The FLUXNET2015 dataset provides ecosystem-scale data on CO₂, water, and energy exchange between the biosphere and the atmosphere, and other meteorological and biological measurements, from 212 sites around the globe (over 1500 site-years, up to and including year 2014). These sites, independently managed and operated, voluntarily contributed their data to create global datasets. Data were quality controlled and processed using uniform methods, to improve consistency and intercomparability across sites. The dataset is already being used in a number of applications, including ecophysiology studies, remote sensing studies, and development of ecosystem and Earth system models. FLUXNET2015 includes derived-data products, such as gap-filled time series, ecosystem respiration and photosynthetic uptake estimates, estimation of uncertainties, and metadata about the measurements, presented for the first time in this paper. In addition, 206 of these sites are for the first time distributed under a Creative Commons (CC-BY 4.0) license. This paper details this enhanced dataset and the processing methods, now made available as open-source codes, making the dataset more accessible, transparent, and reproducible.

A Correction to this paper has been published: https://doi.org/10.1038/s41597-021-00851-9.

The current downturn of the arctic cryosphere, such as the strong loss of sea ice, melting of ice sheets and glaciers, and permafrost thaw, affects the marine and terrestrial carbon cycles in numerous interconnected ways. Nonetheless, processes in the ocean and on land have been too often considered in isolation while it has become increasingly clear that the two environments are strongly connected: Sea ice decline is one of the main causes of the rapid warming of the Arctic, and the flow of carbon from rivers into the Arctic Ocean affects marine processes and the air–sea exchange of CO₂. This review, therefore, provides an overview of the current state of knowledge of the arctic terrestrial and marine carbon cycle, connections in between, and how this complex system is affected by climate change and a declining cryosphere. Ultimately, better knowledge of biogeochemical processes combined with improved model representations of ocean–land interactions are essential to accurately predict the development of arctic ecosystems and associated climate feedbacks.

It is important that climate models can accurately simulate the terrestrial carbon cycle in the Arctic due to the large and potentially labile carbon stocks found in permafrost-affected environments, which can lead to a positive climate feedback, along with the possibility of future carbon sinks from northward expansion of vegetation under climate warming. Here we evaluate the simulation of tundra carbon stocks and fluxes in three land surface schemes that each form part of major Earth system models (JSBACH, Germany; JULES, UK; ORCHIDEE, France). We use a site-level approach in which comprehensive, high-frequency datasets allow us to disentangle the importance of different processes. The models have improved physical permafrost processes and there is a reasonable correspondence between the simulated and measured physical variables, including soil temperature, soil moisture and snow. We show that if the models simulate the correct leaf area index (LAI), the standard C3 photosynthesis schemes produce the correct order of magnitude of carbon fluxes. Therefore, simulating the correct LAI is one of the first priorities. LAI depends quite strongly on climatic variables alone, as we see by the fact that the dynamic vegetation model can simulate most of the differences in LAI between sites, based almost entirely on climate inputs. However, we also identify an influence from nutrient limitation as the LAI becomes too large at some of the more nutrient-limited sites. We conclude that including moss as well as vascular plants is of primary importance to the carbon budget, as moss contributes a large fraction to the seasonal CO₂ flux in nutrient-limited conditions. Moss photosynthetic activity can be strongly influenced by the moisture content of moss, and the carbon uptake can be significantly different from vascular plants with a similar LAI. The soil carbon stocks depend strongly on the rate of input of carbon from the vegetation to the soil, and our analysis suggests that an improved simulation of photosynthesis would also lead to an improved simulation of soil carbon stocks. However, the stocks are also influenced by soil carbon burial (e.g. through cryoturbation) and the rate of heterotrophic respiration, which depends on the soil physical state. More detailed below-ground measurements are needed to fully evaluate biological and physical soil processes. Furthermore, even if these processes are well modelled, the soil carbon profiles cannot resemble peat layers as peat accumulation processes are not represented in the models. Thus, we identify three priority areas for model development: (1) dynamic vegetation including (a) climate and (b) nutrient limitation effects; (2) adding moss as a plant functional type; and an (3) improved vertical profile of soil carbon including peat processes.

Over the past decades, vegetation and climate have changed significantly in the Arctic. Deciduous shrub cover is often assumed to expand in tundra landscapes, but more frequent abrupt permafrost thaw resulting in formation of thaw ponds could lead to vegetation shifts towards graminoid-dominated wetland. Which factors drive vegetation changes in the tundra ecosystem are still not sufficiently clear. In this study, the dynamic tundra vegetation model, NUCOM-tundra (NUtrient and COMpetition), was used to evaluate the consequences of climate change scenarios of warming and increasing precipitation for future tundra vegetation change. The model includes three plant functional types (moss, graminoids and shrubs), carbon and nitrogen cycling, water and permafrost dynamics and a simple thaw pond module. Climate scenario simulations were performed for 16 combinations of temperature and precipitation increases in five vegetation types representing a gradient from dry shrub-dominated to moist mixed and wet graminoid-dominated sites. Vegetation composition dynamics in currently mixed vegetation sites were dependent on both temperature and precipitation changes, with warming favouring shrub dominance and increased precipitation favouring graminoid abundance. Climate change simulations based on greenhouse gas emission scenarios in which temperature and precipitation increases were combined showed increases in biomass of both graminoids and shrubs, with graminoids increasing in abundance. The simulations suggest that shrub growth can be limited by very wet soil conditions and low nutrient supply, whereas graminoids have the advantage of being able to grow in a wide range of soil moisture conditions and have access to nutrients in deeper soil layers. Abrupt permafrost thaw initiating thaw pond formation led to complete domination of graminoids. However, due to increased drainage, shrubs could profit from such changes in adjacent areas. Both climate and thaw pond formation simulations suggest that a wetter tundra can be responsible for local shrub decline instead of shrub expansion.

Peat decomposition in managed peatlands is responsible for a decrease of 0.52 GtC yr⁻¹ in global carbon stock and is strongly linked to drainage to improve the agricultural bearing capacity, which increases aeration of the soil. Microbial aerobic decomposition is responsible for the bulk of the net CO₂ emission from the soil and could be reduced by wetting efforts or minimizing drainage. However, the effects of rewetting efforts on microbial respiration rate are largely unknown. In this study, we aimed to obtain more process-based understanding of these rewetting effects on peat decomposition by integrating high-quality field measurements and literature relationships with an advanced hydrological modelling approach where soil moisture and temperature are centralized as the main drivers for peat decomposition. In 2020 and 2021, two dairy farming peatlands, where subsoil irrigation and drainage (SSI) was tested against a control situation, were continuously monitored for CO₂ fluxes, groundwater table, soil moisture and soil temperature. After successfully representing field hydrology and carbon dynamic measurements within our process-based model, we further explored the effects of rewetting under different weather conditions, water management strategies (raising ditchwater levels and SSI) and hydrological seepage settings. To represent peat carbon dynamics we introduced a methodology to estimate potential aerobic microbial respiration rate, based on potential respiration rate curves for soil temperature and water-filled pore space (WFPS). Measurements show that rewetting with SSI resulted in higher summer groundwater levels, soil temperatures and WFPS. SSI reduced the net ecosystem carbon balance (NECB) by 1.58 ± 0.56 kg CO₂ m⁻² yr⁻¹ (83 ± 25 %) and 0.66 ± 0.62 kg CO₂ m⁻² yr⁻¹ (28 ± 15 %) for Assendelft and Vlist respectively in 2020. SSI had a negligible effect in 2021 for both research locations, due to more precipitation, lower temperatures and different SSI management (in Assendelft) as compared to 2020. Simulated rewetting effects were in agreement with measured rewetting effects. Model simulations indicate that raising ditchwater levels always reduces peat respiration rates. Furthermore, we found that the application of SSI (i) reduces yearly peat respiration rates in a dry year and/or with downward hydrological fluxes and (ii) increases peat respiration rates in a wet year and/or when upward groundwater seepage is present. Moreover, combining SSI with high ditchwater levels or pressurizing SSI systems will further reduce peat respiration rates. We show that our process-based approach based on temperature and WFPS soil conditions to determine NECB represents observed variance to a greater extent than empirical relationships that involve average groundwater level observations or simulations. Therefore, we recommend using this kind of approach to estimate the effectiveness of rewetting. When this is not possible, we recommend using mean summer groundwater level instead of mean annual groundwater level as a proxy to estimate NECB. Such relations between mean groundwater levels and NECB are prone to underestimating NECB for SSI parcels.