Arctic Terrestrial Modelling Workshop
14-15 September 2017

ArcTEM workshop

Thursday, September 14, 2017 to Friday, September 15, 2017

St Anne’s College — University of Oxford

Sponsored by the Canadian Network for Regional Climate and Weather Processes (CNRCWP) – Natural Sciences and Engineering Research Council of Canada (NSERC) and Next-Generation Ecosystem Experiments-Arctic (NGEE-Arctic).


This workshop will bring together senior and early career scientists to gain insight into the rapidly changing pan-Arctic land surface and boundary layer. The workshop will review current representation of Arctic ecosystem, carbon, water and energy balance processes in the land model component of Earth System Models, including land-atmosphere interactions, and the next steps to address knowledge gaps. The workshop will also focus on developing a pan-Arctic land model assessment that includes a broader range of models, and engage the data community to provide new validation products for the Arctic and sub-Arctic.

Core focus

  • Representation of Arctic terrestrial ecosystems in models
  • Role of observations: calibration, validation, assimilation
  • Water, nitrogen, carbon, and energy dynamics
  • Land-atmosphere interactions and feedbacks across spatial and temporal scales
  • Pan-Arctic land model assessment
  • Arctic boundary layer processes
  • Extreme/disturbance events
  • Knowledge gaps

Scientific committee

  • Laxmi Sushama (University of Quebec at Montreal)
  • Stan Wullschleger (Oak Ridge National Laboratory)
  • Cathy Wilson (Los Alamos National Laboratory)
  • Eleanor Blyth (Natural Environment Research Council)
  • Benjamin Smith (Lund University)
  • David Lawrence (National Center for Atmospheric Research)
  • Gerhard Krinner (Centre National de la Recherche Scientifique)
  • Patrick Samuelsson (Swedish Meteorological and Hydrological Institute)
  • Paul Miller (Lund University)
  • Joe Melton (Environment and Climate Change Canada)
  • Charles Miller (Jet Propulsion Laboratory, NASA)
Please, register to be able to attend.

The Arctic Terrestrial Modelling Workshop follows on from the Fifth International iLEAPS Science Conference, which is also being held in Oxford. Registration and abstract submission is open, for more information please visit the iLEAPS Conference website (deadline: 8 June 2017)

Please, specify your academic status and scientific interests
Title and abstract
By submitting this form, you accept the Mollom privacy policy.

Day 1 - Thursday September 14

09:00 - 09:30

Introduction and objectives of the workshop

  • Laxmi Sushama (CNRCWP, University of Quebec at Montreal)
  • Stan Wullschleger (NGEE-Arctic, Oak Ridge National Laboratory)

09:30 - 11:30

1. State-of-the-art and knowledge gaps in modelling Arctic land surface processes

Vladimir ROMANOVSKY (University of Alaska Fairbanks)
Modeling Permafrost with High Spatial Resolution

V. Romanovsky, D. Nicolsky, W. Cable, and S. Panda Geophysical Institute, University of Alaska Fairbanks Permafrost has received much attention recently because surface temperatures are rising in most regions where permafrost is found, bringing permafrost to the edge of widespread thawing and degradation. The thawing of permafrost that already occurs at many locations within the permafrost zone can generate dramatic changes in ecosystems, water and carbon cycles, and in infrastructure performance. If the current trends in climate change will continue into the future, there are no doubts that warming of permafrost will eventually trigger widespread permafrost thawing. It is uncertain at what exact locations and areas permafrost will start to thaw first, what will be the rate of this thawing, and what exactly will be the consequences of this thawing for the other components of the Arctic, sub-Arctic and Global Natural Systems. To better understand the possible rates and pathways of permafrost degradation in the future and to predict the local, regional, and global consequences to human society, accurate high spatial resolution permafrost models are needed to be developed. Establishment of these models is possible only by integrating available high resolution environmental data and by assimilation of existing field and remote sensing data and observations into these models. In this presentation we will demonstrate the possibility to generate such permafrost models and describe the necessary steps to develop such a model. Some results of the high spatial resolution permafrost modeling also will be presented.

Guy SCHURGERS (University of Copenhagen)
Representing landscape-scale heterogeneity in large-scale models

The extreme environments found in the Arctic pose particular challenges for a representation in large-scale models. The combination of variability in environmental conditions in the landscape and vegetation growth occurring at the limits of its existence causes a spatial heterogeneity of land surface-atmosphere exchange processes. Moreover, in landscapes that are characterized by large topographic gradients, such as those found along the Greenland coast, the lateral transport of water and nutrients and micrometeorological variations result in hot spots of vegetation growth and plant and soil biogeochemistry in the landscape. In my contribution, I will discuss the challenges of representing these types of landscape-scale heterogeneity in regional-scale or global-scale models. In current models, these are represented only to a limited extent, and explicit or implicit representation of this heterogeneity will help to make these representations more accurate.

Christine DELIRE (Meteo-France)
Modeling snow, permafrost and carbon cycling in high latitudes with the ISBA model

We present an improved representation of snowpack processes and soil properties in the ISBA-SURFEX land surface model and test it over an alpine site and over northern Eurasian regions. The new snow module uses a parameterization of snow albedo and snow compaction/densification adapted from the detailed Crocus snowpack model used in avalanche forecasting by the weather forecasting service. To better represent heat exchange in high latitude soils, we included a dependency on soil organic carbon content for ISBA’s hydraulic and thermal soil properties. The snow module is first evaluated with observed data from the Col de Porte field site (French Alps). The new model version including all of the changes is used over northern Eurasia to evaluate the model’s ability to simulate the snow depth, the soil temperature profile, and the permafrost characteristics. The effect of these modifications on the modelled carbon cycle in high latitudes is assessed.

Terrestrial cryosphere forecasting at ECMWF: status and plans by Linus Magnusson, Peter Bauer, Sarah Keeley and Gianpaolo Balsamo

Polar prediction challenges for accurate forecasting over land include the correct representation of snow phases in both open areas and forest and the treatment of water bodies subject to seasonal freezing. In an operational forecasting system, the result from these simulated processes are confronted with observations on a daily basis. This is a stern test for the model but at the same time serves the model development process. Results will be presented using a series of case studies from operational forecasting experience as well as seasonal verification and also discuss ongoing research diagnostics to highlight the simulated processes and their uncertainties. Perspectives of improved representations of the cryosphere-atmosphere interactions within the APPLICATE project framework at ECMWF will be briefly presented.

Cathy WILSON (Los Alamos National Laboratory)
Reducing Uncertainty in Permafrost Hydrology Models

The CMIP5 Earth System Models were unable to adequately predict the fate of the 16GT of permafrost carbon in a warming climate due to poor representation of Arctic ecosystem processes. The DOE Office of Science Next Generation Ecosystem Experiment, NGEE-Arctic project aims to reduce uncertainty in the Arctic carbon cycle and its impact on the Earth’s climate system by improved representation of the coupled physical, chemical and biological processes that drive how much buried carbon will be converted to CO2 and CH4, how fast this will happen, which form will dominate, and the degree to which increased plant productivity will offset increased soil carbon emissions. These processes fundamentally depend on permafrost thaw rate and its influence on surface and subsurface hydrology through thermal erosion, land subsidence and changes to groundwater flow pathways as soil, bedrock and alluvial pore ice and massive ground ice melts. NGEE scientists are developing data and models to better understand controls on permafrost degradation and improve prediction of the evolution of permafrost and its impact on Arctic hydrology. The LANL Advanced Terrestrial Simulator was built using a state of the art HPC software framework to enable the first fully coupled 3-dimensional surface-subsurface thermal-hydrology and land surface deformation simulations to simulate the evolution of the physical Arctic environment. The model is now being extended to include the biogeochemical, nutrient and vegetation processes and interactions, that when coupled with the physical system drive the carbon, water and energy budgets in the Arctic. The ATS is being used to inform parameterizations of these complex coupled processes for implementation in the DOE ACME land model to better predict the role of warming Arctic landscapes on the global climate system.

Paul BARTLETT (Environment and Climate Change Canada)
The Simulation of Winter Albedo in Boreal Environments Using CLASS

The simulation of winter albedo in boreal and northern environments has been an ongoing challenge for land surface modellers. Many CMIP3 and CMIP5 climate models are characterized by overestimation of albedo in the boreal forest. Recent studies suggest that poor simulation of the snow masking effect of forests, as well as incorrect plant functional type and biases in simulated leaf area index are the primary causes of albedo errors. Different versions of the Canadian Land Surface Scheme (CLASS) have shown positive and negative winter (in the presence of snow) albedo biases in forests, related to the treatment of snow masking by vegetation, canopy snow interaction, bias in canopy cover fraction, and the treatment of subgrid-lakes. We present a brief overview of some key developments in CLASS with respect to the simulation of winter albedo in boreal forests, including historical changes to the representation of snow masking by vegetation, canopy interception and canopy albedo values. Using our latest version (3.6.2), we also examine the sensitivity of simulated albedo to forcing at local and regional scales.

Bakr BADAWY (University of Waterloo)
Uncertainty quantification of snow-related parameters in the Canadian Land Surface Scheme (CLASS)

Snow parameterization in land surface models (LSM) is an important source of uncertainty in climate simulations. However, quantifying this uncertainty is challenging due to the high-dimensional parameter space as well as parameter interaction. In this study, we investigate the sensitivity of varying thirteen snow microphysical parameters in the Canadian Land Surface Scheme (CLASS) using an uncertainty quantification (UQ) approach commonly applied to atmospheric models. In a large ensemble of O(10^3) simulations defined through Latin hypercube sampling (LHS), the parameters are perturbed across their full range of empirical uncertainty determined from available observations and expert elicitation. A statistical model using support vector regression (SVR) is then constructed to efficiently emulate the dynamical CLASS simulations over a much larger O(10^6) set of cases, from which the impact of the parameters on the snow simulation can be fully quantified. The implementation and results of this study will be presented and discussed in detail. Quantifying the importance of snow-related parameters, and their uncertainties, is an important step toward better understanding and quantification of uncertainty within integrated earth system models.

10:30 - 11:00


11:30 - 13:00


  • Leads: Eleanor Blyth, David Lawrence, Diana Verseghy
  • Summary of results from CMIP5 and the Permafrost Carbon Network (David Lawrence)
  • Multi-model ensemble runs of Arctic land surface (Eleanor Blyth)

13:00 - 14:00

Lunch and posters

14:00 - 15:30

2. Land-atmosphere interactions and boundary layer processes

Adam MONAHAN (University of Victoria)
Regimes of the Stably Stratified Boundary Layer: Observations and Models

Observational analyses provide clear evidence that the stably stratified boundary layer can exist in (at least) two distinct regimes. The first, the weakly stable boundary layer (wSBL), is characterized by weak bulk shear, a modest inversion, and sustained turbulence. The second, the very stable boundary layer (vSBL), has strong bulk shear, a strong inversion, and turbulence which is weak and intermittent. This talk will illustrate the structure of these regimes in observations and present a statistical technique (Hidden Markov Model analysis) for distinguishing between them. I will also present results from idealized physical modelling studies designed to understand the dynamics of these regimes. Finally, I will discuss the implications of the existence of these multiple regimes for simulating the stable boundary layer in regional and global models.

Eric BAZILE (Meteo-France)
Overview of the GABLS4 (GEWEX Atmospheric Boundary Layer Study) intercomparison
*Talk will be given remotely during summary of Day 1 - Friday 09:00
Javier FOCHESATTO (University of Alaska Fairbanks)
The role of the Arctic Atmospheric Boundary Layer in the landscape representation of ecosystems fluxes

Surface-atmosphere interactions are central to understanding current and future trends in weather and climate. Large-scale surface fluxes are the quantities often required for model input and/or validation. However, this research is challenging because it involves the analysis of turbulence data that is best understood only under specific surface and atmospheric boundary layer (ABL) conditions, and is often limited to specific levels within the ABL. Such ideal conditions are not always representative of the local surface and flow properties, nor of the spatial and temporal variation of fluxes owing to those properties. Even within what may be classified as the same surface type, local flux values can vary considerably. Quantifying the relationship between local and large-scale fluxes and their connection to surface properties and ABL-flow regime has been the motivation for numerous field campaigns and remains an unsolved problem. In this seminar, I will describe observational platforms and newly developed methodologies that apply to Arctic-ABL research and surface turbulent fluxes. I will discuss multiscale turbulent fluxes experiments carried out in the most vulnerable ecosystems on earth. And, I will introduce new empirical approximations of multiscale turbulent fluxes. To conclude, I will share my research strategy for the next generation of experiments integrating observations and modeling approaches.

Emanuel DUTRA (University of Lisbon)
Land-atmosphere (de)coupling in the presence of snow in the ECMWF and EC-EARTH models

Snow cover presence and its temporal and spatial variability is a key component of the Earth System. In addition to the hydrological implications, such as water storage, it also plays an important role in controlling energy and water exchanges between the land-surface and atmosphere. One of the key characteristics of snow is the high thermal insulation and reduced surface roughness, when compared with snow-free conditions. These particular characteristics, together with dominant stable boundary conditions, modulate land-atmosphere coupling and are challenging to represent in numerical weather prediction (NWP) and climate models. This work will present an overview of the current status of land-atmosphere coupling in the presence of snow in the NWP ECMWF and climate EC-EARTH models. Particular focus will be given to the role of inter-annual snow variability on the Northern Hemisphere winter. On shorter time-scales, the thermal characteristics and roughness of snow will be evaluated in terms of land-atmosphere coupling. In addition to coupled land-atmosphere simulations, surface-only (offline) simulations will be also carried out to evaluate the benefits and drawbacks of this computationally cheaper setup that allows extensive testing when compared with fully coupled simulations.

Gulilat Tefera DIRO (University of Quebec at Montreal)
The role of snow trend and variability on the future snow-atmosphere coupling over North America
Wenxin ZHANG (University of Copenhagen)
Self-amplifying feedbacks accelerate greening and warming of the Arctic

Observational studies provide ample evidence of greening, increased productivity and an increased representation of woody elements in Arctic vegetation in response to recent warming. Biogeophysical feedbacks mediated by a changed land surface energy balance resulting from the effects of vegetation dynamics on albedo and evapotranspiration (ET) are associated with observed changes in climate and vegetation, but their spatio-temporal variations in response to future climate change and the consequences for Arctic vegetation and ecology have not been robustly quantified. We applied a regional Earth system model (RESM) interactively coupling atmospheric dynamics to vegetation productivity, structure and distribution in an ensemble of climate projections bracketing the potential 21st century radiative forcing across the Arctic. We demonstrate that vegetation feedbacks dominated by albedo-mediated warming in the late snow season and evapotranspiration-mediated cooling at the height of summer may strongly modulate the future evolution of the Arctic climate, compared to expectation based on the responses of physical climate processes alone. In the strongest forcing (RCP8.5) scenario, feedbacks lead to an earlier, more thermally-equitable growing-season characterized by enhanced water availability, providing conditions amenable to a continuation and enhancement of vegetation trends already observed across the Arctic.

Bernardo TEUFEL (University of Quebec at Montreal)
The projected evolution of vegetation-climate feedbacks over the pan-Arctic
Frode STORDAL (University of Oslo)
Effects of shrub cover increase on the near surface atmosphere in northern Fennoscandia

Effects of shrub cover increase on the near surface atmosphere in northern Fennoscandia F. STORDAL1, A. BRYN2, J.H. RYDSAA1, L.M. TALLAKSEN1 1Department of Geosciences, University of Oslo, P.O. Box 1047 Blindern, 0316 Oslo, Norway 2Natural History Museum, University of Oslo, P.O Box 1172 Blindern, 0318 Oslo, Norway Shrub expansion in high latitudes can lead to positive feedbacks to the regional climate. We have evaluated the sensitivity to a potential expansion in shrub and tree cover in the northern Fennoscandia region. Two perturbation experiments are performed in which we prescribe a gradual increase of vegetation height in the alpine shrub and tree cover according to empirically established climatic zones within the study region. The first experiment is based on present day climate, and the second is based on a future 1 K increase in temperature. To evaluate the sensitivity of the atmospheric response to inter-annual variations, simulations were conducted for two different years, one with warmer and one with colder spring and summer conditions. We have applied the Weather Research and Forecasting model (WRF) with the Noah-UA land surface module in evaluating biophysical effects of increased shrub cover on the near surface atmosphere on a fine resolution (5.4 km x 5.4 km). We find that shrub cover increase leads to a general increase in near surface temperatures with the peak influence occurring during the snow melting season. It has the largest effect in spring, by advancing the onset of the melting season, and more moderate effect on summer temperature. We find that the net SW absorbed by the surface is sensitive to the shrub and tree heights, which act to strengthen the albedo decrease. Counteracting effects include increased snow cover and enhanced evapotranspiration causing increased cloud cover and precipitation. The strength of the feedback effects resulting from increased shrub cover is more sensitive to snow cover variations than summer temperatures. Taller vegetation has a stronger influence on both spring and summer temperatures. The positive feedback to high latitudes warming induced by increased shrub and tree cover is a robust feature across inter-annual differences in meteorological conditions, and will likely play an important role in the future.

15:30 - 16:00


16:00 - 17:30


  • Leads: Laxmi Sushama, Patrick Samuelsson, Paul Miller
  • Initial references (to be provided in advance)

Day 2 - Friday September 15

09:00 - 09:30

Résumé of day 1:

  • Scientific committee

09:30 - 11:30

3. Role of observations: calibration, validation, assimilation

Jürgen HELMERT (German Meteorological Office)
Contributions from COST-HarmoSnow to validation and data assimilation of snow in models of NWP, hydrology, and climate

The ESSEM COST Action ES1404 Harmosnow ( is focusing on improving the connection between snow measurements and models, between snow observers, researchers and forecasters, for the benefit of various stakeholders and the entire society. The aim is to enhance the capability of the research community and operational services to provide and exploit quality-assured and comparable regional and global observation-based data on the variability of the state and extent of snow. Advancing snow data assimilation in European Numerical Weather Prediction (NWP) and hydrological models with showing its benefit for relevant applications and establishing a validation strategy for climate, NWP and hydrological models against snow observations are objectives of the COST action. Within the HarmoSnow working group on snow data assimilation and validation methods for NWP, hydrology, and climate models a survey of the various snow observations used in NWP, hydrology and climate studies for different purposes including validation and data assimilation was conducted. There are also studies on methods for combining satellite observations with conventional in-situ snow measurements and physical parametrisations of the snowpack in numerical models. This includes also the support of active/passive microwave model developments. An assessment of the impact of a more extended usage of conventional snow observations from high-resolution national networks into NWP was performed and recommendations to increase the availability of snow observation data were formulated. A web-based portal of snow observations, measurements and instruments can be used for monitoring purposes. The information about observational errors are relevant for snow data assimilation and establish links between the modelling and measurement communities.

Remote Sensing of Fractional Snow Cover and Its Applications for Terrestrial Modeling

The information on snow cover is critically important to describe the interactions between the atmosphere and underlying surface influencing the accuracy of land surface models and assimilation schemes determining results of hydrological simulations. However, the problem of reliable snow cover parameterization in models is far from being solved. Snow fraction is one of the most important components of land surface and hydrological models. The errors in estimates of snow fraction lead to very significant inaccuracies in albedo, soil moisture, energy exchange, and atmospheric conditions. Because of an extremely pronounced impact of fractional snow cover, it is necessary to improve the quality of its retrieval from remote sensing. The sub-grid variability of snow within numerical models describing processes of mass and energy exchange calls for knowing the fractional snow cover and its distribution as accurately as possible. In summary, many existing approaches to estimating fractional snow cover are based on a hypothesis that characteristics of individual pixels are linear combinations of spectral signatures of pure snow and non-snow. However the significant variability in spectral signatures of the predetermined endmembers often is not taken into account and observed changes in pixel reflectances are completely ascribed to variable fractions of the endmembers. As a result, the performance of individual algorithms not only depends on the features of the method used, but is influenced to a very great degree by the choice of endmembers for the scenes included in consideration by the authors of the methods. Improved fractional snow cover retrieval can be achieved by an algorithm that takes into account the variability in space and time of snow and non-snow spectral signatures. There are two quite different types of the variability in snow and non-snow reflectances. The endmember reflectances are characterized by a very significant local changes within a scene as well as by marked changes in predominant values from one scene to another. The local variability in snow and non-snow endmembers are approximated by the Normalized Difference Snow Index (NDSI) with a high accuracy. The magnitudes of snow and non-snow NDSI are scene-specific and calculated on the fly. Such an approach employing scene-specific NDSI characterizing local reflective properties of snow and non-snow endmembers is considered as the best way to improve snow fraction retrieval from remote sensing observations on the basis of assuming a linear relationship between NDSI and snow fraction. The quality of snow fraction is evaluated through a quantitative comparison of the snow retrieval results with high resolution Landsat data in the vicinity of a snow line separating snow covered areas from snow free regions. The opportunity of utilizing Landsat data not only as a source of ground truth, but also for investigation of reflectances, applicable to MODIS, VIIRS, Sentinel-3, and GCOM-C presents obvious benefits of choosing Landsat as reference data for validation. The Landsat reference data are used to estimate the performance of the developed fractional snow cover algorithm and to compare with the quality of alternative algorithms. The validation results demonstrate that on the whole the performance of the algorithms using the Normalized Difference Snow Index (NDSI) has advantages in comparison with the algorithms utilizing tie-point interpolation for individual reflectances. The scene-specific realization of the NDSI algorithm is close to its optimal version and therefore could be preferable for snow fraction retrieval. The optimal approach to improve moderate resolution remote sensing of fractional snow cover combines allowing for the local variability of snow and non-snow reflective properties with a scene-specific consideration to create unbiased and consistent information on snow cover distribution required for global, regional and local studies including terrestrial and hydrological applications. In particular, satellite derived fractional snow cover can be effectively used to estimate Snow Water Equivalent (SWE). The use of the Snow Depletion Curve (SDC) illustrating related temporal changes in snow coverage and SWE obtained from remote or in situ observations is a powerful approach to describe local conditions, employed by many researchers. It has been demonstrated that taking into account specific features of SDC varying with elevation and from region to region enhances the quality of SWE simulations. Several different methodologies are created to translate observations on snow fraction into SWE. A promising approach is based on a data assimilation scheme that simulates daily changes of SWE. The essence of the approach is the analysis of discrepancy between calculations of snow fraction and its retrieval from satellite observations to identify the parameters responsible for discrepancies and to recommends their tuning for better results. The approach provides the best quality of results when several consecutive satellite observations in the region under consideration are available for the same season. Snowmelt runoff models could establish the relationship between snow-covered area and water runoff caused by snow melting. The relationships between snow fraction and the volume of melting snow defined by the surface heat balance could be determined for cells of a model domain or uniformed parts of a catchment.

Wolfgang BUERMANN (University of Leeds)
Phenological controls on plant productivity across northern ecosystems

Observational records show strong phenological shifts in response to climate change across northern ecosystems. Yet our understanding of how these phenological shifts influence plant productivity over the course of a growing season is surprisingly limited. Using 30+ years of optical and microwave satellite data, I will determine the phenological model that is best supported by the data for a given locality. These observational-based distributions of phenological controls on plant productivity can then be compared to model simulations.

Annett BARTSCH (Central Institute for Meteorology and Geodynamics)
Remotely sensed data for circumpolar permafrost assessment - the ESA DUE GlobPermafrost project

Permafrost cannot be directly detected from space, but many surface features of permafrost terrains and typical periglacial landforms are observable with a variety of EO sensors ranging from very high to medium resolution at various wavelengths. In addition, landscape dynamics associated with permafrost changes and geophysical variables relevant for characterizing the state of permafrost, such as land surface temperature or freeze-thaw state can be observed with space-based Earth Observation. Suitable regions to examine environmental gradients across the Arctic have been defined in a community white paper (Bartsch et al. 2014). These transects have been revised and adjusted within the ESA DUE GlobPermafrost project. The ESA DUE GlobPermafrost project develops, validates and implements Earth Observation (EO) products to support research communities and international organisations in their work on better understanding permafrost characteristics and dynamics. Prototype product cases will cover different aspects of permafrost by integrating in situ measurements of subsurface properties and surface properties, Earth Observation, and modelling to provide a better understanding of permafrost today. The project will extend local process and permafrost monitoring to broader spatial domains, support permafrost distribution modelling, and help to implement permafrost landscape and feature mapping in a GIS framework. It will also complement active layer and thermal observing networks. Both lowland (latitudinal) and mountain (altitudinal) permafrost issues are addressed. The status of the Permafrost Information System and first results will be presented. This includes prototypes of GlobPermafrost datasets, and the permafrost information system through which they can be accessed.

Pavel KOTOV (Lomonosov Moscow State University)
Forecasting equations for thaw settlement calculation

Currently, thaw settlement estimation is a very important task, especially due to global warming in permafrost area. Two main approaches for settlement estimation have been specified: calculating (using only physical characteristics of soils) and experimental (field or laboratory frozen soil testing). Prediction of thawing soil settlements is focused on experimental determination of deformation characteristics (thawing and compressibility coefficients). But tests are laborious and time-consuming. At the same time, using equations, you can calculate thaw settlement inexpensively. The aim of this work was to conduct calculations thaw settlement for various regions of Western Siberia using different equations and compare with the value, obtained using deformation characteristics. Forecasting equation was chosen for each region. Calculations were performed using 10 different equations. These equations are obtained by the authors using different approaches, but all are based on the generalization of experimental data. These equations do not take into account the pressure thawing rate, cryogenic structure, but allow to calculate a preliminary assessment of thaw settlement. This is a particularly important characteristic on stage of project preparation and choice of key areas for drilling and sampling. Analysis of the equations showed that the thaw settlement depends on: density (frozen soil density, dry soil density, soil particles density), water content indicators (water content, ice content, unfrozen water content). Influence of dispersion was taken into account plastic limit or plasticity index.

Friedrich OBLEITNER (Innsbruck University)
Svalbard glaciers: observing and modelling decadal changes of mass balances and related surface exchange processes

Svalbard (Spitzbergen) is one of the most glaciated regions in the Arctic and has experienced a significant warming since decades now. Correspondingly, its glaciers and ice sheets experienced substantial mass losses and many of them retreated, especially in the south of the archipelago. Such responses to regional climate change were also projected into the future but major uncertainties remain concerning snow/ice-atmosphere exchange processes and feedbacks with changing thermal regimes and dynamics of the glaciers and their environments. Kongsvegen glacier is one of the few Svalbard benchmark glacier where long-term records of mass balance and meteorological conditions are available. We use these data to demonstrate the observed developments and to investigate the inherent processes. The latter is mostly based on the application of glacier mass and enegy balance models driven by observational data or downscaled output from numerical weather prediction models. Investigation of snow processes and related surface exchange processes play a major role in this context.

Wouter HANTSON (University of Maine)
Inter-Comparison and Functional Benchmarking of Ecosystem Models for NASA’s ABoVE Field Campaign.

The Arctic Boreal region (ABR) is the source of among the largest uncertainties to global climate projections (1). High uncertainties for carbon fluxes, defined as the variance among multiple models (2), means that Arctic-Boreal Ecosystem Models can exhibit nearly every possible combination of net carbon flux source / sink pattern as shown by Fisher et al. (2014) for Alaska (3). A key challenge is that there are few data available to benchmark models and guide improvements, and thus to reduce uncertainty. Within our ABoVE project, we are evaluating, identifying and quantifying the sensitivity – vulnerability and resilience – of ecosystem dynamics in the ABR as captured by terrestrial biosphere models (TBMs). Results will be used to improve TBM performance in representing, simulating, and scaling the key indicators of ABR ecosystem dynamics and their associated uncertainties. By leveraging existing TBM simulation results from the MsTMIP-project (4), existing model outputs can be compared to and evaluated against field and remote sensing data among the ABoVE ecosystem dynamics indicators i.e., (vegetation, carbon, permafrost, water, wildlife/habitat). These data include remote sensing observations and products for: phenology (MODIS NDVI, EVI, fAPAR) GPP and NPP (MODIS), fire (MODIS), albedo (MODIS), biomass (ICESat/GLAS), canopy height (ICESat/GLAS), evapotranspiration (MODIS), soil moisture (SMOS), total water storage and derived groundwater (GRACE), and land surface temperature (MODIS). This suite of data provides a robust constraint on interacting biogeochemical components of dynamic ABR ecosystems. Initial model inter- comparison and evaluation against Arctic/Boreal specific benchmarks, derived from field and remote sensing data, show model performance, sensitivities and uncertainties in key ecological and carbon cycle indices across the ABR. 1 Koven et al.; IPCC, 2014; Schaefer et al., 2014; Sneyder and Liess, 2014. 2 IPCC, 2014 3 Fisher et al., 2014 4 Huntzinger et al., 2013 Literature cited Fisher, J. B., M. Sikka, W. C. Oechel, D. N. Huntzinger, J. R. Melton, C. D. Koven, A. Ahlström, M. A. Arain, I. Baker, J. M. Chen, P. Ciais, C. Davidson, M. Dietze, B. El-Masri, D. Hayes, C. Huntingford, A. K. Jain, P. E. Levy, M. R. Lomas, B. Poulter, D. Price, A. K. Sahoo, K. Schaefer, H. Tian, E. Tomelleri, H. Verbeeck, N. Viovy, R. Wania, N. Zeng, and C. E. Miller (2014), Carbon cycle uncertainty in the Alaskan Arctic, Biogeosciences, 11(15), 4271-4288. Huntzinger, D., C. Schwalm, A. Michalak, K. Schaefer, A. King, Y. Wei, A. Jacobson, S. Liu, R. Cook, W. Post, G. Berthier, D. Hayes, M. Huang, A. Ito, H. Lei, C. Lu, J. Mao, C. Peng, S. Peng, B. Poulter, D. Riccuito, X. Shi, H. Tian, W. Wang, N. Zeng, F. Zhao, and Q. Zhu (2013), The North American Carbon Program Multi-scale synthesis and Terrestrial Model Intercomparison Project–Part 1: Overview and experimental design, Geoscientific Model Development, 6, 2121-2133. IPCC (2014), Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, 1132 pp., Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Koven, C. D., B. Ringeval, P. Friedlingstein, P. Ciais, P. Cadule, D. Khvorostyanov, G. Krinner, and C. Tarnocai (2011), Permafrost carbon-climate feedbacks accelerate global warming, Proceedings of the National Academy of Sciences, 108(36), 14769-14774. Schaefer, K., H. Lantuit, V. E. Romanovsky, E. A. Schuur, and R. Witt (2014), The impact of the permafrost carbon feedback on global climate, Environmental Research Letters, 9(8), 085003. Snyder, P. K., and S. Liess (2014), The simulated atmospheric response to expansion of the Arctic boreal forest biome, Climate dynamics, 42(1-2), 487-503.

10:30 - 11:00


11:30 - 13:00


  • Leads: Stan Wullschleger, Cathy Wilson, Charles Miller

13:00 - 14:00

Lunch and posters

14:00 - 15:30

4. Ecosystem processes and carbon/nitrogen cycles

Atul JAIN (University of Illinois)
Dominant role of soil physical states over nitrogen cycle on permafrost soil carbon

Permafrost carbon feedback in earth system models is dependent on the simulation soil organic carbon (SOC) currently stored in the northern high-latitude (NHL) permafrost, which remains highly uncertain. While this is attributed to factors such as lack of nitrogen dynamics, cold region soil/snow physics, and non-uniformity in land-cover and climate across models, their relative importance is unknown. Here, we have used a land surface model, ISAM, to show that simple combination of these factors can produce highly divergent NHL permafrost SOC. Our study suggests that the SOC is strongly dependent on the treatment of cold-region soil/snow physics, the lack of which from a full model configuration reduced it by up to 50% of SOC. This magnitude is greater than that from nitrogen dynamics – excluding which the SOC increased by approximately 31%, and from climate and land-cover driven uncertainties (~±10%). SOC accumulation with the soil/snow processes was primarily driven by cooler and drier soils with minimal impact from litter, while nitrogen impact was dominantly through litter. Our modeling analysis indicates that in future, potential gains in permafrost SOC storage from warming-induced nutrient mobilization should be much weaker than the offsetting decreases from higher decomposition. While continued improvements in biogeochemical schemes are necessary, we argue that detailed treatment of cold-region biogeophysics is more critical as the underlying driver of SOC for the NHL environment.

Michael RAWLINS (University of Massachusetts - Amherst)
Improvements in modeling the export of water and carbon within Arctic river-estuary systems

Improvements in modeling the export of water and carbon within Arctic river-estuary systems Michael A. Rawlins1, James W. McClelland2, Robert G. M. Spencer3 1Climate System Research Center, University of Massachusetts-Amherst, Amherst, Massachusetts, USA 2Marine Science Institute, University of Texas at Austin, Port Aransas, Texas, USA 3Department of Earth, Ocean and Atmospheric Science, Florida State University, Tallahassee, Florida, USA, Quantitative estimates of organic matter, sediment and other constituent exports from land to sea are crucial for understanding influences on coastal food webs and biogeochemical processes that drive carbon-cycle feedbacks in the Arctic. Precipitation input and permafrost state are key controls on the spatial distribution and timing of water movement and the associated export of carbon, nutrients, and materials from the Arctic landmass into coastal regions. Continued warming and attendant hydrological cycle intensification are among the many environmental changes that have the potential to result in alterations in the timing and amount of these exports. Projections of future changes in water and materials export is dependent on improved understanding of the dominant controls on constituent mobilization and transit across watersheds, and the ability to anticipate alterations which may produce non-linear responses over time. For example, increased soil thaw in areas with deep accumulations of organic material may produce increases over time in dissolved organic carbon (DOC) leaching and riverine export. In contrast, regions where thaw is expected to occur through predominately mineral soils may experience decreases in DOC export. It is believed that warming is altering the relative amount of surface vs groundwater inputs and increasing flow path lengths. Warming and landscape disturbances may also be leading to changes in processes which control the production and leaching of DOC and other constituents from soils and into stream and river networks. The relative degree of processing of carbon in streams and rivers by photo- and microbial degradation is not well known at this time. Here we describe how in-situ measurements and hydrological modeling are being used to explore average conditions and seasonal to interannual variability in the export of DOC from Arctic watersheds. We investigate the dominant influences on organic matter inputs that can be leveraged to parametrize hydrological and biogeochemical models. Process rates for biogeochemical cycling across the river-estuary continuum are also explored. In this presentation we will discuss measured data, key geophysical processes, and numerical approaches that enable improved modeling of contemporary and future water and carbon exports and associated impacts to the biology and biogeochemistry of Arctic coastal waters.

Eleanor BURKE (Met Office)
Permafrost Nitrogen Interactions and their Response to Climate Change

Nitrogen is known to be one of the main limiting nutrients for plant growth in the arctic tundra. As the climate changes and the permafrost thaws, the nitrogen available to plants may increase leading to enhanced plant growth. In order to quantify this process, a vertically discretised soil nitrogen scheme was included within the JULES land surface model. This simulated nitrogen cycle is sensitive to the presence of permafrost. This version of JULES compared well with observations over a range of Arctic sites. Simulations under future scenarios of climate change show the simulated nitrogen limitation decreasing as the global mean temperature increases.

Mousong WU (Lund University)
Constraining carbon fluxes in cold regions using a carbon cycle data assimilation system—A model-data fusion approach

Carbon cycle in cold regions is important for understanding connections between climate change and Arctic ecosystem response. In terrestrial biosphere models, one key issue is to better predict carbon fluxes under future scenarios. In this study, we conducted simulation work with a carbon cycle data assimilation system (CCDAS) for four sites in cold regions, Sodankyla in Finland, Soroe in Denmark, Groundhog in Canada, and Siziwang in northern China. They represent different climate zones and plant types. We used two sources of data to assimilate the terrestrial biosphere model—BETHY. The first dataset is SMOS (Soil Moisture and Ocean Salinity) soil moisture data, and the second one is Fapar (Fraction of absorbed photosynthetically active radiation) dataset. We used them to constrain parameters related to soil water, plant phenology as well as soil and plant carbon cycle processes. Results showed that the combinations of SMOS soil moisture data with Fapar dataset could reduce parameter uncertainties more than using them solely in data assimilation system. Fapar data could constrain carbon cycle well when there was limited soil moisture data for some sites. However, soil moisture data was important in constraining water uptake by roots and related evapotranspiration processes. This study has shown the possibility of using multi-source dataset in constraining a terrestrial biosphere model in cold regions. More study will be done in next step to explore the importance of each dataset in carbon cycle simulation as well as to investigate model performance in regional scale modeling.

15:30 - 16:00


16:00 - 17:30


  • Leads: David Lawrence, Gerhard Krinner
  • Towards understanding permafrost carbon, nitrogen, and water interactions under a changing climate (David Lawrence)

17:30 - 18:00


  • Scientific committee


Yuanchao FAN (Uni Research Climate & Bjerknes Centre for Climate Research)
Modeling permafrost landscapes in transformation: from above-ground vegetation dynamics to underground biophysical changes

Climate change has been altering the behaviors of terrestrial ecosystems, with especially augmented influences in the Arctic region including permafrost and the covering vegetation types. Permafrost soils in the high latitudes contain large amount of organic matter and carbon, and the Arctic plants such as bryophytes and lichens provide insulating effects on the soils. The above- and belowground dynamics are thus closely connected in determining the fate of permafrost carbon pools. Global warming is accelerating the rates of permafrost thawing and degradation, which could potentially release large amount of GHGs such as carbon dioxide and methane to the atmosphere and also modify the landscapes through land subsidence and formation of thermokarst wetlands. The underlying biophysical and biogeochemical processes of thawing permafrost are not adequately represented in Earth system models (ESMs). One key process of landscape change (i.e. land subsidence) is induced by melting of excess ground ice which is highly nonlinear and occurs on spatial scales below ESM grids. Another important feedback / interactive process between Arctic vegetation dynamics and underlying permafrost transformation is also rarely represented in ESMs. These knowledge gaps have drawn attentions of climate and ecosystem scientists. As an early career researcher based in the Bjerknes Centre for Climate Research, I have been involved in a project addressing the scale mismatch problem of permafrost landscape changes and ESM grids (using NorESM) as well as studies (field and modeling) on the biological response of Arctic plants to global warming and their interaction with the underlying permafrost through energy and material cycling. The Arctic Terrestrial Modelling Workshop and its core focuses could provide an important opportunity to develop such skills essential for undertaking the above research tasks and potentially contribute to the pan-Arctic land model development and assessment.

Markus TODT (Northumbria University)
Simulation of longwave enhancement beneath coniferous forests

CMIP5 models have been shown to underestimate both trend and variability in northern hemisphere spring snow cover extent, a substantial fraction of which is covered by boreal forests. Forest coverage shades the ground and enhances longwave radiation thereby impacting the radiation budget of the ground which is dominating the snow energy balance in forests. Longwave enhancement is a potential mechanism that contributes to uncertainty in snowmelt modelling. Here we use radiation measurements from an alpine forest to assess the simulation of sub-canopy longwave radiation by CLM4.5, the land component of the NCAR Community Earth System Model. CLM4.5 overestimates the diurnal cycle of sub-canopy longwave radiation and consequently longwave enhancement. Overestimation results from clear sky conditions, due to high absorption of shortwave radiation during daytime and radiative cooling during nighttime. Using recent improvements to the canopy parameterisations of SNOWPACK as a guideline, CLM4.5 simulations of sub-canopy longwave radiation improve through the implementation of a heat mass parameterisation, i.e. including the thermal inertia effect due to biomass. However, this improvement does not substantially reduce the amplitude of the diurnal cycle, a result also found during the development of SNOWPACK.

Heather KROPP (Colgate University)
The Influence of Vegetation on the Decoupling between Air and Soil Temperatures across the Arctic

Vegetation plays a key role in the decoupling between the air and active layer temperature via impacts on surface temperature and soil thermal properties. However, the quantification of vegetative impacts on active layer temperatures is largely limited to local or regional scales. We compiled data on a Pan-Arctic scale that pairs measurements of air and soil temperature with vegetation and ecosystem data. We quantify the impacts of vegetation on the decoupling of air and soil temperatures across the Arctic through semi-mechanistic modeling approaches of soil temperature and empirical models of air-soil temperature decoupling using n factors (degree days soil/degree days air). Our results indicate that empirical models that include vegetation data can explain variation in air-soil temperature decoupling by as much as 75%. We also find that the impact of vegetation functional types on the air-soil temperature decoupling can vary across regions of the Arctic. We find that large-scale quantitative measures of the impacts of vegetation on active layer temperature are critical to understanding how changes in vegetation under climate change can further affect permafrost stability and soil temperature. Furthermore, pan-Arctic scale observations and the relationships established between vegetation and soil temperature have the potential to contribute to benchmarking large scale terrestrial and earth system models.

Hui TANG (University of Oslo)
Improve dynamic vegetation model of community land model in simulating Arctic vegetation and its interaction with climate

Improve dynamic vegetation model of community land model (CLM4.5BGCDV) in simulating Arctic vegetation and its interaction with climate H. TANG(1), F. STORDAL(1), T.K. BERNTSEN(1), A. BRYN(2) (1)Department of Geosciences, University of Oslo, P.O. Box 1047 Blindern, 0316 Oslo, Norway (2)Natural History Museum, University of Oslo, P.O Box 1172 Blindern, 0318 Oslo, Norway Arctic vegetation is undergoing significant changes in response to global warming, which can exert great impact on the Arctic climate and contribute to a stronger warming in the Arctic (i.e., Arctic Amplification). Correctly modelling vegetation dynamics and its climate feedbacks in the Arctic region is therefore essential for a better projection of future Arctic climate changes. Dynamic global vegetation model (BGCDV) has been developed in community land model (CLM4.5), but the description of Arctic vegetation is quite coarse with only 4 plant functional types (PFTs) representing the Arctic vegetation. Its ability in simulating Arctic vegetation distribution and climate feedbacks has not been widely tested. In this study, we attempt to evaluate and improve the performance of CLM4.5BGCDV in simulating vegetation dynamics in the Arctic region. It is found that the default setting of CLM4.5BGCDV can reasonably reproduce total plant cover fraction under present-day atmospheric forcing, but seriously underestimate plant cover fraction of boreal deciduous shrub. As a result, the fully coupled atmosphere-CLM4.5BGCDV run exhibits a strong cold bias in northern Eurasia and Canada due to a positive feedback between Arctic shub cover and temperature. To alleviate the biases of CLM4.5BGCDV in simulating boreal deciduous shrub cover and the strong cold biases in the coupled atmosphere-CLM4.5BGCDV run, several vegetation parameters have been modified for the Arctic PFTs, including photosynthetic capacity, root distribution, light competition and nitrogen limitation to allow a better growth of boreal deciduous shrub in the Arctic. In addition, a new PFT – boreal needleleave deciduous trees, was added to CLM4.5BGCDV to better represent the vegetation in northern Eurasia. The effect of these modifications on simulating Arctic vegetation and its climate feedbacks will be discussed.

Chad THACKERAY (University of Waterloo)
The impact of simulated albedo biases on climate

Snow cover plays a vital role in the Arctic climate system, primarily through impacts on the water and energy budgets. Changes in snow extent are the main driver of seasonal variability in Northern Hemisphere land surface albedo. Prior research has demonstrated that several global climate models exhibit deficiencies in the magnitude and/or seasonality of simulated snow-covered surface albedo, despite reasonable representations of snow cover. This study aims to determine the effects of albedo biases from two models (Community Land Model, Model for Interdisciplinary Research on Climate) with differing issues (related to timing and magnitude, respectively) on climate simulations. The experimental design overrides the model’s (biased) internal calculations of snow albedo, and replaces it with prescribed observational forcing derived from satellite data. By deliberately correcting the model biases, we can expect to directly relate the change in model response of temperature and related variables (relative to a control case) to the perturbation applied. Preliminary results generated from offline simulations show that removing biases related to the magnitude of snow-covered surface albedo is more important than that of biases related to the seasonality. However, both factors are shown to influence important variables such as surface temperature and snow cover. Additional simulations are ongoing to further assess how prevalent these issues are in current land surface models.

Peter HORVATH (University of Oslo)
Distribution model of selected vegetation types in the boreal arctic zone.

I am a PhD student at University of Oslo working in collaboration with the Department of Geosciences (MetOs) and the Natural History Museum (NHM) on a strategic research project - LATICE (Land-ATmospheric Interactions in the Cold Environments). Here I will present the first of my PhD projects: Representation of boreal and arctic ecosystems in climate models is still rather uncertain and there is a clear need for more precise vegetation input data. Species distribution modelling is a common approach for predicting species occurrence in unmapped areas based on presence data and environmental data. Here, we use distribution modelling to predict the probability of occurrence of different Vegetation Types in Norway. Our vegetation data, containing both presence and absence occurrences, is based on The Norwegian area frame survey of land cover and outfield land resources (AR18x18) dataset, a 10-year mapping effort of vegetation carried out by the Norwegian Institute for Bioeconomy Research. In addition, we are supplying our models with high-resolution environmental data (100 meter) covering the whole land area of the country. Using statistical method of GLM, an automatic forward model selection based on MIAmaxent R-package, and independent evaluation datasets, we predict selected Vegetation Types for Norway. Our further aims are to compare outputs from our Distribution model with a different vegetation-modelling approach, Dynamic vegetation modelling (DGVM), and then suggest possible ways of improving the parameterizations of dynamic global vegetation model in the boreal zone. On these topics, I am working together with colleagues from University of Oslo - Frode Stordal and Hui Tang (Dept. of Geosciences, MetOs) as well as Anders Bryn and Rune Halvorsen (NHM).

Robin WOJCIK (German Research Centre for Geosciences)
Carbon budget of Icelandic glacier forefields

Recently deglaciated forefield of retreating glacier are highly biogeochemically active and rapidly changing environments as they are affected by weathering and ecological colonization. The potential feedback that glacier forefield may provide on global biogeochemical cycle has attracted a considerable interest over recent years, particularly since glaciers are expected to retreat at increasing rates in the future. The increasing supply of fresh material exposed by the retreat of glacier may enhance the chemical weathering rates of rocks by carbon dioxide dissolved in rainwater. The resulting carbon dioxide sink from the atmosphere may provide negative feedbacks on global warming. Besides, carbon fluxes related to weathering, it is expected that the biomass and soil development on High Arctic glacier forefield will result in the uptake carbon from the atmosphere. Future global warming and subsequent degradation of permafrost will most likely result in the remobilization of small amounts of soil organic carbon. Little work has been done on quantifying the balance of diverse carbon fluxes of glacier forefield and further investigate their potential feedback on global biogeochemical cycles. The present work aims to determine the carbon budget of Icelandic glacier forefield over different time scales by considering the carbon fluxes of rock weathering and developing biomass. This project will be based on soil samples collected in various retreating glacier forefield of Iceland. Weathering will be characterized based on the relative variation of chemical mineralogical composition and isotopic composition variability between the samples, which will be determined using X-Ray diffraction, X-Ray fluorescence and mass spectroscopy elemental analyses. The samples will also be analyzed for their total carbon and nitrogen content and isotopic composition. All mineralogical and geochemical analyses will be carried out on of the different size fractions of each sample in order to estimate the sample scale variability. Results of soil elemental and physical analyses will be upscaled using a chronosequence classification in order to estimate the distribution and total storage of soil organic carbon in the different forefields.

Jin-Soo KIM (Pohang University of Science and Technology)
Reduced terrestrial productivity in North America associated with Arctic warming

Warming temperatures in the northern hemisphere have enhanced terrestrial productivity. Despite the warming trend, North America has experienced more frequent and stronger cold weather events during winters and springs linked to anomalous Arctic warming since 1990, which may affect terrestrial processes. Here we analyse multiple observation datasets and numerical model output to evaluate links between Arctic temperatures and primary productivity in North America. We find that positive springtime temperature anomalies in the Arctic have led to negative anomalies in gross primary productivity over most of North America during the last three decades, which amount to a net productivity decline of 0.31 PgC yr−1 across the continent. This decline is mainly explained by two factors: severe cold conditions in the northern North America and lower precipitation in the South Central United States. In addition, United States crop-yield data reveal that during years experiencing anomalous warming in the Arctic, yields declined by approximately 1 to 4% on average, with individual states experiencing declines of up to 20%. We conclude that the strengthening of Arctic warming anomalies in the past decades has remotely reduced productivity over North America.

Elena PARFENOVA (Forest Institute of Siberian Branch of the Russian Academy of Sciences)
Changing forest-tundra border in response to climate change in Central Siberia and methane-emissions

Our study area is the tundra zone over Central Siberia within the window 80-120°E and 70-78°N. Vegetation is presented generally erect dwarf-shrub, low shrub and non-tussock sedge moss tundra. Graminoid and forb tundra, prostrate dwarf-shrub herb tundra, as well sedge, moss, dwarf-shrub and low-shrub wetlands are less common. Climate data of January and July temperature, annual precipitation from 100 weather stations over the study area were used to calculate climatic indices to model the tundra distribution: Growing-degree days above 5ºC, Negative degree-days below 0ºC, and annual precipitation for the 1960-1990 baseline period using a digital elevation model of 1 km resolution using Hutchinson’s thin plate splines. The outcomes from twenty global climate models (CMIP5) of two scenarios rcp 2.6 and rcp 8.5 and their ensemble means were used to characterize a range of warming by the 2080s. Additionally, the permafrost distribution was modeled to identify both the forest advance into tundra and “hot spots” that may be responsible for methane emissions in a warming climate by the 2080s. The contemporary permafrost border was calculated as a function of the July and January temperatures and annual precipitation (R2 = 0.70). Stefan’s theoretical equation was used to calculate the future permafrost distribution and map its border. As predicted from the CMIP5 models, the 2080 climate would be characterized by milder and more moderate climates with less permafrost coverage by the 2080s. Climate change effects should shift vegetation zones. Habitats for northern vegetation classes tundra and forest-tundra would 2.5 times shrink and the border between forest and the tundra area would nearly disappear according to the 8.5 scenario and left only some percent according to the 2.6 scenario by the 2080s. Decreasing of the wetland-to-forest ratio will decline landscape CH4 emissions.

Martijn PALLANDT (Max Planck Institute for Biogeochemistry)
Evaluation and improvement of the pan-Arctic greenhouse gas monitoring network

The Arctic is a mayor driver of global climate and weather, and currently is one of the regions most affected by climate change. Understanding the processes that govern this region is thus essential. An extensive network of greenhouse gas (GHG) monitoring sites spans the Arctic. There are however uncertainties on its ability to monitor all essential processes. Proposed methane outgassing in the East Siberian Sea and wintertime fluxes are examples of areas that may be under represented. This study aims at assessing the current Arctic GHG monitoring network on its ability to represent the Arctic ecosystems, both spatially and temporally. An inventory will be made of existing monitoring sites and their data coverage. Based on this inventory ecosystem eddy covariance sites will be linked to ecoregions in a modeling exercise to assess the networks coverage of the ecoregions of the Arctic. For the atmospheric tower network an observing system synthetic experiment (OSSE) will be performed. In this OSSE a global climate model will prescribe an environment from which synthetic samples are generated as input for an inverse model. This allows for: the creation of events (methane outgassing), adding extra “towers” to assess their impact on the network, and comparison to perfectly know conditions. A final step will be the installation of an automatic flask sampler in Ambarchik (Russia) near the east Siberian sea, with as main goal to differentiate between anthropogenic and natural carbon sources in the Arctic, and between newly metabolized and paleo-methane in the East Siberian sea.

Efrén LÓPEZ BLANCO (Aarhus University / University of Edinburgh)
Analysis on inter-annual variability of CO2 exchange in Arctic tundra: a model-data approach

Arctic ecosystems are exposed to rapid changes influenced by the climate variability; hence a major concern is how the carbon (C) exchange balance will respond to climate change. The process-based understanding of C balances in the Arctic helps us to identify the mechanisms and responses of tundra ecosystems. Here we report the independent predictions of net ecosystem exchange (NEE), gross primary production (GPP) and ecosystem respiration (Reco) calculated from the soil-plant-atmosphere (SPA) model across eight full annual cycles. The model products are validated with observational data obtained from the Greenland Ecosystem Monitoring (GEM) program in West Greenland tundra (64° N). Overall, the model results explain ca. 70%, 75% and 50% of the variance in NEE, GPP and Reco respectively using data on meteorology and local vegetation and soil structure. The estimated leaf area index (LAI) is able to explain more than 80% of the plant greenness variation, which was used as a plant phenology proxy. The full annual cumulated NEE during the 2008-2015 period was -0.13 g C m-2 on average (range -30.6 to 34.1 g C m-2), while GPP was -214.6 g C /m2 (-126.2 to -332.8 g C m-2) and Reco was 214.4 g C m-2 (213.9 to 302.2 g C m-2). We found that the model supports the main finding from our previous analysis on flux responses to meteorological variations and biological disturbance. Here, large inter-annual variations in GPP and Reco are also compensatory, and so NEE remains stable across climatically diverse snow-free seasons. Further, we note evidence that autotrophic respiration (Ra) is highly correlated with GPP (R2 = 0.93, p < 0.001), concluding that this relation drives the insensitivity of NEE. Interestingly, the model quantifies the contribution of the larvae outbreak occurred in 2011 in about 40%. With this study, we demonstrate the importance of incorporate wintertime periods allowing a more comprehensive understanding of complete C budgets and the delayed effect of wintertime conditions on the C fluxes.

Zhan WANG (Yonsei University)
The Influence of Shrub Expansion on Permafrost Thawing in Arctic

The vegetation in Arctic has increased for decades with land cover and vegetation type change. In order to isolate the influence of the shrub expansion on the permafrost thawing, the Arctic terrestrial ecosystem in recent decades will be simulated using the Community Land Model (CLM) with and without the vegetation dynamics. The simulated snowpack, surface energy exchange, soil heat flux and the active layer thickness will be compared between the CLM-Carbon Nitrogen Model (CN) and the CLM-Carbon Nitrogen Dynamic Global Vegetation Model (CNDV). Furthermore the Moderate Resolution Imaging Spectroradiometer (MODIS) land cover product (MCD12Q1) will be utilized to evaluate the outcomes of vegetation dynamics in CLM-CNDV.

Chao WU (Tsinghua University / University of Exeter)
Contribution of fires in regulating global boreal forest composition and carbon cycle

Climate is well known as an important determinant of biogeography. Although the climate is directly important for vegetation composition in the boreal forest, and indirectly these ecosystems, as with other flammable ecosystems of the world, are strongly sensitive to fire disturbance. However, the specific contribution of fire disturbance in global boreal forest composition and carbon cycle is poorly understood. Here we used eight different fire models from FireMIP to explore the contribution of fire in global boreal forests composition and carbon cycle. The results will benefit the fire management and policy makers.

Arman GANJI (University of Quebec at Montreal)
Sub-grid representation of snow in Land Surface Models
Dae Il JEONG (University of Quebec at Montreal)
Detection and attribution of Spring Snow Equivalent (SWE) changes over the Northern Hemisphere
Gulilat Tefera DIRO (University of Quebec at Montreal)
Added value of dynamical downscaling of winter seasonal forecasts over North America
Oleksandr HUZIY (University of Quebec at Montreal)
On the representation of heavy lake-effect snow events for the Laurentian Great Lakes region in a Regional Climate Model
Bernardo TEUFEL (University of Quebec at Montreal)
The projected impact of dynamic vegetation on climate change over the pan-Arctic
Caio RUMAN (University of Quebec at Montreal)
Retreating Canadian glaciers and their implications for regional climate and hydrology in future climate

Arctic Terrestrial Modelling Workshop Participants

  1. Adam MONAHAN (University of Victoria, Canada)
  2. Adrian LOCK (Met Office, United Kingdom)
  3. Annett BARTSCH (Central Institute for Meteorology and Geodynamics, Austria)
  4. Atul JAIN (University of Illinois, USA)
  5. Bakr BADAWY (University of Waterloo, Canada)
  6. Benjamin SMITH (Lund University, Sweden)
  7. Bernardo TEUFEL (University of Quebec at Montreal, Canada)
  8. Cathy WILSON (Los Alamos National Laboratory, USA)
  9. Chad THACKERAY (University of Waterloo, Canada)
  10. Chao WU (Tsinghua University, China / University of Exeter, United Kingdom)
  11. Charles MILLER (Jet Propulsion Laboratory, USA)
  12. Christine DELIRE (Meteo-France, France)
  13. David LAWRENCE (National Center for Atmospheric Research, USA)
  14. Diana VERSEGHY (Environment and Climate Change Canada, Canada)
  15. Efrén LÓPEZ BLANCO (Aarhus University, Denmark / University of Edinburgh, United Kingdom)
  16. Eleanor BLYTH (Natural Environment Research Council, United Kingdom)
  17. Eleanor BURKE (Met Office, United Kingdom)
  18. Elena PARFENOVA (Forest Institute of Siberian Branch of the Russian Academy of Sciences, Russia)
  19. Emanuel DUTRA (University of Lisbon, Portugal)
  20. Eric BAZILE (Meteo-France, France)
  21. Friedrich OBLEITNER (Innsbruck University, Austria)
  22. Frode STORDAL (University of Oslo, Norway)
  23. Gabriele ARDUINI (ECMWF, United Kingdom)
  24. Gerhard KRINNER (National Center for Scientific Research, France)
  25. Gianpaolo BALSAMO (ECMWF, United Kingdom)
  26. Gulilat Tefera DIRO (University of Quebec at Montreal, Canada)
  27. Guy SCHURGERS (University of Copenhagen, Denmark)
  28. Heather KROPP (Colgate University, USA)
  29. Hui TANG (University of Oslo, Norway)
  30. Igor APPEL (TAG LLC, USA)
  31. Javier FOCHESATTO (University of Alaska Fairbanks, USA)
  32. Jin-Soo KIM (Pohang University of Science and Technology, South Korea)
  33. Joe MELTON (Environment and Climate Change Canada, Canada)
  34. Jürgen HELMERT (German Meteorological Office, Germany)
  35. Laxmi SUSHAMA (University of Quebec at Montreal, Canada)
  36. Markus TODT (Northumbria University, United Kingdom)
  37. Martijn PALLANDT (Max Planck Institute for Biogeochemistry, Germany)
  38. Michael RAWLINS (University of Massachusetts - Amherst, USA)
  39. Mousong WU (Lund University, Sweden)
  40. Patrick SAMUELSSON (Swedish Meteorological and Hydrological Institute, Sweden)
  41. Paul BARTLETT (Environment and Climate Change Canada, Canada)
  42. Paul MILLER (Lund University, Sweden)
  43. Pavel KOTOV (Lomonosov Moscow State University, Russia)
  44. Peter HORVATH (University of Oslo, Norway)
  45. Robin WOJCIK (German Research Centre for Geosciences, Germany)
  46. Stan WULLSCHLEGER (Oak Ridge National Laboratory, USA)
  47. Vladimir ROMANOVSKY (University of Alaska Fairbanks, USA)
  48. Wenxin ZHANG (University of Copenhagen, Denmark)
  49. Wolfgang BUERMANN (University of Leeds, United Kingdom)
  50. Wouter HANTSON (University of Maine, USA)
  51. Yuanchao FAN (Uni Research Climate & Bjerknes Centre for Climate Research, Norway)
  52. Zhan WANG (Yonsei University, South Korea)