Linking water and carbon cycles: modeling latent heat exchange and dissolved organic carbon

Abstract

In this study latent heat flux (kE) measurements made at 65 boreal and arctic eddy-covariance (EC) sites were analyses by using the Penman–Monteith equation. Sites were stratified into nine different ecosystem types: harvested and burnt forest areas, pine forests, spruce or fir forests, Douglas-fir forests, broadleaf deciduous forests, larch forests, wetlands, tundra and natural grasslands. The Penman–Monteith equation was calibrated with variable surface resistances against half-hourly eddy-covariance data and clear differences between ecosystem types were observed. Based on the modeled behavior of surface and aerodynamic resistances, surface resistance tightly control kE in most mature forests, while it had less importance in ecosystems having shorter vegetation like young or recently harvested forests, grasslands, wetlands and tundra. The parameters of the Penman–Monteith equation were clearly different for winter and summer conditions, indicating that phenological effects on surface resistance are important. We also compared the simulated kE of different ecosystem types under meteorological conditions at one site. Values of kE varied between 15% and 38% of the net radiation in the simulations with mean ecosystem parameters. In general, the simulations suggest that kE is higher from forested ecosystems than from grasslands, wetlands or tundra-type ecosystems. Forests showed usually a tighter stomatal control of kE as indicated by a pronounced sensitivity of surface resistance to atmospheric vapor pressure deficit. Nevertheless, the surface resistance of forests was lower than for open vegetation types including wetlands. Tundra and wetlands had higher surface resistances, which were less sensitive to vapor pressure deficits. The results indicate that the variation in surface resistance within and between different vegetation types might play a significant role in energy exchange between terrestrial ecosystems and atmosphere. These results suggest the need to take into account vegetation type and phenology in energy exchange modeling.