1. Measuring Turbulent Flows over a Forested Hill Rosey Grant¹, Jennifer Hutton¹, Ian Brooks¹, Barry Gardiner², Kevin Jones³, Stephen Mobbs¹, Andrew Ross¹ ¹School of Earth and Environment, University of Leeds, Leeds, UK. ²Forest Management Division, Forestry Commission, Midlothian, UK, ³School of Geosciences, University of Edinburgh, Edinburgh, UK Acknowledgements We would like to thank all those who helped out with the field campaign and with work relating to this project so far. Special thanks to Anna Pederzoli, Bob Logan, Cathryn Birch, Clare Grant, Flea Perry, James Groves, Matt Hobby, Matt Woodhouse, Nick Hewitt and Tor Smith without whose help there would have been a lot of left over pie. This project is funded by NERC. 1 Introduction The effect of complex terrain on air flow has been widely studied and the theory surrounding the problem under idealized conditions is well developed (e.g. Jackson and Hunt, 1975). The central features of turbulent flows through forest canopies are also well documented (e.g. Kaimal and Finnigan, 1994 and Finnigan, 2000). However, despite the fact that a significant portion of the worlds mountainous terrain is forested the interactions between the air flow and canopy over complex terrain have not yet been fully investigated. The intent of this work is to introduce a field campaign conducted by the University of Leeds in the winter 2006-07 in collaboration with the Forestry Commission, the University of Edinburgh and the UK Met Office. The measurements collected will provide a dataset for validating some of the latest model developments and theories surrounding the problem of canopy air flow interaction over complex terrain and will also be useful for improving numerical weather prediction schemes and estimating wind damage to commercial forests. 7 References Finnigan, J. J., 2000. "Turbulence in plant canopies". Annual Reviews in Fluid Mechanics, 32, 519- 571. Finnigan, J. J. and S. E. Belcher, 2004. "Flow over a hill covered with a plant canopy". Quarterly Journal of the Royal Meteorological Society, 130, 1-29. Jackson, P. S. and J. C. R. Hunt, 1975. "Turbulent wind flow over a low hill". Quarterly Journal of the Royal Meteorological Society, 101, 929-955. Kaimal, J. C. and J. J. Finnigan, 1994. Atmospheric boundary layer flows: their structure and measurement. Oxford University Press. Ross, A. N. and S. B. Vosper, 2004. "Neutral turbulent flow over forested hills". Quarterly Journal of the Royal Meteorological Society, 131, 1841-1862. 2 Theory An analytical model created by Finnigan and Belcher (2004) has predicted the following unique effects in boundary layer flows over a hill covered with a canopy- The pressure gradient associated with the flow over a hill causes additional perturbations to the flow within the canopy. The pressure gradient within the canopy is partly balanced by the canopy drag. In the upper reaches of the canopy downwards turbulent transport of momentum is also important. Deep within the canopy this balancing of the pressure gradient will produce maximum streamwise winds on the upwind slope of the hill while in the upper reaches of the canopy the maximum streamwise winds are near the crest of the hill. In the lower reaches of the canopy on the lee of the hill the pressure gradient causes decreased wind speeds and, if the canopy is sufficiently deep this leads to flow separation with a reversal. The presence of a canopy on a hill will decrease the expected speedup over the hill due to the extra turbulent mixing within the canopy and the maximum velocity will occur further upwind of the hill crest than on an unforested hill. 6 Conclusions The motivation behind this field campaign was to collect a dataset that could validate the latest model developments surrounding the problem of flow over a forested hill. Four months of data was successfully collected and is now being analysed. Initial results seem to indicate that some of the features predicted by Finnigan and Belcher (2004) can be observed. The coming months will see a full analysis of the data collected during the campaign. This data set should provide a new and valuable set of measurements for understanding flow and turbulence within forest canopies and for validating our current models of these processes. Figure 2.2 1:25000 OS Map of the field site with individual instrument sites labelled. T represents the towers and AR the AWS sites.© Crown Copyright/database right 2005. An Ordnance Survey/EDINA supplied service. Figure 2.1 Map of Britain with the Scottish island of Arran labelled. 5 Case Study Initial data analysis has indicated that there is a potentially interesting case study from 27.04.07 with easterly winds and a full data set. The plot in figure 4 shows averaged wind vectors from 7 AWS sites located within the forest. On the east slope winds are predominantly easterly, aligned with the synoptic flow. On the western leeward slope flow separation is indicated by the abrupt change in direction, directly below the ridge summit. This flow separation is a ubiquitous feature of both numerical and analytical results for forested hills. The University of Edinburgh The field campaign took place on the Scottish island of Arran, located approximately 14miles off the south-west coast of mainland Scotland (Figure 2.1). The field site is a ridge on the north-east coast of the island, 1.5km in length and varying in height from sea level to 261m. The campaign was run over a period of 8 months from October 2006 to May 2007 with intensive data collection taking place from February until May 2007. The winter and spring months were chosen for data collection because winds at this time of year are stronger and will provide clearer indications of the flow dynamics. The southern end of the ridge is forested with a majority of coniferous trees while the northern end is unforested, allowing a direct comparison between flows over a forested and flows over an unforested hill to be made. 4 Field Campaign Vertical Profiles Three towers of 18m, 18m and 22m were erected at the sites labelled on figure 2.2. Purpose: to provide vertical velocity and temperature profiles within and above the canopy. Turbulent momentum and heat fluxes will be calculated from this data. Canopy heights:8-12m Instrumentation: four 3-D sonic anemometers, five temperature sensors, five cup-anemometers. Logging frequency: 10Hz Surface Measurements 12 automatic weather stations (AWS) were placed at each of the sites marked AR in figure 2.2. Purpose: to provide data showing the dynamics of surface flows within and outside the canopy over the ridge. Measurements: wind speed and wind direction at 2m, temperature, relative humidity, pressure. Logging frequency: 3seconds. Figure 3.1 View looking up tower T1 showing 3-D sonic anemometers. Figure 3.2 AWS at site ARJ, located on the coast on the east side of the ridge. Figure 4 Wind vectors from 7 AWS sites over the southern, forested end of the ridge. Measurements are averaged over 14:00-15:00 on 27.04.07. Also shown are contours of height. Numerical simulations of flow over a forested hill produced by Ross and Vosper (2004) have provided evidence to support the theories of Finnigan and Belcher (2004). For small hills though the linear theory breaks down. This leads to significant flow into / out of the canopy and a shift in the pressure field relative to the hill. The plot in figure 1 illustrates these features including the flow separation within the canopy over the lee slope. Figure 1 Numerical simulation showing U velocity and streamlines for a forested hill. This illustrates the predicted flow separation and flow in and out of the canopy. 3 The Numerical Model U (msˉ¹) Figure 5 shows vertical profiles of wind speed at each of the towers. The data was taken from the same period as in figure 4. The plot shows a variation across the hill with T2 (positioned on the summit of the ridge) showing a much greater increase in wind speed with height than T1 and T3. Profiles are consistent with measurements over flat terrain and with numerical simulations. T3 in particular displays the expected logarithmic increase above the canopy (~12m) and an exponential decrease within the canopy. Figure 5 Vertical wind speed profiles from the 3 tower sites. The 3-D sonic anemometer measurements are denoted by +, the cup anemometer measurements are denoted by x. Measurements are averaged over 14:00-15:00 on 27.04.07. T1 T2 T3