hotwire anemometry. - Calibration data shows excellent correlation to Irwin s original calibration: Our calibration: Irwins calibration: saltation uτ h = ν uτ h = ν suspension REDEPOSITION Fig. : Sketch of aeolian mechanisms. - Plants influence these processes mainly through momentum absorption, local stress concentration, trapping particles in motion and reducing the area of sediment exposed to the wind. - The onset of erosion depends on the magnitude and distribution of surface shear stresses τ s on the ground beneath plant canopies. - Prior studies of shear stress distributions on rough surfaces have not used real / living plants. - Real plants (lolium perenne) have highly irregular structures that can be extremely flexible and porous. surface = 0.453 8 ph.6 0.89 + ρν 0.453 8 ph.0 0.93 + ρν τs air τ τ ( axis exaggerated ) Fig. 3: a.) Sketch of Irwin sensor and b.) sensor calibration with shear stress velocity u τ. y/d ( axis exaggerated ) y/d IRWIN SENSOR sensor tube τ = f ( P, h, ν, ρ) 0 - (τ surface - τ 0 ) / τ 0-0.5 0 0.5 - - 0 4 6 8 0 x/d (τ surface - τ 0 ) / τ 0-0.5 0 0.5 0 - s } sensor hole h Shear stress distribution - - 0 4 6 8 0 x/d Fig. 4: Normalized surface shear stress distribution around a.) a single rigid cylinder and b.) a single flexible live plant. Tab. : Comparison of cylinder and plant geometry. - σ = element u τ h/ν 0 4 0 3 0 0 Cylinder Plant Ø [cm] 5 4 7 height [cm] 9 0 4 σ 0.5 0. 0.6 ReØ 4,000 33,000 58,000 element basal area frontal area cylinder plant - Mean surface shear stress in the undisturbed flow for both cases was τ 0 0.5 N/m. - Shear Stress at position of cylinder/plant is assumed to be τ 0. - Limited measurement resolution directly around the elements. Resolution is fine enough to resolve the shear stress distribution further away from the elements. - Measurement positions are marked. data h =,7 / 5 / 7 mm fitted curve Irwin 980 0 0 0 0 4 0 6 ph /ρν - Basal to frontal area indexe σ as well as Reynolds number Re Ø agree between experiments. Results Differences in shear stress patterns that can be attributed to the influence of the plant s porosity and flexibility (Fig. 4): - The sheltered zone behind the plant is narrower in cross-stream and longer in streamwise direction than that of the cylinder. - Lowest shear stresses in the sheltered zone are 50% (plant) and 70% (cylinder) lower than the surface shear stress τ 0 in the undisturbed flow. - Peak shear stress values τ peak are 60% (plant) and 30% (cylinder) above background (τ 0 ). τ peak [N/m ] 0.4 0.35 0.3 0.5 0. 0.5 0. τ 0 no element plant cylinder Fig. 5: Peak shear stress values (left) and total shear force F on surface of evaluated area in percent of the no element case (right). Impact of Results Outlook 05 00 95 90 85 80 75 F [%] While the sheltering effect of the plant is smaller in size and magnitude than that of the cylinder, the peak shear stresses in the lateral speed up zones are significantly lower (Fig. 5). Since the onset of soil erosion occurs when a critical threshold shear stress is experienced, the lower peak shear stress means that plants provide better protection against soil erosion than rigid elements. Thus, parameterizations of flow over vegetated surfaces based on measurements of rigid elements may overestimate erosion. - Measurements of shear stress distributions in real plant canopies (Fig.6) of different densities at multiple wind velocities. Fig. 6: Wind tunnel test section covered with real plants of a density of ρ 5 plants/sqm ( λ 0. ), ( Burri et al., in prep.). - Custom multi channel pressure scanner that allows for simultaneous measurements of 3 Irwin sensors with frequencies up to f = 50Hz. Funded by: Vontobel-Stiftung and the Swiss National Science Foundation Page /