About NGEE Arctic

The Arctic may be the most climatically sensitive region on Earth. High latitudes have experienced the greatest regional warming in recent decades and are projected to warm twice as much as the rest of the globe by the end of the twenty-first century (Allison et al. 2009). These areas are uniquely characterized by the presence of permafrost, defined as ground that has been continuously frozen for two or more years. Observations suggest that permafrost degradation is now common in high-latitude ecosystems (Jorgenson et al. 2006) and is expected to drive changes in climate forcing through biogeochemical and biophysical feedbacks. Biogeochemical feedbacks are dominated by the potential to release a large amount of currently stored carbon back into the atmosphere as CO2 and CH4 (Zimov et al. 2006, Schuur et al. 2009), whereas biophysical feedbacks include terrestrial energy budgets that are changing in response to warming in high-latitude ecosystems (Chapin et al. 2005, Euskirchen et al. 2009). These feedbacks will take place in an environment undergoing dramatic geomorphic change and landscape reorganization (Rowland et al. 2010, Grosse et al. 2011). Thawing of ice-rich permafrost can lead to subsidence and deformation of land surfaces that range from localized depressions to deep and extensive thermokarst events. These landscape features, along with thermal erosion, gully formation, and drainage network expansion, are dramatically changing topography, surface hydrology, and vegetation structure on time scales of years to decades.

Coupled climate-carbon models project that the northern high latitudes will serve as a substantial land carbon sink during the twenty-first century because both climate warming and elevated global [CO2] favor increased productivity and carbon uptake in the region (Friedlingstein et al. 2006, Qian et al. 2010, Sitch et al. 2008). However, these models lack many of the key processes governing high-latitude ecosystem behavior, and the magnitude of predicted permafrost thaw and subsequent amount of carbon made available for decomposition (release of CO2) or methanogenesis (release of CH4) varies widely among modeling studies. In contrast, results based on incorporating all of the major factors controlling the high-latitude carbon budget in uncoupled, process-based model simulations generally suggest that the net effect of increasing temperatures over the Arctic is a positive feedback to climate warming (McGuire et al. 2010, Hayes et al. 2011). Models that have projected permafrost carbon losses estimate a substantial, but highly uncertain, magnitude of cumulative emissions to the atmosphere over the next 100 to 200 years (Koven et al. 2011, Schaefer et al. 2011, Schneider von Deimling et al. 2011, Zhuang et al. 2006). Fewer negative feedbacks have been identified, and they may not be large enough to counterbalance the large positive feedbacks (Euskirchen et al. 2010). These feedbacks are generally most pronounced at the regional scale and amplify the rate of regional warming.

Multiple carbon, water, and energy feedbacks that occur in response to permafrost degradation must be resolved if we are to improve model prediction of climate. Permafrost soils store almost as much organic carbon (approximately 1670 Pg; Tarnocai et al. 2009) as is found in the rest of the world’s soils. Because of widespread permafrost thaw (Schuur et al. 2009), much of this soil organic matter may be vulnerable to rapid mineralization. Surprisingly little is known about the vulnerability of permafrost and how the landscape would evolve in the future. Key questions are the extent to which permafrost carbon is stabilized by processes other than cold temperatures and the extent to which the active layer becomes thicker as well as saturated and anaerobic. This is largely a function of how the landscape will evolve over time as a result of strong surface-subsurface interactions and impacts on local to regional hydrology. Anaerobic processes slow the rate of decomposition and favor production of CH4 rather than CO2, thus increasing the climate impact of carbon release because of the higher global warming potential of CH4 (Figure 1). There is evidence of old carbon mineralization upon permafrost thaw (Nowinski et al. 2010, Schuur et al. 2009, Mack et al. 2004), indicating the high vulnerability of the organic matter previously stored in permafrost. Understanding the turnover times of carbon released due to thawing permafrost is critical for modeling the decomposition of organic matter. Moreover, accelerated decomposition may increase nitrogen availability, which promotes vegetation growth and may promote further microbial activity (Nowinski et al. 2008). However, the dynamics and mechanisms of plant response to changes in nitrogen availability are limited to only a few experiments. Furthermore, relatively little is known about the feedbacks that arise due to different forms of nitrogen released upon decomposition of labile vs recalcitrant carbon pools, thus further impeding model assessments (Xu et al. 2011).

While existing representations of land surface processes in Earth System models describe some interrelationships that exist among vegetation, biogeochemistry, and climate, many of the coupled arctic system properties and processes related to permafrost degradation are not currently explicitly represented. The presence of ice wedges, for example, and their influence on surface topography appear to be critical drivers of plot-scale processes but cannot be resolved at even the highest resolutions presently conceived for global-scale climate models. Subsurface geochemical conditions that influence greenhouse gas (GHG) emissions can vary laterally on the order of meters due to interactions between surface water and microtopography induced by thermokarst features or polygonal ground (e.g., Zona et al. 2011). Similarly, the formation, erosion, and drainage of thermokarst lakes (Walter et al. 2007) may provide important feedbacks to climate in high-latitude systems because of their role in the surface energy balance and CO2 and CH4 emissions. Accurately representing these dynamics in Earth System models is difficult, although progress has recently been made to introduce these processes into the Community Land Model (Subin et al. 2011a,b). There is a need for improved high-resolution Arctic terrestrial simulation capabilities that allow explicit representation of properties and processes at the spatial and temporal scales where they occur. Such high-resolution modeling can only be achieved through synthesis of new knowledge and understanding of Arctic system processes emerging from mechanistic studies carried out in the field and in the laboratory.