CM2Mc-LPJmL v1.0: biophysical coupling of a process-based dynamic vegetation model with managed land to a general circulation model

The terrestrial biosphere is exposed to land-use and climate change, which not only affects vegetation dynamics but also changes land–atmosphere feedbacks. Specifically, changes in land cover affect biophysical feedbacks of water and energy, thereby contributing to climate change. In this study, we couple the well-established and comprehensively validated dynamic global vegetation model LPJmL5 (Lund–Potsdam–Jena managed Land) to the coupled climate model CM2Mc, the latter of which is based on the atmosphere model AM2 and the ocean model MOM5 (Modular Ocean Model 5), and name it CM2Mc-LPJmL. In CM2Mc, we replace the simple land-surface model LaD (Land Dynamics; where vegetation is static and prescribed) with LPJmL5, and we fully couple the water and energy cycles using the Geophysical Fluid Dynamics Laboratory (GFDL) Flexible Modeling System (FMS). Several improvements to LPJmL5 were implemented to allow a fully functional biophysical coupling. These include a sub-daily cycle for calculating energy and water fluxes, conductance of the soil evaporation and plant interception, canopy-layer humidity, and the surface energy balance in order to calculate the surface and canopy-layer temperature within LPJmL5. Exchanging LaD with LPJmL5 and, therefore, switching from a static and prescribed vegetation to a dynamic vegetation allows us to model important biospheric processes, including fire, mortality, permafrost, hydrological cycling and the impacts of managed land (crop growth and irrigation). Our results show that CM2Mc-LPJmL has similar temperature and precipitation biases to the original CM2Mc model with LaD. The performance of LPJmL5 in the coupled system compared to Earth observation data and to LPJmL offline simulation results is within acceptable error margins. The historical global mean temperature evolution of our model setup is within the range of CMIP5 (Coupled Model Intercomparison Project Phase 5) models. The comparison of model runs with and without land-use change shows a partially warmer and drier climate state across the global land surface. CM2Mc-LPJmL opens new opportunities to investigate important biophysical vegetation–climate feedbacks with a state-of-the-art and process-based dynamic vegetation model.

Capturing site-to-site variability through Hierarchical Bayesian calibration of a process-based dynamic vegetation model

Process-based ecosystem models help us understand and predict ecosystem processes, but using them has long involved a difficult choice between performing data- and labor-intensive site-level calibrations or relying on general parameters that may not reflect local conditions. Hierarchical Bayesian (HB) calibration provides a third option that frees modelers from assuming model parameters to be completely generic or completely site-specific and allows a formal distinction between prediction at known calibration sites and “out-of-sample” prediction to new sites. Here, we compare calibrations of a process-based dynamic vegetation model to eddy-covariance data across 12 temperate deciduous Ameriflux sites fit using either site-specific, joint cross-site, or HB approaches. To be able to apply HB to computationally demanding process-based models we introduce a novel emulator-based HB calibration tool, which we make available through the PEcAn community cyberinfrastructure. Using these calibrations to make predictions at held-out tower sites, we show that the joint cross-site calibration is falsely over-confident because it neglects parameter variability across sites and therefore underestimates variance in parameter distributions. By showing which parameters show high site-to-site variability, HB calibration also formally gives us a structure that can detect which process representations are missing from the models and prioritize errors based on the magnitude of the associated uncertainty. For example, in our case-study, we were able to identify large site-to-site variability in the parameters related to the temperature responses of respiration and photosynthesis, associated with a lack of thermal acclimation and adaptation in the model. Moving forward, HB approaches present important new opportunities for statistical modeling of the spatiotemporal variability in modeled parameters and processes that yields both new insights and improved predictions.

Cereal yield forecasting in Morocco using the CARAIB dynamic vegetation model driven by HadGEM2-AO projections

<p>Food security, in Morocco as in many parts of the world, depends heavily on cereal production which fluctuates relying on weather conditions. In fact, Morocco has a production system for cereals which is dominated by rainfed. It is therefore necessary to further develop knowledge about climate change and strengthen forecasting systems for predicting the impacts of climate change.</p><p>Our research, funded by a bilateral project of Wallonie-Bruxelles International, aims to study the response of cereal production to climate change, using the dynamic vegetation model CARAIB (CARbon Assimilation In the Biosphere) developed within the Unit for Modelling of Climate and Biogeochemical Cycles (UMCCB) of the University of Liège. This spatial model includes crops and natural vegetation and may react dynamically to land use changes. Originally constructed to study vegetation dynamics and carbon cycle, it includes coupled hydrological, biogeochemical, biogeographical and fire modules. These modules respectively describe the exchange of water between the atmosphere, the soil and the vegetation, the photosynthetic production and the evolution of carbon stocks and fluxes in this vegetation-soil system. For crops, a specific module describes basic management parameters (sowing, harvest, rotation) and phenological phases.</p><p>The simulations are performed across all Morocco using different input data. The three main cereal crops simulated include soft wheat, durum wheat and barley, they are grown in all provinces and all agro-ecological zones. Regarding climatic inputs, we’re using two sets of data: the first one is interpolated and bias-corrected fields from the climate model HadGEM2-AO for the historical period (1990-2005), in addition to three different Representative Concentration Pathway scenarios (RCP2.6, RCP4.5 and RCP8.5) from 2005 to 2100. The second one is high resolution (30 arc sec) gridded climate data derived from WorldClim combined with interpolated anomalies from CRU (Climatic Research Unit) over the historical period 1990 to 2018.</p><p>After obtaining preliminary results for the past period, and in order to improve the prediction using the field data which are the observed yields, we performed a sensitivity analysis. We used the One-at-a-time (OAT) approach by moving one input variable, keeping others at their baseline (nominal) values, then, returning the variable to its nominal value, then repeating for each of the other inputs in the same way. Sensitivity may then be measured by monitoring changes in the output, using linear regression. The inputs studied are the initial value of carbon pool, leaf C/N ratio, water stress, sowing date, GDD harvest, stomatal conductance parameters, specific leaf area, and rooting depth.</p>

Assessing the effects of climate and land use land cover changes on recent carbon storage in terrestrial ecosystem using model-satellite approach over Wallonia, Belgium

<p>The use of a dynamic vegetation model, CARAIB, to estimate carbon sequestration from land-use and land-cover change (LULCC) offers a new approach for spatial and temporal details of carbon sink and for terrestrial ecosystem productivity affected by LULCC. Using the remote sensing satellite imagery (Landsat) we explore the role of land use land cover change (LULCC) in modifying the terrestrial carbon sequestration. We have constructed our LULCC data over Wallonia, Belgium, and compared it with the ground-based statistical data. However, the results from the satellite base LULCC are overestimating the forest data due to the single isolated trees. We know forests play an important role in mitigating climate change by capturing and sequestering atmospheric carbon. Overall, the conversion of land and increase in urban land can impact the environment. Moreover, quantitative estimation of the temporal and spatial pattern of carbon storage with the change in land use land cover is critical to estimate. The objective of this study is to estimate the inter-annual variability in carbon sequestration with the change in land use land cover. Here, with the CARAIB dynamic vegetation model, we perform simulations using remote sensing satellite-based LULCC data to analyse the sensitivity of the carbon sequestration. We propose a new method of using satellite and machine learning-based observation to reconstruct historical LULCC. It will quantify the spatial and temporal variability of land-use change during the 1985-2020 periods over Wallonia, Belgium at high resolution. This study will give the space to analyse past information and hence calibrate the dynamic vegetation model to minimize uncertainty in the future projection (until 2070). Further, we will also analyse the change in other climate variables, such as CO<sub>2</sub>, temperature, etc. Overall, this study allows us to understand the effect of changing land-use patterns and to constrain the model with an improved input dataset which minimizes the uncertainty in model estimation.</p>

CM2Mc-LPJmL v1.0: Biophysical coupling of a process-based dynamic vegetation model with managed land to a general circulation model

The terrestrial biosphere is exposed to land-use and climate change, which not only affects vegetation dynamics, but also changes land-atmosphere feedbacks. Specifically, changes in land-cover affect biophysical feedbacks of water and energy, therefore contributing to climate change. In this study, we couple the well established and comprehensively validated Dynamic Global Vegetation Model LPJmL5 to the coupled climate model CM2Mc, which is based on the atmosphere model AM2 and the ocean model MOM5 (CM2Mc-LPJmL). In CM2Mc, we replace the simple land surface model LaD (where vegetation is static and prescribed) with LPJmL5 and fully couple the water and energy cycles using the Geophysical Fluid Dynamics Laboratory (GFDL) Flexible Modeling System (FMS). Several improvements to LPJmL5 were implemented to allow a fully functional biophysical coupling. These include a sub-daily cycle for calculating energy and water fluxes, a conductance of the soil evaporation and plant interception, a canopy-layer humidity, and the surface energy balance in order to calculate the surface and canopy layer temperature within LPJmL5. Exchanging LaD by LPJmL5, and therefore switching from a static and prescribed vegetation to a dynamic vegetation, allows us to model important biosphere processes, including fire, mortality, permafrost, hydrological cycling, and the impacts of managed land (crop growth and irrigation). Our results show that CM2Mc-LPJmL has similar temperature and precipitation biases as the original CM2Mc model with LaD. Performance of LPJmL5 in the coupled system compared to Earth observation data and to LPJmL offline simulation results is within acceptable error margins. The historic global mean temperature evolution of our model setup is within the range of CMIP5 models. The comparison of model runs with and without land-use change shows a partially warmer and drier climate state across the global land surface. CM2Mc-LPJmL opens new opportunities to investigate important biophysical vegetation-climate feedbacks with a state-of-the-art and process-based dynamic vegetation model.

Accounting for forest management in the estimation of forest carbon balance using the dynamic vegetation model LPJ-GUESS (v4.0, r9333): Implementation and evaluation of simulations for Europe

Global forests are the main component of the land carbon sink, which acts as a partial buffer to CO2 emissions into the atmosphere. Dynamic vegetation models offer an approach to making projections of the development of forest carbon sink capacity in a future climate. Forest management capabilities in dynamic vegetation models are important to include the effects of age and species structure and wood harvest on carbon stocks and carbon storage potential. This article describes the introduction of a forest management module in the dynamic vegetation model LPJ-GUESS. Different age- and species-structure setup strategies and harvest alternatives are introduced. The model is used to represent current European forests and an automated harvest strategy is applied. Modelled carbon stocks and fluxes are evaluated against observed data at the continent and country levels. Including wood harvest in simulations increases the total European carbon sink by 32 % in 1991–2015 and improves the fit to the reported European carbon sink, growing stock and net annual increment (NAI). Growing stock (156 m3 ha−1) and NAI (5.4 m3 ha−1 y−1) densities in 2010 are close to reported values, while the carbon sink density in 2000–2007 (0.085 kgC m−2 y−1) is 63 % of reported values. The fit of modelled values and observations for individual European countries vary, but NAI is generally closer to observations when including wood harvest in simulations.