Determination of surface energy fluxes over small water bodies using micrometeorological measurements presents a number of challenges relating to terrain complexity, fetch and homogeneity. Water storages vary greatly in spatial scale, from small irrigation dams on farms to large reservoirs essential in managing water for entire regions. However, studies of open water evaporation tend to be biased towards larger bodies of water (Rosenberry et al., 2007). This is partly due to the theoretical and practical problems associated with determining surface fluxes in airflow above small water bodies. Over a small reservoir with limited upwind fetch the measured signal can be contaminated by fluxes originating from upwind of the reservoir. In addition, the dimensions of small reservoirs means that direct measurements of surface fluxes can't always be made in airflow that meets the conditions of idealized flow (planar-homogeneous, stationary and in the absence of advection or subsidence) (Assouline et al., 2008).
The ability to accurately measure surface energy balance components in the complex environment of a small reservoir in south-east Queensland, Australia, using the Eddy covariance (EC) and Scintillometry techniques was assessed. The initial focus was on how frequently airflow above a small reservoir could be assumed to be stationary, homogenous and without subsidence. On-water measurements of mean and turbulent flow characteristics were made using a sonic anemometer situated above the water surface while upwind and downwind measurements of the wind profile were made using four tower mounted anemometers.
Results suggested that the main cause of turbulence above the water surface was from land-based sources. Analysis showed that a logarithmic wind profile was a reasonably common occurrence immediately upwind and downwind of the reservoir and at the EC system. In addition, close agreement was observed between upwind, downwind and central measurements of friction velocity. It was found that in moderately strong winds turbulence could be advected hundreds of meters downwind and remain homogenous and fully developed. However, non-ideal airflow conditions were observed during light winds. Airflow approaching from a direction where trees were present was more frequently stationary, homogenous and without subsidence than airflow approaching from a direction where terrain was smoother. However, these conditions were also the most conducive to producing instances of significant subsidence. Therefore, it is expected that EC and scintillometry measurements of water surfaceatmosphere energy exchanges conducted on small reservoirs will have a reasonable amount of data elimination due to non-ideal flow conditions.
For surface energy flux measurements to be performed accurately using techniques such as EC and scintillometry, it is also necessary to have some knowledge about the spatial context that the measurements represent. To insure that measurements from the reservoir are not affected by non-local advection it is essential that the spatial dimensions of the measurement footprint lie above the water surface. To determine the footprints of EC and scintillometry measurements a simple method for estimating the spatial dimensions of footprints in complex terrain using the SCADIS two-dimensional footprint calculator was used. The SCADIS footprint model is designed to estimate measurement footprints in complex terrain where there is a variety of surface types causing turbulence production.
The techniques outlined in this research are ideally suited for researchers wishing to make surface energy flux measurements in complex terrain but who also do not have the time or resources to apply a more complex footprinting approach. This study represents the first time the spatial dimensions of scintillometer footprints have been estimated in complex terrain. It was found that in neutral atmospheric stability the majority of the contribution to the footprints for both EC and scintillometry, originated from the water surface, due largely to the rough terrain surrounding the reservoir (e.g. Figure 1). The increased mechanical turbulence over the lake, caused by the interaction of the airflow with trees upwind of the site, resulted in a reduction in footprint size when compared to the situation over a larger lake. However, analysis suggested that possible instances of footprint contamination at small reservoirs may occur during conditions of moderate to intense stability.
Figure 1: Surface colour plots of footprint function at Logan's Dam for (a) the EC system and (b) the scintillometer. Both plots are for a south-easterly geostrophic wind of 10 m s-1 in neutral atmospheric conditions with instrument heights of 2 m.