Quantifying the Spatial Representativeness of Carbon Flux Footprints of a Grassland Ecosystem in the Semi‐Arid Region

Abstract

AbstractQuantifying the single‐site representativeness of carbon flux footprints plays a crucial role in land‐atmosphere interaction, especially in semi‐arid regions with high‐frequency turbulence. In this study, we used multi‐platform datasets, including observational data derived from the eddy covariance flux monitoring systems of the Semi‐Arid Climate and Environment Observatory of Lanzhou University (SACOL) and multivariate satellite remote sensing, to explore the distribution characteristics of carbon flux footprints during 2007–2016, using a Flux Footprint Prediction model. The relative importance of atmospheric boundary layer factors and vegetation factors on the area of carbon flux footprints was quantified by correlation analysis, multiple stepwise regression, and random forest. Physical mechanisms affecting the differences in the spatial distribution of carbon flux footprints were also analyzed. The results show that the carbon flux footprints are usually distributed on the prevailing wind side of the station, whose main contributing source area is the relatively flat grassland at the summit range of SACOL, with a spatial representativeness of 67.4–507.2 m in length and 54,982–105,329 m2 in area. Mechanism studies have shown that the frictional velocity, vegetation Index, atmospheric instability parameters, and wind speed have significant importance on the area of carbon flux footprints due to dynamic and thermal forcing, with frictional velocity being dominant, while wind direction has a major effect on the profile shape. This study provides references for the representativeness of site fluxes for semi‐arid regions dominated by grassland ecosystems where observations are often scarce.