Atmospheric models used for weather and climate prediction are traditionally formulated in a deterministic manner. In other words, given a particular state of the resolved scale variables, the most likely forcing from the subgrid scale processes is estimated and used to predict the evolution of the large-scale flow. However, the lack of scale separation in the atmosphere means that this approach is a large source of error in forecasts. Over recent years, an alternative paradigm has developed: the use of stochastic techniques to characterize uncertainty in small-scale processes. These techniques are now widely used across weather, subseasonal, seasonal, and climate timescales. In parallel, recent years have also seen significant progress in replacing parametrization schemes using machine learning (ML). This has the potential to both speed up and improve our numerical models. However, the focus to date has largely been on deterministic approaches. In this position paper, we bring together these two key developments and discuss the potential for data-driven approaches for stochastic parametrization. We highlight early studies in this area and draw attention to the novel challenges that remain.

Stochastic generators are useful for estimating climate impacts on various sectors. Projecting climate risk in various sectors, e.g. energy systems, requires generators that are accurate (statistical resemblance to ground-truth), reliable (do not produce erroneous examples), and efficient. Leveraging data from the North American Land Data Assimilation System, we introduce TemperatureGAN, a Generative Adversarial Network conditioned on months, regions, and time periods, to generate 2 m above ground atmospheric temperatures at an hourly resolution. We propose evaluation methods and metrics to measure the quality of generated samples. We show that TemperatureGAN produces high-fidelity examples with good spatial representation and temporal dynamics consistent with known diurnal cycles.

The new mineral manganonewberyite (IMA2024–004), Mn(PO3OH)(H2O)3, was found underground at the Cassagna mine, Liguria, Italy, where it is a secondary phase formed by the interaction of bat guano with Mn-rich rock. Manganonewberyite occurs with niahite, kutnohorite, sampleite and serrabrancaite on a tinzenite–quartz–braunite matrix. Crystals are prisms and blades, up to ∼0.15 mm long, elongated parallel to [001], flattened on {100} and exhibiting the forms {100}, {010} and {111}. Crystals are colourless and transparent, with vitreous lustre and white streak. The mineral is brittle with curved fracture. The Mohs hardness is ∼3. Cleavage is perfect on {010}. The density is 2.34(2) g·cm–3. Optically, manganonewberyite is biaxial (+) with α = 1.541(2), β = 1.547(2) and γ = 1.559(2) (white light). The 2V is 71.6(3)°. The optical orientation is X = a, Y = b and Z = c. The empirical formula is (Mn0.960Mg0.016Ca0.015)Σ0.991(H1.02P1.00O4)(H2O)3. Manganonewberyite is orthorhombic, space group Pbca, with cell parameters: a = 10.4273(6), b = 10.8755(8), c = 10.2126(4) Å, V = 1158.13(11) Å3 and Z = 8. The crystal structure (R1 = 2.79% for 892 I > 2σI reflections) is the same as that of newberyite with Mn in place of Mg.

Airborne radar sensors capture the profile of snow layers present on top of an ice sheet. Accurate tracking of these layers is essential to calculate their thicknesses, which are required to investigate the contribution of polar ice cap melt to sea-level rise. However, automatically processing the radar echograms to detect the underlying snow layers is a challenging problem. In our work, we develop wavelet-based multi-scale deep learning architectures for these radar echograms to improve snow layer detection. These architectures estimate the layer depths with a mean absolute error of 3.31 pixels and 94.3% average precision, achieving higher generalizability as compared to state-of-the-art snow layer detection networks. These depth estimates also agree well with physically drilled stake measurements. Such robust architectures can be used on echograms from future missions to efficiently trace snow layers, estimate their individual thicknesses, and thus support sea-level rise projection models.

Climate models are biased with respect to real-world observations. They usually need to be adjusted before being used in impact studies. The suite of statistical methods that enable such adjustments is called bias correction (BC). However, BC methods currently struggle to adjust temporal biases. Because they mostly disregard the dependence between consecutive time points. As a result, climate statistics with long-range temporal properties, such as the number of heatwaves and their frequency, cannot be corrected accurately. This makes it more difficult to produce reliable impact studies on such climate statistics. This article offers a novel BC methodology to correct temporal biases. This is made possible by rethinking the philosophy behind BC. We will introduce BC as a time-indexed regression task with stochastic outputs. Rethinking BC enables us to adapt state-of-the-art machine learning (ML) attention models and thereby learn different types of biases, including temporal asynchronicities. With a case study of number of heatwaves in Abuja, Nigeria and Tokyo, Japan, we show more accurate results than current climate model outputs and alternative BC methods.

Bias correction is a critical aspect of data-centric climate studies, as it aims to improve the consistency between observational data and simulations by climate models or estimates by remote sensing. Satellite-based estimates of climatic variables like precipitation often exhibit systematic bias when compared to ground observations. To address this issue, the application of bias correction techniques becomes necessary. This research work examines the use of deep learning to reduce the systematic bias of satellite estimations at each grid location while maintaining the spatial dependency across grid points. More specifically, we try to calibrate daily precipitation values of tropical rainfall measuring mission based TRMM_3B42_Daily precipitation data over Indian landmass with ground observations recorded by India Meteorological Department (IMD). We have focused on the precipitation estimates of the Indian Summer Monsoon Rainfall (ISMR) period (June–September) since India gets more than 75% of its annual rainfall in this period. We have benchmarked these deep learning methods against standard statistical methods like quantile mapping and quantile delta mapping on the above datasets. The comparative analysis shows the effectiveness of the deep learning architecture in bias correction.

A quadrotor unmanned aerial vehicle (UAV) must achieve desired flight missions despite internal uncertainties and external disturbances. This paper proposes an adaptive trajectory tracking control method that attenuates unknown uncertainties and disturbances. Although the quadrotor is underactuated, a fully actuated controller is designed using backstepping control. To avoid repeated derivatives of control inputs, a dynamic surface method introduces a filter and auxiliary controller. Lyapunov criteria guide adaptive laws for tuning controller gain and filters. A low-power observer is integrated for state estimation. Additionally, a disturbance observer is developed and combined with the control scheme to handle unknown disturbances. Simulations on a DJI F450 quadrotor demonstrate that the proposed control algorithm offers strong trajectory-tracking performance and system stability under multiple uncertainties and external disturbances during flight.

The transport industry of Ukraine is an integral part of its economy. According to the National Transport Strategy of Ukraine, a critical strategic goal is to enhance transport safety. Currently, there is a gap in mobile devices capable of automatically measuring slopes and evenness of both runways and road surfaces in two coordinates. This paper addresses the creation of new methods for assessing longitudinal and transverse slopes using micromechanical systems. The study highlights international experiences, presents practical applications and proposes strategies for overcoming implementation challenges. A detailed roadmap for deployment and further improvements is provided.

This paper analyses the performance of the Australian and New Zealand Satellite-Based Augmentation System (Aus-NZ SBAS) test-bed to evaluate its use in civil aviation applications with a focus on dual-frequency multi-constellation (DFMC) signals. The Aus-NZ SBAS test-bed performance metrics were determined using kinematic data recorded in flight across a variety of environments and operational conditions. A total of 14 tests adding up to 32 h of flight were evaluated. Flight test data were processed in both the L1 SBAS and DFMC SBAS modes supported by the test-bed broadcasts. The performance results are reviewed regarding accuracy, availability and integrity metrics and compared with the requirement thresholds defined by the International Civil Aviation Organisation (ICAO) for Precision Approach (PA) flight operations. The experimentation performed does not allow continuity assessment as specified in the standard due to a long-term statistical requirement and inherent limitations imposed by the reference station network. Analysis of flight test results shows that DFMC SBAS provides several performance improvements over single-frequency SBAS, tightening both horizontal and vertical protection levels and resulting in greater service availability during the approach.

This research employs an enhanced Polar Operation Limit Assessment Risk Indexing System (POLARIS) and multi-scale empirical analysis methods to quantitatively evaluate the risks in icy region navigation. It emphasises the significant influence of spatial effects and external environmental factors on maritime accidents. Findings reveal that geographical location, environmental and ice conditions are crucial contributors to accidents. The models indicate that an increase in ports, traffic volume and sea ice density directly correlates with higher accident rates. Additionally, a novel risk estimation model is introduced, offering a more accurate and conservative assessment than current standards. This research enriches the understanding of maritime accidents in icy regions, and provides a robust framework for different navigation stages and conditions. The proposed strategies and model can effectively assist shipping companies in route planning and risk management to enhance maritime safety in icy regions.

With increased global navigation satellite system (GNSS) signals and degraded observation environments, the correctness of ambiguity resolution is disturbed, causing unexpected real-time kinematic (RTK) positioning solutions. This paper presents an improved fault detection and exclusion (FDE) method based on the generalized least squares (GLS) model. The correlated GLS model is constructed by regarding double-differencing (DD) integer ambiguities as the known parameters. Meanwhile, the validity of residuals as crucial components of fault detection could be enhanced by the iterative re-weighted least squares (IRLS) method rather than the least squares (LS) without robustness. A static test with artificial faults and a dynamic test with natural faults were carried out, respectively. By analyzing test statistics of the enhanced FDE algorithm and comparing its positioning errors with those from the classical LS, it is shown that our method can provide high-precision and high-reliability RTK solutions facing wrong DD fixed ambiguities due to observation faults.

Egg masses of Aplysia depilans consist of long and intertwined strings containing numerous capsules with eggs. Light microscopy stains and transmission electron microscopy revealed four layers in the gelatinous sheath that encircled and aggregated the chain of egg capsules. The outermost layer has a fluffy structure. The second, third, and fourth layers consisted of reticulated matrices with different densities. The second and third layers were divided into 5‒6 strata each. The fourth and innermost layer of the gelatinous sheath has a higher density and no visible stratification. This layer glues the tightly packed capsules to one another and to the outer layers of the gelatinous sheath. The thin wall of the capsules is formed by a homogeneous and highly electron-dense material. Inside the capsules, the eggs or embryos were bathed in an electron-lucent aqueous medium. Bacteria and diatoms were the most abundant microorganisms on the surface of egg strings. Bacteria penetrate the gelatinous sheath and appear to be involved in the degradation of the upper strata, but were never found inside the egg capsules. Metagenomic analysis revealed a large taxonomic diversity of bacteria associated with egg masses of A. depilans. Although 15 phyla could be recognized, the families Flavobacteriaceae (Bacteroidota), Lentisphaeraceae (Lentisphaerota), and Rhodobacteraceae (Pseudomonadota) represented 67.9% ± 11.6% of the relative abundance in the microbiome of the egg string samples. The presence of genera capable of decomposing polysaccharides, such as Tenacibaculum and Cellulophaga, supports the idea that bacteria are responsible for the degradation of the gelatinous layers of the egg strings.

Vessel collision risk estimation is crucial in navigation manoeuvres, route planning, risk control, safety management and forewarning issues. The interaction possibility is a good method to quantify the near-miss collision risks of multi-ships. Current models, however, are mostly concerned about the movements in an unrestricted isotropic travel environment or network environment. This article simultaneously addresses these issues by developing a novel environment–kinetic compound space–time prism to capture potential spatial–temporal interactions of multi-ships in constrained dynamic environments. The approach could significantly reduce the overestimation of the individual vessel’s potential travel area and the interaction possibility of encountering vessels in restricted water. The proposed environmental–kinetical compound space–time prism (EKC-STP)-based method enables identifying where and when multi-ships possibly interacted in the constraint water area, as well as how the interaction possibility pattern changed from day to day. The collision risk evaluation results were validated through comparison with other methods. The full picture of hierarchical collision risk distribution in port areas is determined and could be employed to provide quantifiable references for efficient and practical anti-collision measures establishment.

The Plane or Plain Scale is a navigational device that dates back to the early 1600s but has long since ceased to be used in practice. It could perform the function of a protractor and be used to solve problems in plane trigonometry. In addition, coupled with a suite of remarkable geometric constructions based on stereographic projection, the Plane Scale could efficiently solve problems in spherical trigonometry and hence navigation on a sphere. The methods used seem today to be largely unknown. This paper describes the Plane Scale and how it was used.

The optimisation of inter-island transportation systems constitutes a critical determinant of regional economic development and the efficacy of mobility infrastructure. This study presents a comparative analysis of passenger mode selection between short-sea shipping (SSS) and road transport alternatives through stated preference surveys conducted via anonymised questionnaires. Employing advanced discrete choice modelling techniques – specifically the multinomial logit (MNL), random parameter logit (RPL) and latent class (LC) frameworks – we quantitatively disentangle the complex determinants influencing modal preferences. Our systematic sensitivity analysis reveals distinct behavioural patterns: passengers opting for SSS prioritise journey convenience, whereas road transport users exhibit stronger cost sensitivity. These findings provide actionable insights for formulating evidence-based policies to enhance intermodal transportation networks in the Zhoushan Archipelago of China. Beyond its immediate geographical focus, this research contributes methodological innovations by applying finite mixture models to capture unobserved heterogeneity in maritime transport decisions. The framework demonstrates significant transferability potential for island territories globally and urban freight corridor optimisation challenges, particularly in contexts requiring trade-off analyses between maritime efficiency and terrestrial logistics constraints.

The propagation of elastic-flexural–gravity waves through an ice shelf is modeled using full three-dimensional elastic models that are coupled with a treatment of under-shelf sea-water flux: (i) finite-difference model (Model 1), (ii) finite-volume model (Model 2) and (iii) depth-integrated finite-difference model (Model 3). The sea-water flow under the ice shelf is described by a wave equation involving the pressure (the sea-water flow is treated as a “potential flow”). Numerical experiments were undertaken for an ice shelf with ‘rolling’ surface morphology, which implies a periodic structure of the ice shelf. The propagation of ocean waves through an ice shelf with rolling surface morphology is accompanied by Bragg scattering (also called Floquet band insulation). The numerical experiments reveal that band gaps resulting from this scattering occur in the dispersion spectra in frequency bands that are consistent with the Bragg’s law. Band gaps render the medium opaque to wave, that is, essentially, the abatement of the incident ocean wave by ice shelf with rolling surface morphology is observed in the models. This abatement explains the ability of preserving of ice shelves like the Ward Hunt Ice Shelf, Ellesmere Island, Canadian Arctic, from the possible resonant-like destroying impact of ocean swell.

The cause of megafauna extinction in South America remains clouded in controversy, since it took place at a time of intense climate change and almost at the same time as the initial human influx into the continent. In this paper, we aimed to assess the effects of climate change on open vegetation habitats and, consequently, on megafauna extinction in South America by using a species distribution model, fossil records, and paleoclimatic projections. In addition, we evaluated the effects of climatic variables on the distribution of suitable habitats across South America. Our results demonstrated alternating intervals of expansion and contraction of suitable areas for megafauna persistence, mainly in response to lower and higher precipitation, in the last 21 ka in all regions of South America. However, the amplitude of this oscillation was more significant in the Brazilian Northeast. In the Andean and Chaco–Pampas regions, greater precipitation stability resulted in greater stability in habitat suitability; therefore, for these regions, other factors must have predominated for the extinction of the megafauna. We therefore concluded that in the Andean and Chaco–Pampas regions, climate change was not solely responsible for the disappearance of megafauna, but in the Brazilian Northeast, it may have been decisive.