Volcanic meteo-tsunamis, though rare, can pose significant threats to people, as exemplified by the 2022 Hunga Tonga – Hunga Ha’apai (HT-HH) eruption in the SW Pacific. While various studies have delved into the complexities of such phenomena, none have explored analogous scenarios in regions with potential occurrence of large eruptions near or under the sea. We focus on coastal areas along the South China Sea (SCS), among the most densely populated on Earth and historically prone to volcanic activity, including the catastrophic 1883 Krakatau eruption. Here we strategically chose one intra-basin volcano, KW-23612 in the northern SCS, and three extra-basin volcanoes, Banua Wuhu in the Celebes Sea, and Kikai and Fukutoku-Oka-no-Ba in the northern Philippines Sea (southern Japan), from which we simulated volcanic meteo-tsunamis with scaled intensities of the HT-HH event, to assess which countries around the SCS could be more at risk from the occurrence of such phenomena. Our results show that the worst-case scenarios are produced by eruption/tsunamis from the northern SCS, producing offshore waves up to 10 cm offshore Macau and Hong Kong, and up to 20 cm offshore Manila. In contrast, countries bordering the shallow Sunda Shelf (Malaysia, Thailand, Cambodia, and southern Vietnam) seem less at risk from volcanic meteo-tsunamis, though we observed some amplification effects along the deeper Singapore Strait. This study is the first of its kind in the region and sets the basis to investigate amplification effects, and shallow coastal dynamics at key locations, after integrating higher resolution bathymetry data.

A document by Shreyaa Sanjay. Click on the document to view its contents.

Evaporation decreases the mass and increases the isotope composition of falling drops. Combining and integrating the dependence of the evaporation on the drop diameter and on the drop-environment humidity difference, the square of drop diameter is found to decrease with the square of vertical distance below cloud base. Drops smaller than 0.5 mm evaporate completely before falling 700 m in typical subtropical marine boundary layer conditions. The effect on the isotope ratio of equilibration with the environment, evaporation, and kinetic molecular diffusion is modeled by molecular and eddy diffusive fluxes after Craig and Gordon (1965), with a size-dependent parameterization of diffusion that enriches small drops more strongly, and approaches the rough aerodynamic limit for large drops. Rain shortly approaches a steady state with the subcloud vapor by exchange with a length scale of 40 m. Kinetic molecular diffusion enriches drops up to as they evaporate by up to +5~\permil~for deuterated water (HDO) and +3.5~\permil~for H$_2$$^{18}$O. Rain evaporation enriches undiluted subcloud vapor by +12~\permil~per mm rain, explaining enrichment of vapor in evaporatively cooled downdrafts that contribute to cold pools. Microphysics enriches the vapor lost by the early and complete evaporation of smaller drops in the distribution. Vapor from hydrometeors is more enriched than it would be by Rayleigh distillation or by mixtures of liquid rain and vapor in equilibrium with rain.

The occurrence of tropical cyclones (TC) in the Horn of Africa and nearby areas is for the first time examined to document their contribution to local rainfall and their trends over the period 1990-2020. An average 1.5 TC (of any intensity) per year was observed over the Western Arabian Sea, with two asymmetrical seasons, namely May-June (30% of cyclonic days) and September-December (70%). Case studies reveal that in many instances, TC-related rainfall extends beyond 500 km from the TC center, and that substantial rains occur one to two days after the lifecycle of the TC. Despite their rarity, in the otherwise arid to semi-arid context characteristic of the region, TCs contribute in both seasons to a very high percentage of total rainfall (up to 30 to 60%) over the northwestern Arabian Sea, the Gulf of Aden and their coastlines. Over inland northern Somalia, contributions are much lower. TCs disproportionately contribute to some of the most intense daily falls, which are often higher than the mean annual rainfall. A strong increase in the number of TCs is found from 1990 to 2020, hence their enhanced contribution to local rainfall. This increase is associated with a warmer eastern / southern Arabian Sea, a decrease in vertical wind shear, and a strong increase in tropospheric moisture content.

The water vapor transport in the extratropics is mainly organized in narrow elongated filaments. These filaments are referred to with a variety of names depending on the contexts. When making landfall on a coastline, they are generally referred to as atmospheric rivers; when occurring at high latitudes, many authors regard them as warm moist intrusions; when occurring ahead of a cold front towards the core on an extratropical cyclone, the most commonly used term is warm conveyor belt. Here, we propose an algorithm that detects these various lines of moisture transport in instantaneous maps of the vertically integrated water vapor transport. The detection algorithm extracts well-defined maxima in the water vapor transport and connects them to lines that we refer to as moisture transport axes. By only requiring a well-defined maximum in the vapor transport, we avoid imposing a threshold in the absolute magnitude of this transport (or the total column water vapor). Consequently, the algorithm is able to pick up moisture transport axes at all latitudes without requiring region-specific tuning or normalization. We demonstrate that the algorithm can detect both atmospheric rivers and warm moist intrusions, but also prominent monsoon air streams. Atmospheric rivers sometimes consist of several distinct moisture transport axes, indicating the merging of several moisture filaments into one atmospheric river. We showcase the synoptic situations and precipitation patterns associated with the occurrence of the identified moisture transport axes in example regions in the low, mid, and high latitudes.

Existing weather models are known to have poor skill at forecasting rainfall over East Africa, where there are regular threats of drought and floods. Improved forecasts could reduce the effects of these extreme weather events and provide significant socioeconomic benefits to the region. We present a novel machine learning-based method to improve precipitation forecasts in East Africa, using postprocessing based on a conditional generative adversarial network (cGAN). This addresses the challenge of realistically representing tropical rainfall, where convection dominates and is poorly simulated in conventional global forecast models. We postprocess hourly forecasts made by the European Centre for Medium-Range Weather Forecasts Integrated Forecast System at 6-18h lead times, at $0.1^{\circ}$ resolution. We combine the cGAN predictions with a novel neighbourhood version of quantile mapping, to integrate the strengths of machine learning and conventional postprocessing. Our results indicate that the cGAN substantially improves the diurnal cycle of rainfall, and improves predictions up to the $99.9^{\text{th}}$ percentile ($\sim 10 \text{mm}/\text{hr}$). This improvement extends to the 2018 March–May season, which had extremely high rainfall, indicating that the approach has some ability to generalise to more extreme conditions. We explore the potential for the cGAN to produce probabilistic forecasts and find that the spread of this ensemble broadly reflects the predictability of the observations, but is also characterised by a mixture of under- and over-dispersion. Overall our results demonstrate how the strengths of machine learning and conventional postprocessing methods can be combined, and illuminate what benefits machine learning approaches can bring to this region.

Stratocumulus clouds, a key component of global climate, are sensitive to aerosol properties. Aerosol-cloud-precipitation interactions in these clouds influence their closed-to-open cell dynamical transition and hence cloud cover and radiative forcing. This study uses large-eddy simulations with Lagrangian super-particle and bin microphysics schemes to investigate impacts of aerosol scavenging and physical processing by clouds on drizzle initiation and the cellular transition process. The simulation using Lagrangian microphysics with explicit representation of cloud-borne aerosol and scavenging shows significant aerosol processing that impacts precipitation generation and consequently the closed-to-open cell transition. Sensitivity simulations using the bin scheme and their comparison with the Lagrangian microphysics simulation suggest that reduced aerosol concentration due to scavenging is a primary microphysical catalyst for enhanced precipitation using the Lagrangian scheme. However, changes in the aerosol distribution shape through processing also contribute appreciably to the differences in precipitation rate. Thus, both aerosol scavenging and processing drive earlier rain formation and the transition to open cells in the simulation with Lagrangian microphysics. This study also highlights a shortcoming of Eulerian bin microphysics producing smaller mean drop radius and cloud water mixing ratios owing to numerical diffusion. Initially larger mean radius and cloud mixing ratios using the Lagrangian scheme induce faster rain development compared to the bin scheme. A positive feedback in turn accelerates aerosol removal and further rain production using the Lagrangian scheme and, consequently, reduced cloud droplet number, increased mean size, and increased droplet spectral width.

This study analyzes hydrometeor evolution during rapid intensification (RI) and tropical cyclone (TC) intensity dependence using satellite data. Previous studies have suggested that ice cloud water or non-convective precipitation can serve as predictors of RI from different perspectives. However, few studies have focused on the impact of TC intensity or comprehensive comparisons to identify better indicators. During RI, the hydrometeor contents in weak TCs (WTCs) increase over the entire region, whereas they increase mainly in the inner-core region and decrease in advance in the outer-core region for strong TCs (STCs). Hydrometeor contents are higher in RI than in slow intensification, and their maximum locations are related to TC intensity and intensification rate. Cloud water content (CWC) in the inner-core region has the largest correlation with the intensification rate, especially in WTCs. Therefore, the CWC can serve as a predictor of RI and can be applied to all TC intensities.

Existing research has employed various methods to quantify urban heat island (UHI) effects, but the ideal method for individual cities remains unclear. This study investigated how different methods influence UHI understanding in Addis Ababa, a tropical city facing UHI challenges. Three methods were compared: dynamic urbanization, natural and built-up fractions, and urban center vs. surrounding rural areas. Satellite data and spatial analyses revealed maximum daytime UHIs of 4°C and 3.1°C in summer and autumn, respectively. Examining the mean temperature differences between urban and rural areas across methods yielded diverse results. This suggests that while the ‘dynamic urbanization’ method is statistically favorable in this specific case, averaging results from multiple methods produced a more robust and generalizable approach to understanding UHIs in different urban contexts. Ultimately, this study highlights the importance of context-specific method selection for accurately understanding the complex interplay between urban and rural environments.

We present a multidisciplinary study of the microphysics, mesoscale and synoptic conditions of a rare radiation-fog event in the central and southern regions of Israel during January 3-6, 2021. The fog developed during nighttime from south to central coastal areas and dissipated at morning. The synoptic conditions were dominated by Red Sea Troughs at the surface without cyclonic upper air circulation, suitable for radiation fog development. In-situ measurements were combined with satellite imagery, high resolution (1-km grid size) Weather Research and Forecast model (WRF) with Real-Time Four-Dimensional Data Assimilation (RTFDDA) forecasts and post-processing algorithms including machine learning (ML) to analyze this event and to evaluate its numerical forecasting. The micro-physical analysis involved measurements of droplet size distribution and visibility range, allowing the calculation of liquid-water content and effective diameter of fog droplets. The measured visibility range was 90 m. The droplet diameter main mode was 1-2 micrometers, followed by another one around 6 micrometers. Typical liquid-water content values were 0.01-0.025 g/m3. WRF-RTFDDA mesoscale forecasts, post-processed by simple thresholds-based and ML algorithms, largely succeeded in predicting the temporal and spatial development of the dense fog. They proved useful in distinguishing between near-surface fog and elevated fog/low clouds, a distinction not possible from satellite imagery only. Clear patches at coastal areas covered in part by urban landuse were observed both in satellite imagery and model forecasts. WRF-RTFDDA forecasts proved their usefulness in forecasting this massive fog and low clouds events and in providing alerts to operational users and field campaign deployments.

The Indian summer monsoon precipitation trend from 1979 to 2022 shows a substantial 40% increase over Northwest India, which is in agreement with the future projections of the Coupled Model Intercomparison Project 6 (CMIP6). The observationally constrained reanalysis dataset reveals that a prominent sea surface warming in the western equatorial Indian Ocean and the Arabian Sea might be responsible for the rainfall enhancement through strengthening the cross-equatorial monsoonal flow and associated evaporation. We show that the cross-equatorial monsoon winds over the Indian Ocean are strengthening due to the merging of Pacific Ocean trade winds and rapid Indian Ocean warming. These winds also enhance the latent heat flux (evaporation), and in combination, this results in increased moisture transport from the ocean toward the land.

Temperature and water vapor are known to fluctuate on multiple scales. In this study 27 years of airborne measurements of temperature and relative humidity from IAGOS (In-service Aircraft for a Global Observing System) are used to parameterize the distribution of water vapor in the upper troposphere and lower stratosphere (UTLS). The parameterization is designed to simulate water vapor fluctuations within gridboxes of atmospheric general circulation models (AGCMs) with typical size of a few tens to a few hundreds kilometers. The distributions currently used in such models are often not supported by observations at high altitude. More sophisticated distributions are key to represent ice supersaturation, a physical phenomenon that plays a major role in the formation of natural cirrus and contrail cirrus. Here the observed distributions are fitted with a beta law whose parameters are adjusted from the gridbox mean variables. More specifically the standard deviation and skewness of the distributions are expressed as empirical functions of the average temperature and specific humidity, two typical prognostic variables of AGCMs. Thus, the distribution of water vapor is fully parameterized for a use in these models. The new parameterization simulates the observed distributions with a determination coefficient always greater than 0.917, with a mean value of 0.997. Moreover, the ice supersaturation fraction in a model gridbox is well simulated with a determination coefficient of 0.983. The parameterization is robust to a selection of various geographical subsets of data and to gridbox sizes varying between 25 to 300 km.

Understanding the physical mechanisms which contribute towards the rapid intensification of flash droughts is crucial for improving their forecasts. These mechanisms are difficult to elucidate using statistical techniques due to the complex interactions between land surface and atmospheric processes. In order to overcome this limitation, we use a slab model to model the coupled energy and water balance of the land and atmosphere. We develop an analytical framework to disentangle the influence of external forcings and system response driven by the state variables using the energy and water balance equations of the model. We apply the model to six locations selected from different climate regions of India to identify the physical mechanisms of flash droughts. We find that most flash droughts in India happen during the monsoon season, with higher frequency in humid regions of Northeast India and Southern Peninsular India. We find that all flash droughts occur during periods of deficient rainfall and the drying is predominantly driven by net shortwave radiation. However, the flash droughts differ in terms of contribution of winds towards drying, based on which we classify the flash drought mechanisms into three types: (a) flash droughts with wind-driven intensification due to land-atmospheric feedback (b) flash droughts with minimal contribution of winds towards drying and (c) flash droughts with wind-driven intensification due to advected heat. We also show that although the enhanced vapor pressure deficit is a frequently recurring feature of flash droughts, it is not necessarily the most relevant contributor in their development.

A document by Ehsan Erfani. Click on the document to view its contents.

A document by Ehsan Erfani. Click on the document to view its contents.

The atmospheric convection is a phenomenon that has a length scale smaller than the resolution at which the state-of-the-art general circulation numerical models are running, and thus, the convection needs to be parameterized. In this work, a theoretical formulation for the parameterization of subgrid-scale convection based on subgrid averaging that represents a generalized eddy-diffusivity mass-flux (GEM) formulation is presented. The subgrid fluxes are derived by considering a decomposition of subgrid variables into convective and turbulent variables, and by assuming that the convection is modeled by round convective plumes with generic radial profiles. The main difference between our formulation and the mass-flux formulations is that the condition of a very small fractional area occupied by the convection is replaced with the condition that the convective vertical velocity goes to the mean state far from the updraft region of the plume. The plume model can be also generalized by considering that the convective elements are energy-consistent plumes, governed by the conservation of mass, momentum, kinetic energy, and buoyancy, which provides a physical-based closure for the entrainment. Therefore, the closing problem of our formulation consists of the specification of the convective radial profiles and the boundary conditions at the initial vertical level. Furthermore, our formulation allows one to consider more realistic profiles for the convective variables, making the formulation suitable for the parameterization of atmospheric convection at the convective gray-zone. This aspect is discussed in the last part of this work, where it is showed how the formulation can be implemented at any resolution. Moreover, in the present framework, no distinct assumptions are required for each convective type, thus facilitating the parameterization of convection in a unified way.

Climate change is posing unprecedented challenges, necessitating the development of effective climate adaptation. Conventional computational models of climate adaptation frameworks inadequately account for our capacity to learn, update, and enhance decisions as exogenous information is collected. Here we investigate the potential of reinforcement learning (RL), a machine learning technique that exhibits efficacy in acquiring knowledge from the environment and systematically optimizing dynamic decisions, to model and inform adaptive climate decision-making. To illustrate, we derive adaptive stratigies for coastal flood protections for Manhattan, New York City, considering continuous observations of sea-level rise throughout the 21st century. We find that, when designing adaptive seawalls to protect Manhattan, the RL-derived strategy leads to a significant reduction in the expected cost, 6% to 36% under the moderate emissions scenario SSP2-4.5 (9% to 77% under the high emissions scenario SSP5-8.5), compared to previous methods. When considering multiple adaptive policies (buyout, accommodate, and dike), the RL approach leads to a further 5% (15%) reduction in cost, showcasing RL’s flexibility in addressing complex policy design problems when multiple policies interact. RL also outperforms conventional methods in controlling tail risk (i.e., low probability, high impacts) and avoiding losses induced by misinformation (e.g., biased sea-level projections), demonstrating the importance of systematic learning and updating in addressing extremes and uncertainties related to climate adaptation. The analysis also reveals that, given the large uncertainty and potential misjudgment about climate projection, “preparing for the worst” is economically more beneficial when adaptive strategies, such as those supported by the RL approach, are applied.

The Ocean Heat Content (OHC) anomaly has become an increasingly important climate parameter for the Intergovernmental Panel on Climate Change (IPCC) assessment and evaluation of climate change. One good reason is that the OHC appears to be less prone to climate variability, typically experienced with surface temperature and other climate parameters. Therefore, a reasonable estimate of OHC increase is important for research and climate related policies. Levitus et al. (2012) (https://doi.org/10.1029/2012GL051106) is a relevant ocean heat content related paper, and their analysis and estimate of OHC increase between 1955 and 2010 is high, about four to seven times greater than what the law of conservation of energy may allow. The source of discrepancy is analyzed in this commentary and it appears to be a result of using corrected ocean data sources. Therefore, verification of the observed increase in OHC using alternative ocean data sources is recommended.

It has been said, arguably, that causality analysis should pave a promising way to interpretable deep learning and generalization. Incorporation of causality into artificial intelligence (AI) algorithms, however, is challenged with its vagueness, non-quantitiveness, computational inefficiency, etc. During the past 18 years, these challenges have been essentially resolved, with the establishment of a rigorous formalism of causality analysis initially motivated from atmospheric predictability. This not only opens a new field in the atmosphere-ocean science, namely, information flow, but also has led to scientific discoveries in other disciplines, such as quantum mechanics, neuroscience, financial economics, etc., through various applications. This note provides a brief review of the decade-long effort, including a list of major theoretical results, a sketch of the causal deep learning framework, and some representative real-world applications in geoscience pertaining to this journal, such as those on the anthropogenic cause of global warming, the decadal prediction of El Niño Modoki, the forecasting of an extreme drought in China, among others.

The higher-order turbulence scheme, Cloud Layers Unified by Binormals (CLUBB), is known for effectively simulating the transition from cumulus to stratocumulus clouds within leading atmospheric climate models. This study investigates an underexplored aspect of CLUBB: its capacity to simulate near-surface winds and the Planetary Boundary Layer (PBL), with a particular focus on its coupling with surface momentum flux. Using the GFDL atmospheric climate model (AM4), we examine two distinct coupling strategies, distinguished by their handling of surface momentum flux during the CLUBB’s stability-driven substepping performed at each atmospheric time step. The static coupling maintains a constant surface momentum flux, while the dynamic coupling adjusts the surface momentum flux at each CLUBB substep based on the CLUBB-computed zonal and meridional wind speed tendencies. Our 30-year present-day climate simulations (1980-2010) show that static coupling overestimates 10-m wind speeds compared to both control AM4 simulations and reanalysis, particularly over the Southern Ocean (SO) and other midlatitude ocean regions. Conversely, dynamic coupling corrects the static coupling 10-m winds biases in the midlatitude regions, resulting in CLUBB simulations achieving there an excellent agreement with AM4 simulations. Furthermore, analysis of PBL vertical profiles over the SO reveals that dynamic coupling reduces downward momentum transport, consistent with the found wind-speed reductions. Instead, near the tropics, dynamic coupling results in minimal changes in near-surface wind speeds and associated turbulent momentum transport structure. Notably, the wind turning angle serves as a valuable qualitative metric for assessing the impact of changes in surface momentum flux representation on global circulation patterns.

During the northern spring (approximately Ls≈33°) in Martian Year 35, Mars experienced an unusual dust storm characterized by significantly increased dust in the northern troposphere. As observed by the Mars Climate Sounder (MCS), temperature significantly increases in the mid-latitude troposphere of both hemispheres and decreases in the northern mesosphere during the event. The temperature response simulated by the Martian General Circulation Model (GCM) agrees with the MCS observations. The radiative heating from dust is responsible for the increased temperature in the northern troposphere. In contrast, the dynamic heating/cooling contributes to the temperature variations in the southern troposphere and northern mesosphere. The increased dissipation of planetary waves enhances the residual meridional circulation and causes the temperature warming in the Southern Hemisphere. In addition, the enhanced meridional circulation related to this event leads to ~36% increase in water vapor transport from the Northern to the Southern Hemisphere as compared to the net interhemispheric transport over an entire Martian Year.

In this study, we explore the impact of oceanic moisture fluxes on atmospheric blocks using the ECMWF Integrated Forecast System. Artificially suppressing surface latent heat flux over the Gulf Stream region leads to a significant reduction (up to 30%) in atmospheric blocking frequency across the northern hemisphere. Affected blocks show a shorter lifespan (-6%), smaller spatial extent (-12%), and reduced intensity (-0.4%), with an increased detection rate (+17%). These findings are robust across various blocking detection thresholds. Analysis indicates a resolution-dependent response, with resolutions lower than Tco639 (~18km) showing no significant change in some blocking characteristics, even with reduced blocking frequency. Exploring the broader Rossby wave pattern, we observe that diminished moisture flux favours eastward propagation and higher zonal wavenumbers, while air-sea interactions promotes stationary and westward-propagating waves with zonal wavenumber 3. This study underscores the critical role of western boundary current’s moisture fluxes in modulating atmospheric blocking.

Subgrid-scale processes, such as atmospheric gravity waves, play a pivotal role in shaping the Earth’s climate but cannot be explicitly resolved in climate models due to limitations on resolution. Instead, subgrid-scale parameterizations are used to capture their effects. Recently, machine learning has emerged as a promising approach to learn parameterizations. In this study, we explore uncertainties associated with a machine learning parameterization for atmospheric gravity waves. Focusing on the uncertainties in the training process (parametric uncertainty), we use an ensemble of neural networks to emulate an existing gravity wave parameterization. We estimate both offline uncertainties in raw neural network output and online uncertainties in climate model output, after the neural networks are coupled. We find that online parametric uncertainty contributes a significant source of uncertainty in climate model output that must be considered when introducing neural network parameterizations. This uncertainty quantification provides valuable insights into the reliability and robustness of machine learning-based gravity wave parameterizations, thus advancing our understanding of their potential applications in climate modeling.

Improved urban greenhouse gas (GHG) flux estimates are crucial for informing policy and mitigation efforts. Atmospheric inversion modelling (AIM) is a widely used technique combining atmospheric measurements of trace gas, meteorological modelling, and a prior emission map to infer fluxes. Traditionally, AIM relies on mid-afternoon observations due to the well-represented atmospheric boundary layer in meteorological models. However, confining flux assessement to daytime observations is problematic for the urban scale, where air masses typically move over a city in a few hours and AIM therefore cannot provide improved constraints on emissions over the full diurnal cycle. We hypothesized that there are atmospheric conditions beyond the mid-afternoon under which meteorological models also perform well. We tested this hypothesis using tower-based measurements of CO2 and CH4, wind speed observations, weather model outputs from INFLUX (Indianapolis Flux Experiment), and a prior emissions map. By categorizing trace gas vertical gradients according to wind speed classes and identifying when the meteorological model satisfactorily simulates boundary layer depth (BLD), we found that non-afternoon observations can be assimilated when wind speed is >5 m/s. This condition resulted in small modeled BLD biases (<40%) when compared to calmer conditions (>100%). For Indianapolis, 37% of the GHG measurements meet this wind speed criterion, almost tripling the observations retained for AIM. Similar results are expected for windy cities like Auckland, Melbourne, and Boston, potentially allowing AIM to assimilate up to 60% the total (24-h) observations. Incorporating these observations in AIMs should yield a more diurnally comprehensive evaluation of urban GHG emissions.