Carbon storage, processing, and transport in freshwater systems are important components of the global carbon cycle and sensitive to global change. However, in large-scale modeling this part of the boundless carbon cycle is often lacking or represented in a very simplified way. A new process-oriented lake biogeochemical model is used for investigating impacts of changes in atmospheric CO2 concentrations and organic carbon loading from the catchment on future greenhouse gas emissions from lakes across two boreal to subarctic regions (Northern Sweden and Alaska). Aquatic processes represented include carbon, oxygen, phytoplankton, and nutrient dynamics leading to CO2 and CH4 exchanges with the atmosphere. The model is running inside a macroscale hydrological model and may be easily implemented into a land surface scheme. Model evaluation demonstrates the validity in terms of average concentration of nutrients, algal biomass, and organic and inorganic carbon. Cumulative annual emissions of CH4 and CO2, as well as pathways of CH4 emissions, also compare well to observations. Model calculations imply that lake emissions of CH4 may increase by up to 45% under the Representative Concentration Pathway 8.5 scenario until 2100, and CO2 emissions may increase by up to 80% in Alaska. Increasing organic carbon loading to the lakes resulted in a linear response in CO2 and CH4 emissions across both regions, but increases in CO2 emissions from subarctic lakes in Sweden were lower than for southern boreal lakes, probably due to the higher importance of imported vegetation-generated inorganic carbon for CO2 emission from subarctic lakes.

Effects of the short-term temporal variability of meteorological variables on soil temperature in northern high-latitude regions have been investigated. For this, a process-oriented land surface model has been driven using an artificially manipulated climate dataset. Short-term climate variability mainly impacts snow depth, and the thermal diffusivity of lichens and bryophytes. These impacts of climate variability on insulating surface layers together substantially alter the heat exchange between atmosphere and soil. As a result, soil temperature is 0.1 to 0.8 degrees C higher when climate variability is reduced. Earth system models project warming of the Arctic region but also increasing variability of meteorological variables and more often extreme meteorological events. Therefore, our results show that projected future increases in permafrost temperature and active-layer thickness in response to climate change will be lower (i) when taking into account future changes in short-term variability of meteorological variables and (ii) when representing dynamic snow and lichen and bryophyte functions in land surface models.