Data-driven predictions of the time remaining until critical global warming thresholds are reached

Proceedings of the National Academy of Sciences, Год журнала: 2023, Номер 120(6)

Опубликована: Янв. 30, 2023

Leveraging artificial neural networks (ANNs) trained on climate model output, we use the spatial pattern of historical temperature observations to predict time until critical global warming thresholds are reached. Although no used during training, validation, or testing, ANNs accurately timing from maps annual temperature. The central estimate for 1.5 °C threshold is between 2033 and 2035, including a ±1σ range 2028 2039 in Intermediate (SSP2-4.5) forcing scenario, consistent with previous assessments. However, our data-driven approach also suggests substantial probability exceeding 2 even Low (SSP1-2.6) scenario. While there limitations approach, results suggest higher likelihood reaching scenario than indicated some assessments—though possibility that could be avoided not ruled out. Explainable AI methods reveal focus particular geographic regions Our framework provides unique, quantifying signal change constraining uncertainty projections. Given existing evidence accelerating risks natural human systems at °C, provide further high-impact over next three decades.

Язык: Английский

---

Co-Pyrolysis of biomass and plastic waste: Process prediction and optimization based on Artificial Intelligence and response optimizer surrogate model

Process Safety and Environmental Protection, Год журнала: 2024, Номер 186, С. 612 - 624

Опубликована: Апрель 10, 2024

Язык: Английский

---

The transition to sustainable combustion: Hydrogen- and carbon-based future fuels and methods for dealing with their challenges

Proceedings of the Combustion Institute, Год журнала: 2024, Номер 40(1-4), С. 105638 - 105638

Опубликована: Янв. 1, 2024

While the world is already facing substantial impacts of global warming, transition towards a sustainable-energy future slow because sheer scale energy needs that are presently satisfied mostly by combustion fossil fuels. Chemical carriers likely to play an essential role in systems, where harvesting and utilization renewable occur not necessarily at same time or place, hence long-time storage long-range transport needed. For this, hydrogen-based chemical carriers, such as hydrogen ammonia, will very important systems. Furthermore, there significant promise carbon-based fuels made from upgrading CO2, lignocellulosic biomass, combination both with electricity-derived hydrogen, yielding electro-fuels, biofuels, bio-hybrid fuels, respectively. The these combustion-based conversion has many advantages, e.g., versatile use for heat power, robust flexible technologies, suitability continuous transition. However, also challenges, which need be addressed discussed present paper. Hydrogen-based well known possess properties different conventional occurrence intrinsic flame instabilities lean premixed flames, can lead several-fold increase consumption speeds wide range conditions. Bio-hybrid show enormous molecular diversity allowing task-specific optimization fuel structure, however, call fuel-design methodology based on quantitative fuel-structure/property relationships. requires adjustments devices processes ensure clean, safe, efficient, fuel-flexible combustion, have accomplished relatively quickly. Computational methods vital element modern design particular importance when rapid developments required complex objectives pursued. Yet, highly non-linear nature complexities associated resulting difficulties development predictive models, this new methods. Recently, machine-learning-based been embraced pillar modeling, especially situations physics-based approaches reached maturity, but still limited accuracy applicability. Some interesting examples machine-learning model discussed.

Язык: Английский

---

IMO 2023 strategy-Where are we and what’s next?

Marine Policy, Год журнала: 2023, Номер 160, С. 105953 - 105953

Опубликована: Дек. 1, 2023

Язык: Английский

---

Enhancement in hydrogen sorption behaviour of MgH2 catalyzed by graphene quantum dots

International Journal of Hydrogen Energy, Год журнала: 2024, Номер 67, С. 1026 - 1032

Опубликована: Фев. 20, 2024

Язык: Английский

---

Multi-decadal trends of low-clouds at the Tropical Montane Cloud Forests

Ecological Indicators, Год журнала: 2024, Номер 158, С. 111599 - 111599

Опубликована: Янв. 1, 2024

Clouds are critical to the biodiversity and function of Tropical Montane Cloud Forests (TMCF) as they control water regimes sunlight that can be perceived by plants. These ecosystems provide a key role in ecosystem services humanity considered hotspots endemism, given number species is restricted their microclimates. The cloudiness these projected decline owing global warming, but recent temporal trends remain unclear. Here, we evaluated low-cloud fractions (CF) (e.g., proportion an area covered low-cloud) other Essential Climatic Variables (ECV) surface temperature, pressure, soil moisture, precipitation) for 521 sites worldwide with TMFCs from 1997 2020. We hypothesize traces warming over last few decades have led decreases CF on TMCFs. previous was also assessed globally among biogeographic realms identify regional trends. calculated aggregating hourly observations ERA5 reanalysis CHIRPS into annual averages then using linear regressions calculate slopes (i.e., rate change) (Δ, year−1). Our results suggest at TMCFs range between −64.7×10−4 51.4×10−4 year−1, revealing 70 % experienced reductions CF. Declines low-clouds 253 more severe than tropical landmasses when peak values density distribution compared (TMCFs: −7.8×10−4 year−1; −2.3×10−4 Despite this, differ realms, those Neotropics Indomalayan most pronounced declines. Decreases were associated increases temperature pressure TMCF's climate changing warmer environments. climatic shifts may represent imprints change TMCFs, highlighting current threat essential provide.

Язык: Английский

---

Antarctica in 2025: Drivers of deep uncertainty in projected ice loss

Science, Год журнала: 2025, Номер 387(6734), С. 601 - 609

Опубликована: Фев. 6, 2025

Antarctica is a vital component of Earth’s climate system, influencing global sea level, ocean circulation, and planetary albedo. Major knowledge gaps in critical processes—spanning the atmosphere, ocean, ice sheets, underlying beds, shelves, ice—create uncertainties future projections, hindering adaptation risk assessments intervention strategies. Antarctica’s sheet could contribute 28 centimeters to level by 2100, potentially more if we surpass warming thresholds that trigger instabilities rapid retreat. We review recent advances understanding changing stability margins identify key processes require further research. Progress requires high-resolution satellite data, targeted field campaigns, improved modeling, refined theory. Increased investment interdisciplinary collaboration are essential uncovering hidden reducing projections.

Язык: Английский

---

Global warming and obesity: External heat exposure as a modulator of energy balance

FASEB BioAdvances, Год журнала: 2025, Номер unknown

Опубликована: Март 18, 2025

Abstract In obesity research, the importance of core body temperature ( CBT ) regulation is often neglected. thermogenic regulation, however, plays a crucial role in heat management through convection, radiation, and conduction processes to remove from body, as well metabolic that sequester lipogenesis. This review emphasizes even small changes can significantly impact events ranging ATP production fat deposition. Accordingly, case made physical events, such external exposure, also compositional changes, do work processes. Examples are provided suggest independent diet exercise, where one lives, have an on composition obesity. For example, below 35 degrees earth's latitude, rates 40 percent or greater among adults. However, regions between 45 50 US ‐Canadian border, 25%–30%.

Язык: Английский

---

Machine learning–based extreme event attribution

Science Advances, Год журнала: 2024, Номер 10(34)

Опубликована: Авг. 21, 2024

The observed increase in extreme weather has prompted recent methodological advances event attribution. We propose a machine learning–based approach that uses convolutional neural networks to create dynamically consistent counterfactual versions of historical events under different levels global mean temperature (GMT). apply this technique one heat (southcentral North America 2023) and several have been previously analyzed using established attribution methods. estimate temperatures during the southcentral were 1.18° 1.42°C warmer because warming similar will occur 0.14 0.60 times per year at 2.0°C above preindustrial GMT. Additionally, we find learned relationships between daily GMT are influenced by seasonality forced response meteorological conditions. Our results broadly agree with other techniques, suggesting learning can be used perform rapid, low-cost events.

Язык: Английский

---

Multi-year analysis of physical interactions between solar PV arrays and underlying soil-plant complex in vegetated utility-scale systems

Applied Energy, Год журнала: 2024, Номер 365, С. 123227 - 123227

Опубликована: Апрель 25, 2024

Язык: Английский

---