Proposing a Representative Scale for Aggregating Surface Energy Fluxes for the Qinghai-Tibet Plateau Based on Fine-Scale Modeling and Topographical Analysis

Abstract The study focuses on the scale dependence of surface energy fluxes in topographically diverse landscapes, where variations in soil moisture and vegetation types significantly influence evapotranspiration processes. Land-atmosphere interaction plays a major role in atmospheric processes in the Qinghai-Tibet Plateau (QTP) region, and the inability of current land-atmosphere models to capture the landscape-scale heterogeneity in the latent heat flux introduces uncertainties in these models. To address this issue, the remote sensing-based fine-scale energy balance modeling (METRIC) was combined here with topographical analysis (2D FFT spectral analysis and LUMP hillslope discretization) to identify a Representative Elementary Area (REA) for latent heat flux. A scale dependence analysis within a 1-10 km range is performed, and two flux aggregation methods are proposed: the Mechanistic Approach to define REA, which identifies optimal aggregation areas for latent heat through window size variations, and the Landscape-Explicit Approach, which upscales latent heat based on hillslope elements of dominant quasi-periodic topographic features. It was established that the representative elementary area for latent heat at QTP is in the 2-4 km range. Moreover, REA is formed by characteristic topographical patterns that can be defined through regularity and repetition, which can be seen in representative hillslopes. Benchmark simulations conducted with the METRIC model indicate that ignoring topographical influences results in the absence of REA. The Landscape-Explicit Approach may offer greater interpretability and predictive capability for future land-atmosphere model development as it considers topographic effects on lateral water redistribution and surface energy transfer.