Strong Skin Cooling and Its Impacts on Lake Thermal Processes in a Large Lake on the Tibetan Plateau

Abstract

Abstract Skin cooling, wherein the surface temperature of a water body Tskin is lower than the temperature below the surface, is a widespread phenomenon. Previous studies have almost ignored this effect on the Tibetan Plateau (TP), despite the presence of thousands of lakes on the TP and the fact that extraordinary solar heating leads to very strong energy exchanges on the lake surfaces. This study utilizes in situ observations and MODIS-derived Tskin data at Lake Nam Co, one of the largest lakes on the TP, to quantify the skin cooling effect. The observed nighttime skin cooling is approximately 0.52°C on average, with the maximum of about 1°C, during the lake water turnover period (from October to mid-November), which obviously surpasses reported values for oceans (less than 0.4°C). To understand the impact of the skin cooling on the lake thermal processes, a skin cooling parameterization is validated and incorporated into the WRF-lake model. Simulations with the updated model show that accounting for the skin cooling process systematically lowers sensible and latent heat fluxes by a few watts per square meter, which yields an increase in water temperature by 0.45°C at the end of December and may delay the onset of lake freeze. Finally, we show that the inclusion of the skin cooling process in a lake model needs simultaneous adjustment of the parameterization of heat/water vapor transfer. Significance Statement Skin cooling is a widespread phenomenon for a water surface, and its intensity depends on the energy flux exchange of the water surface. The Tibetan Plateau possesses the presence of thousands of lakes, but early studies have ignored the skin cooling effect. We found that the nighttime skin cooling magnitude during the lake water turnover period in this region obviously exceeds reported values for oceans, due to the strong surface energy exchange in the Tibetan Plateau. Neglecting the skin cooling process may lead to systematic overestimation of turbulent heat fluxes and underestimation of water temperature. We highlight that accounting for this skin cooling process is crucial to select appropriate parameterization schemes for heat/water vapor transfer in lake thermal process modeling.