Search for: All records

Abstract. Reducing methane emissions from the oil and gas (oil–gas) sector has been identified as a critically important global strategy for reducing near-term climate warming. Recent measurements, especially by satellite and aerial remote sensing, underscore the importance of targeting the small number of facilities emitting methane at high rates (i.e., “super-emitters”) for measurement and mitigation. However, the contributions from individual oil–gas facilities emitting at low emission rates that are often undetected are poorly understood, especially in the context of total national- and regional-level estimates. In this work, we compile empirical measurements gathered using methods with low limits of detection to develop facility-level estimates of total methane emissions from the continental United States (CONUS) midstream and upstream oil–gas sector for 2021. We find that of the total 14.6 (12.7–16.8) Tg yr−1 oil–gas methane emissions in the CONUS for the year 2021, 70 % (95 % confidence intervals: 61 %–81 %) originate from facilities emitting <100kgh-1 and 30 % (26 %–34 %) and ∼80 % (68 %–90 %) originate from facilities emitting <10 and <200kgh-1, respectively. While there is variability among the emission distribution curves for different oil–gas production basins, facilities with low emissions are consistently found to account for the majority of total basin emissions (i.e., range of 60 %–86 % of total basin emissions from facilities emitting <100kgh-1). We estimate that production well sites were responsible for 70 % of regional oil–gas methane emissions, from which we find that the well sites that accounted for only 10 % of national oil and gas production in 2021 disproportionately accounted for 67 %–90 % of the total well site emissions. Our results are also in broad agreement with data obtained from several independent aerial remote sensing campaigns (e.g., MethaneAIR, Bridger Gas Mapping LiDAR, AVIRIS-NG (Airborne Visible/Infrared Imaging System – Next Generation), and Global Airborne Observatory) across five to eight major oil–gas basins. Our findings highlight the importance of accounting for the significant contribution of small emission sources to total oil–gas methane emissions. While reducing emissions from high-emitting facilities is important, it is not sufficient for the overall mitigation of methane emissions from the oil and gas sector which according to this study is dominated by small emission sources across the US. Tracking changes in emissions over time and designing effective mitigation policies should consider the large contribution of small methane sources to total emissions. 

Abstract Sensors are indispensable tools of modern life that are ubiquitously used in diverse settings ranging from smartphones and autonomous vehicles to the healthcare industry and space technology. By interfacing multiple sensors that collectively interact with the signal to be measured, one can go beyond the signal-to-noise ratios (SNR) attainable by the individual constituting elements. Such techniques have also been implemented in the quantum regime, where a linear increase in the SNR has been achieved via using entangled states. Along similar lines, coupled non-Hermitian systems have provided yet additional degrees of freedom to obtain better sensors via higher-order exceptional points. Quite recently, a new class of non-Hermitian systems, known as non-Hermitian topological sensors (NTOS) has been theoretically proposed. Remarkably, the synergistic interplay between non-Hermiticity and topology is expected to bestow such sensors with an enhanced sensitivity that grows exponentially with the size of the sensor network. Here, we experimentally demonstrate NTOS using a network of photonic time-multiplexed resonators in the synthetic dimension represented by optical pulses. By judiciously programming the delay lines in such a network, we realize the archetypal Hatano-Nelson model for our non-Hermitian topological sensing scheme. Our experimentally measured sensitivities for different lattice sizes confirm the characteristic exponential enhancement of NTOS. We show that this peculiar response arises due to the combined synergy between non-Hermiticity and topology, something that is absent in Hermitian topological lattices. Our demonstration of NTOS paves the way for realizing sensors with unprecedented sensitivities. 

Abstract Rapid advancements in deep learning over the past decade have fueled an insatiable demand for efficient and scalable hardware. Photonics offers a promising solution by leveraging the unique properties of light. However, conventional neural network architectures, which typically require dense programmable connections, pose several practical challenges for photonic realizations. To overcome these limitations, we propose and experimentally demonstrate Photonic Neural Cellular Automata (PNCA) for photonic deep learning with sparse connectivity. PNCA harnesses the speed and interconnectivity of photonics, as well as the self-organizing nature of cellular automata through local interactions to achieve robust, reliable, and efficient processing. We utilize linear light interference and parametric nonlinear optics for all-optical computations in a time-multiplexed photonic network to experimentally perform self-organized image classification. We demonstrate binary (two-class) classification of images using as few as 3 programmable photonic parameters, achieving high experimental accuracy with the ability to also recognize out-of-distribution data. The proposed PNCA approach can be adapted to a wide range of existing photonic hardware and provides a compelling alternative to conventional photonic neural networks by maximizing the advantages of light-based computing whilst mitigating their practical challenges. Our results showcase the potential of PNCA in advancing photonic deep learning and highlights a path for next-generation photonic computers. 

Abstract. Accurate and comprehensive quantification of oil and gas methane emissions is pivotal in informing effective methane mitigation policies while also supporting the assessment and tracking of progress towards emissions reduction targets set by governments and industry. While national bottom-up source-level inventories are useful for understanding the sources of methane emissions, they are often unrepresentative across spatial scales, and their reliance on generic emission factors produces underestimations when compared with measurement-based inventories. Here, we compile and analyze previously reported ground-based facility-level methane emissions measurements (n=1540) in the major US oil- and gas-producing basins and develop representative methane emission profiles for key facility categories in the US oil and gas supply chain, including well sites, natural-gas compressor stations, processing plants, crude-oil refineries, and pipelines. We then integrate these emissions data with comprehensive spatial data on national oil and gas activity to estimate each facility's mean total methane emissions and uncertainties for the year 2021, from which we develop a mean estimate of annual national methane emissions resolved at 0.1° × 0.1° spatial scales (∼ 10 km × 10 km). From this measurement-based methane emissions inventory (EI-ME), we estimate total US national oil and gas methane emissions of approximately 16 Tg (95 % confidence interval of 14–18 Tg) in 2021, which is ∼ 2 times greater than the EPA Greenhouse Gas Inventory. Our estimate represents a mean gas-production-normalized methane loss rate of 2.6 %, consistent with recent satellite-based estimates. We find significant variability in both the magnitude and spatial distribution of basin-level methane emissions, ranging from production-normalized methane loss rates of < 1 % in the gas-dominant Appalachian and Haynesville regions to > 3 %–6 % in oil-dominant basins, including the Permian, Bakken, and the Uinta. Additionally, we present and compare novel comprehensive wide-area airborne remote-sensing data and results for total area methane emissions and the relative contributions of diffuse and concentrated methane point sources as quantified using MethaneAIR in 2021. The MethaneAIR assessment showed reasonable agreement with independent regional methane quantification results in sub-regions of the Permian and Uinta basins and indicated that diffuse area sources accounted for the majority of the total oil and gas emissions in these two regions. Our assessment offers key insights into plausible underlying drivers of basin-to-basin variabilities in oil and gas methane emissions, emphasizing the importance of integrating measurement-based data when developing high-resolution spatially explicit methane inventories in support of accurate methane assessment, attribution, and mitigation. The high-resolution spatially explicit EI-ME inventory is publicly available at https://doi.org/10.5281/zenodo.10734299 (Omara, 2024). 

Abstract Photonics offers unique capabilities for quantum information processing (QIP) such as room-temperature operation, the scalability of nanophotonics, and access to ultrabroad bandwidths and consequently ultrafast operation. Ultrashort pulse sources of quantum states in nanophotonics are an important building block for achieving scalable ultrafast QIP; however, their demonstrations so far have been sparse. Here, we demonstrate a femtosecond biphoton source in dispersion-engineered periodically poled lithium niobate nanophotonics. We measure 17 THz of bandwidth for the source centered at 2.09 µm, corresponding to a few optical cycles, with a brightness of 8.8 GHz/mW. Our results open new paths toward realization of ultrafast nanophotonic QIP. 

We introduce and experimentally demonstrate a FROG-based ultrashort pulse characterization technique using nanophotonic parametric amplification as a crucial tool for ultrafast nanophotonic circuits, and measure sub-50-femtosecond pulses. 

We experimentally demonstrate an end-to-end all-optical recurrent neural net-work utilizing short laser pulses and ultrafastχ⁽²⁾nonlinear optics to unlock unprecedented computational clock rates and massively parallel information processing for time-series classification and prediction tasks. 

We show that soliton pulse compression in lithium niobate nanophotonics can enable formation of few-cycle pulses. We experimentally confirm such nonlinear dynamics and measure chirped 44-fs output pulses consistent with numerical simulations. 

We experimentally demonstrate a recurrent optical neural network based on a nanophotonic optical parametric oscillator fabricated on thin-film lithium niobate. Our demonstration paves the way for realizing optical neural networks exhibiting ultra-low la-tencies.