{"type": "FeatureCollection", "features": [{"id": "10.1111/gcb.15441", "type": "Feature", "geometry": null, "properties": {"license": "Open Access", "updated": "2026-09-21T16:17:34Z", "type": "Journal Article", "created": "2020-11-07", "title": "Ensemble modelling, uncertainty and robust predictions of organic carbon in long\u2010term bare\u2010fallow soils", "description": "Abstract

Simulation models represent soil organic carbon (SOC) dynamics in global carbon (C) cycle scenarios to support climate\uffe2\uff80\uff90change studies. It is imperative to increase confidence in long\uffe2\uff80\uff90term predictions of SOC dynamics by reducing the uncertainty in model estimates. We evaluated SOC simulated from an ensemble of 26 process\uffe2\uff80\uff90based C models by comparing simulations to experimental data from seven long\uffe2\uff80\uff90term bare\uffe2\uff80\uff90fallow (vegetation\uffe2\uff80\uff90free) plots at six sites: Denmark (two sites), France, Russia, Sweden and the United Kingdom. The decay of SOC in these plots has been monitored for decades since the last inputs of plant material, providing the opportunity to test decomposition without the continuous input of new organic material. The models were run independently over multi\uffe2\uff80\uff90year simulation periods (from 28 to 80\uffc2\uffa0years) in a blind test with no calibration (Bln) and with the following three calibration scenarios, each providing different levels of information and/or allowing different levels of model fitting: (a) calibrating decomposition parameters separately at each experimental site (Spe); (b) using a generic, knowledge\uffe2\uff80\uff90based, parameterization applicable in the Central European region (Gen); and (c) using a combination of both (a) and (b) strategies (Mix). We addressed uncertainties from different modelling approaches with or without spin\uffe2\uff80\uff90up initialization of SOC. Changes in the multi\uffe2\uff80\uff90model median (MMM) of SOC were used as descriptors of the ensemble performance. On average across sites, Gen proved adequate in describing changes in SOC, with MMM equal to average SOC (and standard deviation) of 39.2 (\uffc2\uffb115.5)\uffc2\uffa0Mg\uffc2\uffa0C/ha compared to the observed mean of 36.0 (\uffc2\uffb119.7)\uffc2\uffa0Mg\uffc2\uffa0C/ha (last observed year), indicating sufficiently reliable SOC estimates. Moving to Mix (37.5\uffc2\uffa0\uffc2\uffb1\uffc2\uffa016.7\uffc2\uffa0Mg\uffc2\uffa0C/ha) and Spe (36.8\uffc2\uffa0\uffc2\uffb1\uffc2\uffa019.8\uffc2\uffa0Mg\uffc2\uffa0C/ha) provided only marginal gains in accuracy, but modellers would need to apply more knowledge and a greater calibration effort than in Gen, thereby limiting the wider applicability of models.There is a clear need for transformative change in the land management and food production sectors to address the global land challenges of climate change mitigation, climate change adaptation, combatting land degradation and desertification, and delivering food security (referred to hereafter as \uffe2\uff80\uff9cland challenges\uffe2\uff80\uff9d). We assess the potential for 40 practices to address these land challenges and find that: Nine options deliver medium to large benefits for all four land challenges. A further two options have no global estimates for adaptation, but have medium to large benefits for all other land challenges. Five options have large mitigation potential (>3\uffc2\uffa0Gt CO2eq/year) without adverse impacts on the other land challenges. Five options have moderate mitigation potential, with no adverse impacts on the other land challenges. Sixteen practices have large adaptation potential (>25 million people benefit), without adverse side effects on other land challenges. Most practices can be applied without competing for available land. However, seven options could result in competition for land. A large number of practices do not require dedicated land, including several land management options, all value chain options, and all risk management options. Four options could greatly increase competition for land if applied at a large scale, though the impact is scale and context specific, highlighting the need for safeguards to ensure that expansion of land for mitigation does not impact natural systems and food security. A number of practices, such as increased food productivity, dietary change and reduced food loss and waste, can reduce demand for land conversion, thereby potentially freeing\uffe2\uff80\uff90up land and creating opportunities for enhanced implementation of other practices, making them important components of portfolios of practices to address the combined land challenges.There is a clear need for transformative change in the land management and food production sectors to address the global land challenges of climate change mitigation, climate change adaptation, combatting land degradation and desertification, and delivering food security (referred to hereafter as \uffe2\uff80\uff9cland challenges\uffe2\uff80\uff9d). We assess the potential for 40 practices to address these land challenges and find that: Nine options deliver medium to large benefits for all four land challenges. A further two options have no global estimates for adaptation, but have medium to large benefits for all other land challenges. Five options have large mitigation potential (>3\uffc2\uffa0Gt CO2eq/year) without adverse impacts on the other land challenges. Five options have moderate mitigation potential, with no adverse impacts on the other land challenges. Sixteen practices have large adaptation potential (>25 million people benefit), without adverse side effects on other land challenges. Most practices can be applied without competing for available land. However, seven options could result in competition for land. A large number of practices do not require dedicated land, including several land management options, all value chain options, and all risk management options. Four options could greatly increase competition for land if applied at a large scale, though the impact is scale and context specific, highlighting the need for safeguards to ensure that expansion of land for mitigation does not impact natural systems and food security. A number of practices, such as increased food productivity, dietary change and reduced food loss and waste, can reduce demand for land conversion, thereby potentially freeing\uffe2\uff80\uff90up land and creating opportunities for enhanced implementation of other practices, making them important components of portfolios of practices to address the combined land challenges.Simulation models represent soil organic carbon (SOC) dynamics in global carbon (C) cycle scenarios to support climate\uffe2\uff80\uff90change studies. It is imperative to increase confidence in long\uffe2\uff80\uff90term predictions of SOC dynamics by reducing the uncertainty in model estimates. We evaluated SOC simulated from an ensemble of 26 process\uffe2\uff80\uff90based C models by comparing simulations to experimental data from seven long\uffe2\uff80\uff90term bare\uffe2\uff80\uff90fallow (vegetation\uffe2\uff80\uff90free) plots at six sites: Denmark (two sites), France, Russia, Sweden and the United Kingdom. The decay of SOC in these plots has been monitored for decades since the last inputs of plant material, providing the opportunity to test decomposition without the continuous input of new organic material. The models were run independently over multi\uffe2\uff80\uff90year simulation periods (from 28 to 80\uffc2\uffa0years) in a blind test with no calibration (Bln) and with the following three calibration scenarios, each providing different levels of information and/or allowing different levels of model fitting: (a) calibrating decomposition parameters separately at each experimental site (Spe); (b) using a generic, knowledge\uffe2\uff80\uff90based, parameterization applicable in the Central European region (Gen); and (c) using a combination of both (a) and (b) strategies (Mix). We addressed uncertainties from different modelling approaches with or without spin\uffe2\uff80\uff90up initialization of SOC. Changes in the multi\uffe2\uff80\uff90model median (MMM) of SOC were used as descriptors of the ensemble performance. On average across sites, Gen proved adequate in describing changes in SOC, with MMM equal to average SOC (and standard deviation) of 39.2 (\uffc2\uffb115.5)\uffc2\uffa0Mg\uffc2\uffa0C/ha compared to the observed mean of 36.0 (\uffc2\uffb119.7)\uffc2\uffa0Mg\uffc2\uffa0C/ha (last observed year), indicating sufficiently reliable SOC estimates. Moving to Mix (37.5\uffc2\uffa0\uffc2\uffb1\uffc2\uffa016.7\uffc2\uffa0Mg\uffc2\uffa0C/ha) and Spe (36.8\uffc2\uffa0\uffc2\uffb1\uffc2\uffa019.8\uffc2\uffa0Mg\uffc2\uffa0C/ha) provided only marginal gains in accuracy, but modellers would need to apply more knowledge and a greater calibration effort than in Gen, thereby limiting the wider applicability of models.