Application and Evaluation of a Two-Wavelength Scintillometry System for Operation in a Complex Shallow Boreal-Forested Valley

Two-wavelength scintillometer systems can provide much needed measurements of area-averaged sensible and latent heat fluxes. However, these devices rarely have been deployed on canopy-covered complex terrain, and never in the circumpolar boreal biome, where large-scale fluxes are essential to hydroclimate modellers. We present a comparison of fluxes measured above a boreal-forested valley with a two-wavelength scintillometer and an eddy covariance system. Instruments were deployed in late summer 2017, and 19 days of data were retained for the analysis. The scintillometer path was 1347-m long and projected across the valley between 5 and 100 m above the ground, with an effective height of ≈\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\approx $$\end{document}88 m. The limitations of deriving surface fluxes using scintillometry in complex terrain are discussed, and the effects of atmospheric conditions on the flux comparison are quantified. Fluxes are calculated with the scintillometer only, and using a number of atmospheric variables from the eddy-covariance system; impacts of these calculation methods on the correlation between instrumental systems are assessed. Despite a weak agreement of structure parameters between instruments, the comparison of scintillometer and eddy-covariance fluxes yields good correlation (R2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^2$$\end{document} up to 0.82). Scintillometry correlates best with eddy-covariance data when the atmospheric surface-layer top is above the scintillometer effective height, but R2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^2$$\end{document} only drops slightly otherwise (average decrease of 0.11). The validity of scintillometer fluxes appears dubious during night-time and stable periods. We show that area-averaged flux measurements using two-wavelength scintillometers are possible in hilly forests, but more studies are needed to pinpoint the best methodological framework.