Scintillometry and the Surface Energy Balance: Spatial and Temporal Scales for Energy Closure During IPAQS.

Measurements of the surface energy balance (SEB) using eddy-covariance (EC) typically result in a 10-20% energy deficit. Some of the causes for this imbalance are believed to be due to measurement errors, inadequate sampling of large scale motions, and flux divergence. The imbalance is observed even in experiments that are explicitly designed to measure all aspects of the SEB. A measurement technique that has been hypothesized to close the energy balance better is scintillometry. The majority of scintillometer studies report an overestimation of fluxes when compared to EC. One explanation for the overestimation is better closure of the energy balance at larger scales. The scintillometer beam spans horizontally of the O(1 km) sampling many more eddies per measurement than typical EC setups. We explore how spatial and temporal scales affect the SEB closure on a 1 km^2 array of EC stations where thermal heterogeneities dominate. In previous studies scintillometry typically compared more favorably to an ensemble of EC stations, even over seemingly homogeneous terrain.

To explore this hypothesis, we use data from the Idealized Planar Array for Quantifying Surface heterogeneity (IPAQS). These data were gathered during Jun-Jul 2019 at the SLTEST site within the Dugway Proving Ground in Utah. The SLTEST site is a dry lake bed, which results in a very flat surface of uniform roughness. Heterogeneity comes from variations in surface temperature which can vary by as much as 15 K on scales of O(10 m). IPAQS consists of a 1 km x 1 km 'grid cell' with a plethora of instrumentation to calculate the SEB as well as horizontal, and vertical gradients of scalars. The main array is a 4x4 grid of sonic anemometers mounted at 2 m with a 200 m spacing between stations. A two-wavelength scintillometer spans 1150 m across the array providing an overall grid average.

Within the array, the sensible heat flux could vary by upwards of 100 W m^{-2} during the day. This heterogeneity not only means that a single station could compare poorly with the scintillometer, but also the SEB closure would vary within the array. Which raises a fundamental question: how many stations within a 1 km^2 array must be averaged to get better closure of the SEB? Subsequently the effect of the SEB closure on scintillometry v.s. EC fluxes can be quantified.