Publications

Abstract The complexity and interconnectivity of food systems and climate requires new thinking and research designs that better address the real‐world challenges of securing the resilience and sustainability of human and environmental systems. Central to such an approach is coherent action across sectors and scales. Although inter‐and transdisciplinary approaches are widely discussed, no convergence model exists to detect and prepare for food system vulnerabilities emerging from disruptions in climate systems, or to address the contributions to climate change from food system functions. Convergence research is critical to solving these vexing dynamics by integrating knowledge from multiple scientific domains to inform societal action. Here, we present a new convergent science model that incorporates four key components at the global, national and local level. Through the newly created Food and Climate Systems Transformation Alliance, we are now operationalizing, testing and refining the model to promote science convergence for tackling systemic vulnerabilities in the current food paradigm. Globally, funding relating to climate change and food systems transformation needs to pivot to support the levels of ambition, magnitude of need and complexity of challenges posed.

A new estimate of Tibetan soil carbon pool, at a 0.1° spatial resolution, obtained by including paleoclimate, modern climate and landform, soil geochemical properties in multiple machine learning algorithms.

<strong class="journal-contentHeaderColor">Abstract.</strong> Knowledge of the spatial distribution of the fluxes of greenhouse gases and their temporal variability as well as flux attribution to natural and anthropogenic processes is essential to monitoring the progress in mitigating anthropogenic emissions under the Paris Agreement and to inform its Global Stocktake. This study provides a consolidated synthesis of CH₄ and N₂O emissions using bottom-up (BU) and top-down (TD) approaches for the European Union and UK (EU27+UK) and updates earlier syntheses (Petrescu et al., 2020, 2021). The work integrates updated emission inventory data, process-based model results, data-driven sector model results, inverse modelling estimates, and extends the previous period 1990&ndash;2017 to 2020. BU and TD products are compared with European National GHG Inventories (NGHGI) reported by Parties under the United Nations Framework Convention on Climate Change (UNFCCC) in 2021. The uncertainties of NGHGIs were evaluated using the standard deviation obtained by varying parameters of inventory calculations, reported by the EU Member States following the guidelines of the Intergovernmental Panel on Climate Change (IPCC) and harmonized by gap-filling procedures. Variation in estimates produced with other methods, such as atmospheric inversion models (TD) or spatially disaggregated inventory datasets (BU), arise from diverse sources including within-model uncertainty related to parameterization as well as structural differences between models. By comparing NGHGIs with other approaches, the activities included are a key source of bias between estimates e.g. anthropogenic and natural fluxes, which, in atmospheric inversions are sensitive to the prior geospatial distribution of emissions. <strong>For CH₄ emissions</strong>, over the updated 2015&ndash;2019 period, which covers a sufficiently robust number of overlapping estimates, and most importantly the NGHGIs, the anthropogenic BU approaches are directly comparable, accounting for mean emissions of 20.5 Tg CH₄ yr^&minus;1 (EDGAR v5v6.0, last year 2018) and 18.4 Tg CH₄ yr^&minus;1 (GAINS, 2015), close to the NGHGI estimates of 17.5 &plusmn; 2.1 Tg CH₄ yr^&minus;1. TD inversions estimates give higher emission estimates, as they also detect natural emissions. Over the same period, high resolution regional TD inversions report a mean emission of 34 Tg CH₄ yr^&minus;1. Coarser-resolution global-scale TD inversions result in emission estimates of 23 Tg CH₄ yr^&minus;1 and 24 Tg CH₄ yr^&minus;1 inferred from GOSAT and surface (SURF) network atmospheric measurements, respectively. The magnitude of natural peatland and mineral soils emissions from the JSBACH-HIMMELI model, natural rivers, lakes and reservoirs emissions, geological sources and biomass burning together could account for the gap between NGHGI and inversions and account for 8 Tg CH₄ yr^&minus;1. <strong>For N₂O emissions</strong>, over the 2015&ndash;2019 period, both BU products (EDGAR v5v6.0 and GAINS) report a mean value of anthropogenic emissions of 0.9 Tg N₂O yr^&minus;1, close to the NGHGI data (0.8 &plusmn; 55 % Tg N₂O yr^&minus;1). Over the same period, the mean of TD global and regional inversions was 1.4 Tg N₂O yr^&minus;1 (excluding TOMCAT which reported no data). The TD and BU comparison method defined in this study can be "operationalized" for future annual updates for the calculation of CH₄ and N₂O budgets at the national and EU27+UK scales. Future comparability will be enhanced with further steps involving analysis at finer temporal resolutions and estimation of emissions over intra-annual timescales, of great importance for CH₄ and N₂O, which may help identify sector contributions to divergence between prior and posterior estimates at the annual/inter-annual scale. Even if currently comparison between CH₄ and N₂O inversions estimates and NGHGIs is highly uncertain because of the large spread in the inversion results, TD inversions inferred from atmospheric observations represent the most independent data against which inventory totals can be compared. With anticipated improvements in atmospheric modelling and observations, as well as modelling of natural fluxes, TD inversions may arguably emerge as the most powerful tool for verifying emissions inventories for CH₄, N₂O and other GHGs. The referenced datasets related to figures are visualized at <a href="https://doi.org/10.5281/zenodo.6992472" target="_blank" rel="noopener">https://doi.org/10.5281/zenodo.6992472</a> (Petrescu et al., 2022).

Abstract. Reliable quantification of the sources and sinks of atmospheric carbon dioxide (CO2), including that of their trends and uncertainties, is essential to monitoring the progress in mitigating anthropogenic emissions under the Kyoto Protocol and the Paris Agreement. This study provides a consolidated synthesis of estimates for all anthropogenic and natural sources and sinks of CO2 for the European Union and UK (EU27 + UK), derived from a combination of state-of-the-art bottom-up (BU) and top-down (TD) data sources and models. Given the wide scope of the work and the variety of datasets involved, this study focuses on identifying essential questions which need to be answered to properly understand the differences between various datasets, in particular with regards to the less-well-characterized fluxes from managed ecosystems. The work integrates recent emission inventory data, process-based ecosystem model results, data-driven sector model results and inverse modeling estimates over the period 1990–2018. BU and TD products are compared with European national greenhouse gas inventories (NGHGIs) reported under the UNFCCC in 2019, aiming to assess and understand the differences between approaches. For the uncertainties in NGHGIs, we used the standard deviation obtained by varying parameters of inventory calculations, reported by the member states following the IPCC Guidelines. Variation in estimates produced with other methods, like atmospheric inversion models (TD) or spatially disaggregated inventory datasets (BU), arises from diverse sources including within-model uncertainty related to parameterization as well as structural differences between models. In comparing NGHGIs with other approaches, a key source of uncertainty is that related to different system boundaries and emission categories (CO2 fossil) and the use of different land use definitions for reporting emissions from land use, land use change and forestry (LULUCF) activities (CO2 land). At the EU27 + UK level, the NGHGI (2019) fossil CO2 emissions (including cement production) account for 2624 Tg CO2 in 2014 while all the other seven bottom-up sources are consistent with the NGHGIs and report a mean of 2588 (± 463 Tg CO2). The inversion reports 2700 Tg CO2 (± 480 Tg CO2), which is well in line with the national inventories. Over 2011–2015, the CO2 land sources and sinks from NGHGI estimates report −90 Tg C yr−1 ± 30 Tg C yr−1 while all other BU approaches report a mean sink of −98 Tg C yr−1 (± 362 Tg of C from dynamic global vegetation models only). For the TD model ensemble results, we observe a much larger spread for regional inversions (i.e., mean of 253 Tg C yr−1 ± 400 Tg C yr−1). This concludes that (a) current independent approaches are consistent with NGHGIs and (b) their uncertainty is too large to allow a verification because of model differences and probably also because of the definition of “CO2 flux” obtained from different approaches. The referenced datasets related to figures are visualized at https://doi.org/10.5281/zenodo.4626578 (Petrescu et al., 2020a).

In their letter, Leifeld et al. (1) argue that our metaanalysis to identify differences in soil organic carbon (SOC) between organic (OF) and nonorganic farming [conventional farming (CF)] (2) selected CF systems that were nonrepresentative. This was not the case. We included data from all available pairwise field comparisons between OF and CF identified in the literature. The observed difference in external carbon (C) inputs between OF and CF did not result from a bias in the selection of studies/treatments but was attributable to the fact that the field comparisons we analyzed (2) were not from fertilization experiments but from pairwise farming system comparisons where the design and the underlying treatments reflected the current farming practices in the region in which the studies were conducted at the time the experiments were initiated.

Cotton is widely used in textile, decoration, and industry, but it is also threatened by soil salinization. Drip irrigation plays an important role in improving water and fertilization utilization efficiency and ensuring crop production in arid areas. Accurate prediction of soil salinity and crop evapotranspiration under drip irrigation is essential to guide water management practices in arid and saline areas. However, traditional hydrological models such as Hydrus require more variety of input parameters and user expertise, which limits its application in practice, and machine learning (ML) provides a potential alternative. Based on a global dataset collected from 134 pieces of literature, we proposed a method to comprehensively simulate soil salinity, evapotranspiration (ET) and cotton yield. Results showed that it was recommended to predict soil salinity, crop evapotranspiration and cotton yield based on soil data (bulk density), meteorological factors, irrigation data and other data. Among them, meteorological factors include annual average temperature, total precipitation, year. Irrigation data include salinity in irrigation water, soil matric potential and irrigation water volume, while other data include soil depth, distance from dripper, days after sowing (for EC and soil salinity), fertilization rate (for yield and ET). The accuracy of the model has reached a satisfactory level, R 2 in 0.78-0.99. The performance of stacking ensemble ML was better than that of a single model, i.e., gradient boosting decision tree (GBDT); random forest (RF); extreme gradient boosting regression (XGBR), with R 2 increased by 0.02%-19.31%. In all input combinations, other data have a greater impact on the model accuracy, while the RMSE of the S1 scenario (input without meteorological factors) without meteorological data has little difference, which is -34.22%~19.20% higher than that of full input. Given the wide application of drip irrigation in cotton, we recommend the application of ensemble ML to predict soil salinity and crop evapotranspiration, thus serving as the basis for adjusting the irrigation schedule.

Abstract In this paper, we focus on the impact on soil organic carbon ( SOC ) of two dedicated energy crops: perennial grass M iscanthus x G iganteus ( M iscanthus ) and short rotation coppice ( SRC )‐willow. The amount of SOC sequestered in the soil is a function of site‐specific factors including soil texture, management practices, initial SOC levels and climate; for these reasons, both losses and gains in SOC were observed in previous M iscanthus and SRC ‐willow studies. The ECOSSE model was developed to simulate soil C dynamics and greenhouse gas emissions in mineral and organic soils. The performance of ECOSSE has already been tested at site level to simulate the impacts of land‐use change to short rotation forestry ( SRF ) on SOC . However, it has not been extensively evaluated under other bioenergy plantations, such as M iscanthus and SRC ‐willow. Twenty‐nine locations in the U nited K ingdom, comprising 19 paired transitions to SRC ‐willow and 20 paired transitions to M iscanthus , were selected to evaluate the performance of ECOSSE in predicting SOC and SOC change from conventional systems (arable and grassland) to these selected bioenergy crops. The results of the present work revealed a strong correlation between modelled and measured SOC and SOC change after transition to M iscanthus and SRC ‐willow plantations, at two soil depths (0–30 and 0–100 cm), as well as the absence of significant bias in the model. Moreover, model error was within (i.e. not significantly larger than) the measurement error. The high degrees of association and coincidence with measured SOC under M iscanthus and SRC ‐willow plantations in the U nited K ingdom, provide confidence in using this process‐based model for quantitatively predicting the impacts of future land use on SOC , at site level as well as at national level.

Abstract. Agricultural intensification has contributed greatly to the sustained food supply of China's 1.3 billion population over the 30 year period during 1982–2011. Intensification has several and widely recognized negative environmental impacts including depletion of water resources, pollution of water bodies, greenhouse gas emissions and soil acidification. However, there have been few studies over this period on the impacts of intensification on soil organic carbon (SOC) at the regional level. The present study was conducted in Huantai county, a typical intensive farming region in Northern China, to analyze the temporal dynamics of SOC influenced by climate and farming practices. The results indicate that from 1982 to 2011, SOC content and stock in the 0–20 cm layer of the cropland increased from 7.8 ± 1.6 to 11.0 ± 2.3 g kg–1 (41%) and 21 ± 4.3 to 33.0 ± 7.0 Mg ha–1 (54%), respectively. The SOC stock (0–20 cm) of the farmland for the entire county increased from 0.75 to 1.2 Tg (59%). Correlation analysis revealed that incorporation of crop residues significantly increased SOC, while increase in the mean annual temperature decreased the SOC level. Therefore, agricultural intensification has increased crop productivity and contributed to SOC sequestration in Northern China. In the near future, more appropriate technologies and practices must be developed and implemented for a maintenance or enhancement of SOC in this region and elsewhere in Northern China, that also reduce non-CO2 greenhouse gas emissions, since the climate benefit from the additional SOC storage is estimated to be smaller than the negative climate impacts of N2O from N fertilizer additions.