Critical Evaluation of Scalar Roughness Length Parametrizations Over a Melting Valley Glacier

Xiaofeng Guo,Kun Yang,Long Zhao,Wei Yang,Shenghai Li,Meilin Zhu,Tandong Yao,Yingying Chen

DOI: https://doi.org/10.1007/s10546-010-9586-9

2011-01-01

Boundary-Layer Meteorology

Abstract:We present a field investigation over a melting valley glacier on the Tibetan Plateau. In the ablation zone, aerodynamic roughness lengths ( z 0 M ) vary on the order of 10 −4 –10 −2 m, whose evolution corresponds to three melt phases with distinct surface cover and moisture exchange: snow (sublimation/evaporation), bare ice (deposition/condensation), and ice hummocks (sublimation/evaporation). Bowen-ratio similarity is validated in the stably stratified katabatic winds, which suggests a useful means for data quality check. A roughness sublayer is regarded as irrelevant to the present ablation season, because selected characteristics of scalar turbulence over smooth snow are quite similar to those over hummocky ice. We evaluate three parametrizations of the scalar roughness lengths ( z 0 T for temperature and z 0 q for humidity), viz. key factors for the accurate estimation of sensible heat and latent heat fluxes using the bulk aerodynamic method. The first approach is based on surface-renewal models and has been widely applied in glaciated areas; the second has never received application over an ice/snow surface, despite its validity in (semi-)arid regions; the third, a derivative of the first, is proposed for use specifically over rough ice defined as z 0 M > 10 −3 m or so. This empirical z 0 M threshold value is deemed of general relevance to glaciated areas (e.g. ice sheet/cap and valley/outlet glaciers), above which the first approach gives notably underestimated z 0 T , q . The first and the third approaches tend to underestimate and overestimate turbulent heat/moisture exchange, respectively, frequently leading to relative errors higher than 30%. Comparatively, the second approach produces fairly low errors in energy flux estimates both in individual melt phases and over the whole ablation season; it thus emerges as a practically useful choice to parametrize z 0 T , q in glaciated areas. Moreover, we find all three candidate parametrizations unable to predict diurnal variations in the excess resistances to humidity transfer, thus encouraging more efforts for improvement.