ALBERT

Description: As a variant of Support Vector Machine (SVM), Large Margin Distribution Machine (LDM) has been validated to outperform SVM both theoretically and experimentally. Due to the inevitable noise in real applications, the credibility of different samples is not necessarily the same, which is neglected by most existing LDM models. To tackle the above problem, this paper first introduces fuzzy set theory into LDM, and proposes a Fuzzy Large Margin Distribution Machine (FLDM) with better robustness and performance. Considering the noise and uncertainty in datasets, sample points farther from the center of homogenous class are less reliable. Therefore, a fuzzy membership function based on the distance to the class center is utilized to characterize the confidence of each sample, i.e., the degree to which the sample belongs to a certain category. Furthermore, different strategies are developed to obtain class centers for linearly separable and linearly inseparable problems. Experiments conducted on both artificial and UCI datasets verified the superiority of FLDM from different perspectives.

Description: The problem of network disintegration, such as suppression of an epidemic spread and destabilization of terrorist networks, possesses extensive applications and has lately been the focus of growing interest. Many real-world complex systems are represented by spatial networks in which nodes and edges are spatially embedded. However, existing disintegration approaches for spatial network disintegration focus on singular aspects such as geospatial information or network topography, with insufficient modeling granularity. In this paper, we propose an effective and computationally efficient virtual node model that essentially integrates the geospatial information and topology of the network by modeling edges as virtual nodes with weights. Moreover, we employ Kernel Density Estimation, a well-known non-parametric technique for estimating the underlying probability density function of samples, to fit all nodes, comprising both network and virtual nodes, to identify the critical region of the spatial network, which is also the circular geographic region where disintegration occurs. Extensive numerical experiments on synthetic and real-world networks demonstrate that our method outperforms existing methods in terms of both effectiveness and efficiency, which provides a fresh perspective for modeling spatial networks.

Description: Bioenergy with carbon capture and storage (BECCS) can help stabilize the climate by extracting carbon dioxide from the atmosphere while producing renewable energy. However, biomass availability would limit the potential of BECCS, and biomass cropland expansion may threaten biodiversity, food security, and water supply. Replacing land-intensive foods can help unlock sustainable biomass production. Here, we estimated BECCS energy and negative emissions using biomass grown on freed-up land when replacing animal-source foods. Biomass production excludes agricultural expansion to protect biodiversity, ensures enough food supply globally to safeguard food security, and constrains irrigation to secure water for people and ecosystems. Negative emissions consider supply chain emissions and the forgone sequestration from natural revegetation. Results show that replacing 50% of animal products by 2050 could release enough land for BECCS to generate 26.4–39.5 EJelec/year, the scale of coal power today, while removing 5.9–9.3 GtCO2e/year from the atmosphere, almost what coal power emits today.

Description: We investigated the influence of the El Niño-Southern Oscillation (ENSO) on inter-annual precipitation variability in the far-eastern Pacific (FEP) and northern South America (NSA) using an approach based on phase synchronization (PS). First, we carried out a detailed analysis of observational data to define the inter-annual variability, eliminate the seasonal residual frequencies in hydro-climatic anomalies, and assess the statistical significance of PS. Additionally, we characterized the seasonality of regional patterns of sea surface temperature, surface pressure levels, low-level winds and precipitation anomalies associated with the ENSO states. We found that the positive (negative) precipitation anomalies experienced in the FEP and NSA differ from those previously reported in the literature. In particular, the Guianas (northeastern Amazon) and the Caribbean constitute two regions with negative (positive) rainfall anomalies during El Niño (La Niña), separated by a zone of non-significant anomalies along the Orinoco Low-level Jet corridor. Moreover, we showed that the ENSO signal is phase-locked with inter-annual rainfall and low-level wind variability in most of the study regions. Furthermore, we found consistency in the PS between the Central and Eastern Pacific El Niño indices and hydroclimatic anomalies over the Pacific. However, some areas exhibited PS, although they did not show significant precipitation anomalies, suggesting that the influence of ENSO on tropical climatology manifests not only in terms of the magnitude of anomalies but also in terms of the phases only. Our approach advances the understanding of climatic anomalies in tropical regions and provides new insights into the non-linear interactions between ENSO and hydroclimatic processes in tropical Americas.

Description: Lack of nitrogen limits food production in poor countries while excessive nitrogen use in industrial countries has led to transgression of the planetary boundary. However, the potential of spatial redistribution of nitrogen input for food security when returning to the safe boundary has not been quantified in a robust manner. Using an emulator of a global gridded crop model ensemble, we found that redistribution of current nitrogen input to major cereals among countries can double production in the most food insecure countries, while increasing global production of these crops by 12% with no notable regional loss or reducing the nitrogen input to the current production by one third. Redistribution of the input within the boundary increased production by 6–8% compared to the current relative distribution, increasing production in the food insecure countries by two thirds. Our findings provide georeferenced guidelines for redistributing nitrogen use to enhance food security while safeguarding the planet.

Description: Oil seed crops are the second most important field crops after cereals in the agricultural economy globally. The use and demand for oilseed crops such as groundnut, soybean and sunflower have grown significantly, but climate change is expected to alter the agroecological conditions required for oilseed crop production. This study aims to present an approach that utilizes decision-making tools to assess the potential climate change impacts on groundnut, soybean and sunflower yields and the greenhouse gas emissions from the management of the crops. The Decision Support Tool for Agrotechnology Transfer (DSSAT v4.7), a dynamic crop model and the Cool Farm Tool, a GHG calculator, was used to simulate yields and estimate GHG emissions from these crops, respectively. Four representative concentration pathways (RCPs 2.6, 4.5, 6.0, and 8.5), three nitrogen (0, 75, and 150 kg/ha) and phosphorous (0, 30 and 60 P kg/ha) fertilizer rates at three sites in Limpopo, South Africa (Ofcolaco, Syferkuil and Punda Maria) were used in field trials for calibrating the models. The highest yield was achieved by sunflower across all crops, years and sites. Soybean yield is projected to decrease across all sites and scenarios by 2030 and 2050, except at Ofcolaco, where yield increases of at least 15.6% is projected under the RCP 4.5 scenario. Positive climate change impacts are predicted for groundnut at Ofcolaco and Syferkuil by 2030 and 2050, while negative impacts with losses of up to 50% are projected under RCP 8.5 by 2050 at Punda Maria. Sunflower yield is projected to decrease across all sites and scenarios by 2030 and 2050. A comparison of the climate change impacts across sites shows that groundnut yield is projected to increase under climate change while notable yield losses are projected for sunflower and soybean. GHG emissions from the management of each crop showed that sunflower and groundnut production had the highest and lowest emissions across all sites respectively. With positive climate change impacts, a reduction of GHG emissions per ton per hectare was projected for groundnuts at Ofcolaco and Syferkuil and for sunflower in Ofcolaco in the future. However, the carbon footprint from groundnut is expected to increase by 40 to 107% in Punda Maria for the period up to 2030 and between 70-250% for 2050, with sunflower following a similar trend. We conclude that climate change will potentially reduce yield for oilseed crops while management will increase emissions. Therefore, in designing adaptation measures, there is a need to consider emission effects to gain a holistic understanding of how both climate change impacts on crops and mitigation efforts could be targeted.

Description: The planetary boundaries framework defines a safe operating space for humanity. To date, these boundaries have mostly been investigated separately, and it is unclear whether breaching one boundary can lead to the transgression of another. By employing a dynamic global vegetation model, we systematically simulate the strength and direction of the effects of different transgression levels of the climate change boundary (using climate output from ten CMIP6 models for CO2 levels ranging from 350 ppm to 1000 ppm). We focus on climate change-induced shifts of Earth’s major forest biomes, the control variable for the land-system change boundary, both by the end of this century and, to account for the long-term legacy effect, by the end of the millennium. Our simulations show that while staying within the 350 ppm climate change boundary co-stabilizes the land-system change boundary, breaching it (〉450 ppm) leads to its critical transgression with greater severity, the higher the ppm level rises and the more time passes. Specifically, this involves a poleward treeline shift, boreal forest dieback (nearly completely within its current area under extreme climate scenarios), competitive expansion of temperate forest into today’s boreal zone, and a slight tropical forest extension. These interacting changes also affect other planetary boundaries (freshwater change and biosphere integrity) and provide feedback to the climate change boundary itself. Our quantitative process-based study highlights the need for interactions to be studied for a systemic operationalization of the framework.

Description: Ecosystems are under multiple stressors and impacts can be measured with multiple variables. Humans have altered mass and energy flows of basically all ecosystems on Earth towards dangerous levels. However, integrating the data and synthesizing conclusions is becoming more and more complicated. Here we present an automated and easy to apply R package to assess terrestrial biosphere integrity which combines 2 complementary metrics: The BioCol metric quantifies the human colonization pressure exerted on the biosphere through alteration and extraction (appropriation) of net primary productivity, whereas the EcoRisk metric quantifies biogeochemical and vegetation structural changes as a proxy for the risk of ecosystem destabilization. Applied to simulations with the dynamic global vegetation model LPJmL5 for 1500–2016, we find that presently (period 2007–2016), large regions show modification and extraction of 〉25 % of the preindustrial potential net primary production, leading to drastic alterations in key ecosystem properties and suggesting a high risk for ecosystem destabilization. In consequence of these dynamics, EcoRisk shows particularly high values in regions with intense land use and deforestation, but also in regions prone to impacts of climate change such as the arctic and boreal zone. The metrics presented here enable global-scale, spatially explicit evaluation of historical and future states of the biosphere and are designed for use by the wider scientific community, not only limited to assessing biosphere integrity, but also to benchmark model performance. The package will be maintained on GitHub and through that we encourage application also to other models and data sets.

Description: Collective risk social dilemmas are at the heart of the most pressing global challenges we are facing today, including climate change mitigation and the overuse of natural resources. Previous research has framed this problem as a public goods game (PGG), where a dilemma arises between short-term interests and long-term sustainability. In the PGG, subjects are placed in groups and asked to choose between cooperation and defection, while keeping in mind their personal interests as well as the commons. Here, we explore how and to what extent the costly punishment of defectors is successful in enforcing cooperation by means of human experiments. We show that an apparent irrational underestimation of the risk of being punished plays an important role, and that for sufficiently high punishment fines, this vanishes and the threat of deterrence suffices to preserve the commons. Interestingly, however, we find that high fines not only avert freeriders, but they also demotivate some of the most generous altruists. As a consequence, the tragedy of the commons is predominantly averted due to cooperators that contribute only their “fair share” to the common pool. We also find that larger groups require larger fines for the deterrence of punishment to have the desired prosocial effect.

Description: Advances in the field of extreme event attribution allow to estimate how anthropogenic global warming affects the odds of individual climate disasters, such as river floods. Extreme event attribution typically uses precipitation as proxy for flooding. However, hydrological processes and antecedent conditions make the relation between precipitation and floods highly nonlinear. In addition, hydrology acknowledges that changes in floods can be strongly driven by changes in land-cover and by other human interventions in the hydrological system, such as irrigation and construction of dams. These drivers can either amplify, dampen or outweigh the effect of climate change on local flood occurrence. Neglecting these processes and drivers can lead to incorrect flood attribution. Including flooding explicitly, that is, using data and models of hydrology and hydrodynamics that can represent the relevant hydrological processes, will lead to more robust event attribution, and will account for the role of other drivers beyond climate change. Existing attempts are incomplete. We argue that the existing probabilistic framework for extreme event attribution can be extended to explicitly include floods for near-natural cases, where flood occurrence was unlikely to be influenced by land-cover change and human hydrological interventions. However, for the many cases where this assumption is not valid, a multi-driver framework for conditional event attribution needs to be established. Explicit flood attribution will have to grapple with uncertainties from lack of observations and compounding from the many processes involved. Further, it requires collaboration between climatologists and hydrologists, and promises to better address the needs of flood risk management.