Abstract Orography has a dramatic influence on cool-season precipitation, building an upper-elevation snowpack for regional water resources and winter tourism and modulating winter storms that impact transportation and public safety. Using K-band profiling Micro Rain Radar (MRR) observations from the Wasatch Mountains (2682 m MSL) and the adjoining Salt Lake Valley (1372 m MSL), this paper investigates the characteristics of cool-season storms in a continental mountain environment of the eastern Great Basin of North America. Compared to the valley site, the mountain site observed a higher frequency of low-level reflectivity ≥ 5 dB Z e , consistent with greater precipitation frequency due to orographic enhancement. Echoes were deepest during southerly or southwesterly flow with high integrated vapor transport and the passage of cold fronts or baroclinic troughs and shallowest during northwesterly flow postcold-frontal periods when median −10-dB Z e echo tops were only 1260 m AGL. During postcold-frontal periods, reflectivities most often increased with decreasing height near the ground at the mountain site, whereas other storm types featured nearly equal frequencies of increasing and decreasing reflectivity with decreasing height. Due to presumed subcloud sublimation and evaporation, postcold-frontal periods at the valley site featured decreasing modal reflectivity with decreasing height near the ground. These results illustrate important contrasts in precipitation growth and loss processes between mountain and lowland sites, where subcloud sublimation and evaporation represent underappreciated mechanisms for reducing lowland precipitation and enhancing the orographic precipitation gradient. Operational and spaceborne precipitation radars likely inadequately sample these near-surface processes in continental mountain environments. Significance Statement Mountains strongly affect winter storms, complicating forecasting and enabling the development of an upper-elevation snowpack that is vital for water resources and winter tourism. This research shows that contrasts in valley and mountain precipitation in drier, continental mountain environments are strongly influenced by shallow processes near the ground, which are poorly sampled by conventional weather radars and satellites. In the mountains, precipitation growth can occur near the ground, especially in postcold-frontal environments, whereas valley precipitation can decrease as it falls through a dry layer. The latter is an underappreciated contributor to the mean increase of precipitation with elevation found in continental mountain environments.
Abstract Despite decades of research, forecasters still lack the ability to consistently determine whether a given supercell in a tornado-favorable environment will undergo imminent tornadogenesis and, if so, anticipate the tornado’s peak intensity. The key ubiquitous feature of supercells is their cyclonically rotating midlevel updraft—the midlevel mesocyclone. A crucial intermediate step between midlevel mesocyclone generation and tornadogenesis is the production of low-level rotation—the low-level mesocyclone. However, little observational work has been done to investigate low- or midlevel mesocyclone characteristics over long periods prior to the time of tornadogenesis or tornadogenesis failure in a bulk sense. In this study, we use Weather Surveillance Radar-1988 Doppler azimuthal shear data at supercell low and midlevels to investigate mesocyclone intensity and transience differences during the hour preceding genesis and failure for nontornadic, weakly tornadic, and strongly tornadic cases. The mean near-storm environment was approximated by aggregating Rapid Refresh model data from within a given supercell’s near-field inflow region over the hour-long analysis period to determine any relationships between the environment and changes in mesocyclone intensity and transience. Overall, significant differences are found between the aforementioned mesocyclone characteristics of nontornadic and tornadic low-level mesocyclones and also between strongly tornadic and weakly/nontornadic midlevel mesocyclones. We find no evidence to attribute differences in low-level mesocyclone intensity, intensity change, or duration to any near-storm environment characteristics. However, midlevel mesocyclone intensity shows an increasingly significant relationship with convective available potential energy approaching tornadogenesis in both tornadic case types.
Abstract This paper investigates the performance of a unique proof-of-concept hybrid model in a cycling data assimilation scheme. This previously published model combines the Simplified Parameterization, Primitive Equation Dynamics (SPEEDY) model with a machine learning (ML)-based component that itself is capable of modeling the global atmospheric dynamics. Analysis and forecast experiments are carried out assuming that ERA5 reanalyses, interpolated to the model grid, represent the “true” spatiotemporal evolution of the atmosphere. Six-hourly simulated observations are generated for a 30-yr training period and a 1-yr testing period by randomly perturbing the true states. To investigate the effect of the training data on the model performance, the model is trained on different datasets in the different experiments: The training data are either ERA5 reanalyses, analyses prepared using SPEEDY for cycling, or analyses prepared using the hybrid model for cycling. The simulated observations are assimilated with a local ensemble transform Kalman filter (LETKF), and the length of the ensuing forecasts is 10 days in all experiments. The cycled LETKF remains stable for the entire testing period in all experiments. When the hybrid model is trained on ERA5 reanalyses, the biases of the analyses are negligible and the variance of the analysis error is greatly reduced compared to the experiment in which SPEEDY rather than the hybrid model is used for cycling. The gains in analysis accuracy are more modest when the hybrid model is trained on analyses obtained with SPEEDY or a prior trained version of the model. All forecasts with the hybrid model are more accurate than with SPEEDY. Significance Statement This is the first study to investigate the performance of a unique proof-of-concept hybrid model in a cycling data assimilation scheme. This previously published model combines a physics-based model using its original physical parameterization schemes with a machine learning (ML)-based component that itself is capable of modeling the global atmospheric dynamics.
Abstract Ockhi in the Arabian Sea (AS) was unique because of its long track, long duration, rapid intensification (RI), and rapid weakening (RW). Predicting the RI and RW of tropical cyclones is a big challenge due to the complex interaction between the factors responsible for the intensity changes. The present study attempts to understand the ocean’s role in the RI and RW of Ockhi. The changes in a cyclone’s inner-core sea surface temperature (SST) can dramatically change the enthalpy fluxes from the ocean to the atmosphere. The fortuitous encounter between the Ocean Moored Buoy Network in the North Indian Ocean (OMNI) buoys in the AS and Ockhi allowed an understanding of the role of inner-core SST in RI and RW. Surface meteorological measurements from the moorings are used to estimate enthalpy fluxes. The one-dimensional mixed layer Price–Weller–Pinkel (PWP) model is used to estimate the entrainment cooling. The subsurface current and density are used to estimate the shear and buoyancy terms in the Richardson number. Weaker entrainment mixing and reduced inner-core SST cooling facilitated the RI of the cyclone over the Lakshadweep Sea in the southeastern AS. The shallow mixed layer associated with the cold-core eddy region of the northeastern AS favored enhanced entrainment mixing and, hence, facilitated the RW of the cyclone. The current work emphasizes the importance of incorporating precise inner-core SST in models to improve predictions of the RI and RW of cyclones in the AS. Significance Statement The ocean provides the heat required for the cyclone intensification. The strong wind associated with the cyclone causes mixing between the warm upper water and the colder water below it. Sea surface cooling due to mixing may affect cyclone intensification. The results of the present study for the cyclone Ockhi in the Arabian Sea indicate that weaker sea surface cooling in the southeastern Arabian Sea favored storm intensification, whereas enhanced cooling in the northeastern Arabian Sea facilitated weakening. The deep, warm waters in the Lakshadweep Sea during December make the region more prone to weaker sea surface cooling and, consequently, to cyclone intensification.
Abstract The American Society of Civil Engineers (ASCE)/Structural Engineering Institute (SEI)/American Meteorological Society (AMS) Standards Committee on Estimating Wind Speeds in Tornadoes and Other Windstorms has undertaken a long-term effort to develop a multimethod standard for improving estimations of tornado wind speeds. While these methods show promise at discriminating general tornado intensity trends among themselves, at least two significant challenges remain: 1) discriminating finer details of intensity estimates among upper-echelon-intensity tornadoes and 2) comparing across different estimation techniques. Most techniques approach effective limits at high tornado intensities that represent a lower-bound intensity estimate. Furthermore, different intensity estimation techniques are often inherently representing different wind speed variables or characteristics—e.g., the enhanced Fujita (EF) scale estimates a peak 3-s gust speed at essentially single points, treefall pattern estimation methods estimate peak quasi-instantaneous wind gusts over an area, and gust periods associated with radar measurements can vary based on the radar gate spacing and tornado wind intensity. In this study, we use a simple modified Rankine vortex model to illustrate the sensitivity of the relationship between the peak instantaneous wind gust and peak 3-s-average wind gusts associated with idealized tornadic vortices as functions of vortex intensity, size, translation speed, and ratio of inflow to rotational flow. We illustrate that all of these factors impact the instantaneous-to-3-s gust speed relationship nonlinearly, even under the most idealized condition of a modified Rankine vortex. These results highlight the importance of understanding what each tornado intensity estimation method is actually estimating and the need for improved understanding of which gust periods are most relevant for causing tornado damage to various damage indicators. Significance Statement Recent efforts have led to the development of several ways to estimate tornado intensity. However, these methods vary in the duration of the wind gust that they estimate, from an instantaneous peak gust to average speeds over 3 s or longer. This study shows that using comparable gust durations is vital to understanding how tornado intensity estimates compare to each other across different estimation methods. Vortex size, intensity, translation speed, and other key attributes all contribute to substantial complexity between wind speed magnitudes for different gust durations in tornadoes.
Abstract Despite the critical roles of boundary layer turbulent processes in tropical cyclone (TC) development, the distinct contributions of the vertical extent and peak magnitude of vertical eddy diffusivity ( K m ) to TC intensification remain insufficiently understood. Using a series of Global-to-Regional Integrated Forecast System simulations of Typhoon Lekima (2019), this study investigates how the vertical extent and magnitude parameters, embedded within the eddy-diffusivity mass-flux planetary boundary layer scheme, modulate turbulent processes and influence TC intensification. The findings suggest that the vertical extent parameter impacts both TC early spin-up efficiency and rapid intensification (RI) physics, whereas the magnitude parameter primarily affects RI. Specifically, during vortex spin-up, a reduced vertical extent enhances the turbulent moisture flux gradient and diminishes vertical diffusion, thereby promoting feedback in the wind-induced surface heat exchange (WISHE). As RI begins, a reduced vertical extent weakens downward mixing of azimuthal momentum, amplifying boundary layer gradient imbalance. In contrast, the weakening of turbulent mixing resulting from the reduction in peak magnitude becomes pronounced before RI onset. This ensuing boundary layer imbalance thus facilitates RI initiation. These findings establish a unified framework, demonstrating how the vertical extent parameter modulates early WISHE-driven spin-up, while both parameters influence gradient wind imbalance during RI, thereby linking turbulent processes to the stage-dependent dominance of TC intensification mechanisms.
Abstract In ensemble-based data assimilation for numerical weather prediction, ideal covariance localization dampens spurious correlations due to computational limitations in ensemble size while retaining correlations representing any underlying flow-dependent features. This study introduces a new vertical flow-dependent localization (vFDL) method and implements and evaluates the approach within the Global Forecast System (GFS) hybrid four-dimensional ensemble–variational (4DEnVar) data assimilation system. Vertical localization profiles are calculated based upon the vertical autocorrelation depth of 6-h ensemble forecasts of horizontal wind over a period of eight cycles. These profiles are categorized into bins and assigned based upon how much variance is explained by the first eigenvector for a vertical column of the horizontal wind ensemble covariance at each grid point. Global and tropical cyclone track forecasts using vFDL show significant improvement over forecasts using domain-invariant vertical localization at up to 5-day lead times. Forecast improvements are most notable at jet and near-surface levels, which typically exhibit the largest and smallest correlation length scales, respectively, compared to other tropospheric levels. Diagnostics suggest that vFDL allows more accurate background error correlations to be used for data assimilation, including extratropical fronts and large-scale flow in the vicinity of tropical cyclones. Globally, assigning four different vertical localization profiles leads to greater improvements than assigning two or eight different profiles. Additionally, updating which localization profiles are used each cycle does not generally yield statistically significant improvements over keeping the localization profiles fixed over the experiment period.
Abstract Forecasting monsoon precipitation over Arizona is challenging, partly due to its complex terrain. The model grid structure may misrepresent topographic details, and the sparse observation network is insufficient for the initialization of the model at the scale of the topography (∼4 km), particularly in defining the spatial distribution of moisture. Our study aims to assess monsoon precipitation forecast skill over Arizona by simulating 24 precipitation events of the 2021 monsoon season in convective-permitting Weather Research and Forecasting (WRF) Model simulations from a 40-member ensemble coupled with data assimilation (DA). The High-Resolution Rapid Refresh (HRRR) model is used as the initial and boundary conditions. The data being assimilated are hourly global positioning system precipitable water vapor (GPS-PWV) data, collected from 31 sites over the Southwest United States, including special observations collected in Arizona during the 2021 monsoon. The configuration of the simulations is based on seven experiments in which microphysics schemes and horizontal localizations were varied. Our results show: 1) GPS-PWV data assimilation reduced forecast PWV errors and biases during the DA cycle and forecast periods; 2) the assimilation increased the instability of the preconvective atmosphere due to moistening over the Mogollon Rim and southeastern Arizona by as much as 1000 J kg −1 , persisting for at least 6 h into the forecasts; and 3) the assimilation improves the precipitation forecast skill up to 9 h into the forecasts and lowers the biases in the temperature, dewpoint, and mixing ratio within at least 3 km above ground level. Significance Statement Forecasting rainfall is difficult due to complex land–atmosphere interactions. Our study assesses the skill of a high-resolution numerical weather prediction (NWP) model in forecasting monsoon rainfall over Arizona. By better specification of the initial atmospheric moisture through assimilating an observed moisture dataset from global positioning system (GPS) receivers, our study finds that the errors and biases in the modeled moisture are reduced at the initial hour of the forecasts, the modeled atmospheric instability increases, i.e., it is more favorable for convection to occur during the first 6 h of the forecasts, and the modeled precipitation is more comparable with the observed precipitation in the first 9 h into the forecasts.
Abstract Landfalling atmospheric rivers (ARs) deliver critical wintertime precipitation to the western United States, producing substantial socioeconomic impacts. Improving their forecasts across time scales is, therefore, crucial for improved preparedness. While radar data assimilation has proven effective for short-range forecasts of convective systems, its impacts on AR-related precipitation forecasts remain unclear. This study addresses this gap by examining radar data impacts during two high-impact landfalling ARs in January and October 2021. The Gridpoint Statistical Interpolation system is enhanced to directly assimilate reflectivity and dense radial velocity with a horizontal observation spacing comparable to the model grid size. Experiments use a 30-min cycling framework over a 6-h window, followed by a 6-h free forecast. Assimilating radial velocity yields moderate improvements in precipitation forecasts during cycling, with postcycling benefits observed primarily in the more dynamically driven October case. Overall, reflectivity assimilation shows a more beneficial impact than radial velocity, with sustained improvements during heavy precipitation but mixed impacts during the onset phase. Reflectivity sensitivity experiments confirm that degradation is associated with negative innovations. Further analysis shows that those innovations exhibit characteristics consistent with representativeness errors arising from vertical-scale mismatches between radar sampling volumes and model grids, when radar beam geometry is neglected in the reflectivity observation operator. Forecast degradation occurs when innovations are influenced by representativeness errors, although contributions from model deficiencies cannot be ruled out. This study highlights the promising applicability of radar data for improving wintertime precipitation, while it underscores the need to refine the existing reflectivity operator to maximize data benefits.
Abstract Spurious convective cells degrade quantitative precipitation forecasts (QPF) in numerical models. This study develops a background-dependent pseudo-humidity retrieval and assimilation scheme within the WRFDA 3DVar indirect radar data assimilation framework to suppress spurious convection and false precipitation. Pseudo-humidity observations are constructed from background and observed reflectivity through a three-regime strategy: 1) a saturation adjustment for strong echoes (>25 dBZ), 2) a Bayesian retrieval for non-missing weak echoes (≤25 dBZ) and 3) an environmental neighborhood estimation for spurious convection regions (SCRs) characterized by strong background reflectivity but null echoes in observations. The scheme was first examined in the 6 July 2019 Jiangsu squall-line case. Compared with a baseline configuration and a reflectivity-based pseudo-humidity assimilation without the null-echo extension, the proposed scheme substantially reduced spurious moisture within SCRs, weakens moisture convergence and vertical upward motion, and suppresses false precipitation. A one-month batch evaluation further demonstrated robustness by consistently reducing positive precipitation bias and false alarms across thresholds and improving forecasts for both extreme-rainfall and scattered-convection events.
Abstract The assimilation of all-sky infrared brightness temperatures (BTs) into short-term convective-scale forecasts has led to forecast improvements. While most studies employ an ensemble-based approach to achieve this, some limitations, such as the treatment of sampling errors and insufficient model spread, have been addressed in different ways. With the ultimate goal of informing more flexible cloud-adaptive treatments of these issues, the present study aims to better understand the vertical structure of systematic relationships between different atmospheric state variables and BT under different cloud and precipitation conditions. In particular, results are evaluated for high-altitude clouds, both with and without robust precipitation present, and for midaltitude clouds with cloud tops both above and below 400 mb (1 mb = 1 hPa). The systematic relationships to BTs are described across the different cloud categories defined within the study, and case studies are used to connect the systematic results to physical processes or features. In the high-cloud categories, the conclusions made came from a mesoscale convective system. For high-precipitating high clouds, BTs were related to the overall strength of convection and the proximity to the localized updrafts. In nonprecipitating high clouds, BT anomalies were related to the depth and hydrometeor concentration of anvil-like clouds as well as midlevel temperature and dewpoint anomalies associated with organized convection. Results in the midlevel cloud categories appeared to be generally dominated by synoptic-scale features. When cloud tops were above 400 mb, BTs were mainly related to ice-phase hydrometeors. When cloud tops were below 400 mb, BTs were mainly related to liquid-phase hydrometeors and included a contribution from what is considered the clear-air regime above the cloud.
Abstract An extreme precipitation event occurred over the concave terrain in northeastern Taiwan on 26 November 2021 during intensive observing period (IOP) 2 of the Yilan Experiment of Severe Rainfall (YESR). The IOP2 extreme precipitation was associated with a northeasterly monsoon flow exceeding 15 m s −1 , notably stronger than that observed during IOP1, the latter being consistent with typical winter events in Taiwan. Radar wind retrievals over the concave terrain reveal that this strong northeasterly winter monsoon advanced southward and was deflected into a northwesterly flow (NW) over the southern mountains. Orographic ascent and convergence between the northeasterly flow and the NW induced an updraft over the southern mountains, leading to the development of deep convection extending to above 5 km. Convection above 4 km was influenced by a southwesterly flow that advected precipitation northeastward from the updraft region, causing a spatial offset between updraft and precipitation—contrasting with the shallow, collocated convection typical of winter events in Taiwan.
Abstract The complex terrain in mountainous regions makes it extremely difficult to accurately measure or model snowfall, which is a key component of the terrestrial water budget. This study addresses these challenges by using high-altitude frozen lakes as pressure-sensing surfaces to produce accurate observations of the water content of snowfall at a range of sites in the European Alps, west–central Himalayas, and central Rockies, which are subsequently used to test and constrain snowfall output from a 1.5-km resolution version of the atmosphere-only Met Office Unified Model (MetUM). The model resolution is on a similar scale to the size of the lakes, as well as sufficiently fine to represent the critical interactions between atmospheric flows and the complex orography that influences snowfall and especially extremes. The snowfall output from the MetUM is additionally fine tuned by adjusting the fall speed of snow particles so that it is best able to capture the observed snowfall amounts, especially for extreme events. The results presented here show that the MetUM is generally able to accurately simulate both the timing and amounts of the snowfall observations over mountainous regions. Moreover, the model is particularly good at representing extreme snowfall events, with our study also using model hydrometeor output to examine the microphysical conditions related to these conditions. Finally, we suggest that the model output can effectively be used as pseudo-observations and for the generation of high-resolution, long-term gridded snowfall products. Significance Statement Snowfall in mountainous regions is a vital source of freshwater, sustaining rivers that support both large populations and diverse ecosystems. However, the complex topography of these areas poses significant challenges for accurately measuring snowfall. This study tackles these difficulties by using high-altitude frozen lakes as natural pressure sensors to monitor snowfall water content across sites in the European Alps, the west–central Himalayas, and the central Rockies. These observations are used to constrain and refine snowfall simulations from a high-resolution version of the Met Office Unified Model. The results show that the model can accurately simulate both the timing and amounts of the snowfall observations and can effectively be used for the generation of long-term snowfall products.
Abstract This study investigates data assimilation (DA) of all-sky satellite infrared (IR) and visible (VIS) observations for a real-world rain event in China. We focus on brightness temperature from the IR channel 10 (6.9–7.3 μ m) and reflectance from the VIS channel 2 (0.55–0.75 μ m) of the Advanced Geostationary Radiation Imager (AGRI) on board the Fengyun-4A geostationary satellite. The IR and VIS observations are assimilated using a localized particle filter, which is incorporated into the Data Assimilation Research Testbed (DART) coupled with the Weather Research and Forecasting (WRF) Model. The forecasts are verified using multisource-observed precipitation products, radiosonde observations, and equivalent radar reflectivity factor. The results indicate that DA of IR observations improves the 12- and 24-h forecasts of light (0.1–10.0 mm) and moderate (10.1–25.0 mm) precipitation, as well as temperature and humidity. However, it does not improve the forecasts of heavy precipitation (25.1–50.0 mm). Sequential DA of VIS observations following IR observations yields added value, as evidenced by higher equitable threat scores for precipitation forecasts and smaller biases in temperature and humidity. The added value stems from the complementarity of VIS observations with IR observations, providing additional cloud information. Moreover, joint DA of IR and VIS observations mitigates ineffective resampling processes due to nearly identical particle weights compared with assimilating IR observations alone. Furthermore, the study discusses ambiguities in vertical localization and hydrometeor types for the DA of IR and VIS observations. The ambiguities should be a primary research focus for future studies.
Abstract Simulated supercell thunderstorms were initiated using four base-state thermodynamic profiles to test the influence of variation in both the planetary boundary layer and the free-tropospheric relative humidity on downdrafts and cold pool characteristics. The selected base-state environments include different combinations of boundary layer and free-tropospheric relative humidity and closely resemble those of four observed supercells. Bulk cold pool properties were more dependent on boundary layer relative humidity, with more negatively buoyant outflow when the base-state boundary layer was dry. Free-tropospheric relative humidity had a secondary effect on cold pool intensity by modifying precipitation production. The most intense cold pool was the result of the combination of a dry boundary layer that promoted greater rates of evaporation and a moist free troposphere that produced large amounts of precipitation. The relationship between different relative humidity regimes and cold pool characteristics was mostly insensitive to changes in CAPE, shear profile, microphysics, surface drag, random temperature perturbations, and temperature profiles. Significance Statement Supercells are persistently rotating thunderstorms that produce the majority of significant tornadoes. Their rainy downdrafts and associated evaporatively cooled low-level outflow (cold pools) often create severe winds and are thought to strongly influence processes that lead to tornadoes. We explore the role that relative humidity plays in the form and evolution of supercells, focusing on precipitation production and cold pool characteristics. Temperature and moisture profiles near observed supercells are sorted into four distinct relative humidity regimes and used to initiate supercells in a series of high-resolution simulations to show that cold pools are impacted by both near-surface and mid- to upper-level relative humidity. The strongest cold pools are produced when it is dry in the low levels and moist aloft.
Abstract Strong low-level mesocyclones provide vertical acceleration needed to stretch near-surface vertical vorticity into tornadoes. Previous simulations of supercells suggest that coherent tube-like features of enhanced streamwise vorticity called streamwise vorticity currents (SVCs) may strengthen low-level mesocyclones, perhaps increasing the likelihood of tornadogenesis. Such simulations often use a free-slip lower boundary condition, but SVCs exist near the surface and may be influenced by surface drag. Thus, we compare simulations of supercells with semislip and free-slip lower boundary conditions to determine how supercell structure and SVC development may be impacted by surface drag. We find that left-flank convergence boundaries (LFCBs) are more common in free-slip simulations and that forward-flank convergence boundaries (FFCBs) are slightly more common in semislip simulations. Convergence beneath updrafts in semislip simulations is steadier than in free-slip simulations, yielding weaker but steadier low-level updrafts. SVCs in free-slip simulations exhibit a density current head-like shape, while SVCs in semislip simulations exhibit a horizontal tube-like shape. SVCs along LFCBs form similarly regardless of the lower boundary condition, but near-surface streamwise vorticity is less in semislip simulations. SVCs along FFCBs are less susceptible to the effects of drag because trajectories are farther above the surface. Stretching is the dominant contributor to the large streamwise vorticity in SVCs. Baroclinically generated streamwise vorticity adds to the vorticity before stretching occurs, but initial streamwise vorticity is the most important. We conclude that SVCs along LFCBs in free-slip simulations are likely too strong and that SVC influences on low-level mesocyclone intensity may be overestimated in such simulations. Significance Statement Streamwise vorticity currents (SVCs) are tubes of horizontally rotating air from within a supercell thunderstorm. If they are tilted vertically by a supercell updraft, they can enhance vertical rotation in supercell thunderstorms and possibly lead to tornado production. Simulations with surface drag are compared to those without drag to investigate the impact of drag on supercell structure and SVC development. We find that SVCs are likely too strong when drag is not included in the simulations.
Abstract This study compares the climatology and environmental analyses of Global Precipitation Measurement Mission satellite passive microwave (PMW) hail estimates in subtropical South America (SSA) and the continental U. S. (CONUS) to previous analyses conducted using ground- and/ or radar-based reports. In CONUS and SSA, PMW high-probability hail events occurred more frequently near higher terrain, unlike null-probability events which occurred farther east. The high-probability events occurred more frequently later in the day, up to midnight local time in CONUS. This finding differs significantly from past report- and radar-based CONUS climatologies. Both terrain and daily timing findings align more closely with the climatology of mesoscale convective systems. These results agree with recent studies finding PMW-based hail retrievals can disproportionally assign a high probability of hail to large systems with deep layers of graupel, introducing a potential bias toward large (i.e., linear) systems. Subsequent environmental analyses agree with this conclusion, with higher hail probabilities associated with drier low and mid-levels, in contradiction with previous literature and more commonly associated with linear systems. Future work establishing radar- and reports-based hail climatologies in SSA is recommended for further comparison. Environmental profiles selected from an hour prior to convective initiation instead of from a common time during the day showed more discriminatory power among hail probabilities in both CONUS and SSA, highlighting the importance of shorter-timescale modifications prior to convective initiation. When evaluating all convection regardless of hail probability, SSA convection was associated with smaller CAPE, weaker mid-level lapse rates, weaker low-level shear, but higher specific humidity over all layers than CONUS.
Abstract The Taiwan–Luzon Island Arc—comprising Taiwan Island, the Luzon Strait, and Luzon Island—is a unique geographical corridor that strongly modulates tropical cyclone (TC) activity in the western North Pacific. Based on best track observations, TCs are classified into four categories relative to the Taiwan–Luzon barrier: north of Taiwan, crossing Taiwan Island, through the Luzon Strait without landfall, and crossing Luzon Island. Statistical analyses reveal distinct intensity attenuation patterns: TCs crossing Taiwan Island experience the strongest weakening (average 19.2%), followed by those crossing Luzon Island (16.7%). TCs to the north of Taiwan also weaken moderately, whereas those through the Luzon Strait exhibit no significant mean change but high individual variability. The dominant environmental controls on postcrossing intensity evolution vary by region: Weak vertical wind shear favors intensification for TCs traversing the Luzon Strait, whereas warm sea surface temperatures and a moist midtroposphere are more important for postlandfall intensification west of Luzon Island. Pathway-dependent differences are also evident in TC translation speed. Prelandfall acceleration occurs mainly for island-crossing TCs and is primarily controlled by island topography. TCs approaching Taiwan Island exhibit earlier and stronger acceleration than those crossing Luzon Island, reflecting differences in the timing and mechanisms of terrain–TC interaction. Overall, the results highlight strong pathway dependence of TC intensity and motion across the Taiwan–Luzon Island Arc, reflecting contrasting roles of island topography and environmental conditions along different segments of the island arc.
Abstract The forecasts from a numerical weather prediction (NWP) model have systematic biases and cannot be used directly. Statistical calibration is needed to generate ensemble forecast that are accurate and reliable. Typically, it is carried out on an individual gridcell basis, with subsequent application of ensemble-reordering approaches to incorporate the spatial structures in the calibrated ensemble forecast. These ensemble-reordering methods, notably the widely employed Schaake shuffle approach, are based on certain templates and have a few limitations. Using convolutional neural network (CNN), we propose two models for postprocessing the precipitation forecast and for generation of ensemble forecasts. These ensemble forecasts display the spatial structure, thereby removing dependence on ensemble reordering. CNNs are used for forecast calibration and extracting the spatial information; Monte Carlo (MC) dropouts are then used for producing ensemble forecasts. The traditional methods are implemented on individual grid cells, whereas the models we propose are applied to the whole forecast field. The models are implemented on NWP raw forecasts for the Brisbane drainage basin situated in east Australia. They are assessed on all the precipitation levels, including no, low, and high rainfall events. Results demonstrate that ensemble forecasts are well calibrated at basin and gridcell scales, for all precipitation ranges. The uncertainty is estimated reliably, leading to skillfully calibrated ensemble forecasts.
Abstract The comma heads of winter cyclones have a variety of precipitation structures ranging from cells to bands. Much of the previous research has explored the environmental conditions for larger (primary) snowbands in the cyclone comma head, with less work comparing the environments of the broader spectrum of snowband structures. This study looks at these environments for a full range of object sizes and shapes for cool-season cyclones over the northeast United States (NEUS) from 1996 to 2023. The ERA5 reanalysis is used to obtain the environmental parameters and cyclone tracks. Only a weak relationship exists between different object characteristics and parameters such as frontogenesis, stability, and vertical shear. A self-organizing map (SOM) approach was applied to specific regions of the cyclone comma head, and the analysis was separated into different cyclone track orientations over the NEUS. The environmental relationships are somewhat more robust using the SOM technique, such as stronger midlevel frontogenesis in regions with more prevalent large bands and greater low-level vertical shear in regions with more frequent amorphous objects; however, the environments are still not statistically different for each precipitation object type. Given this result and the large spread in environmental ingredients for each object type, it is hypothesized that the objects may have environmental differences that evolve from the development to mature stages. Significance Statement A broad spectrum of organized precipitation structures exists in winter storms leading to variations in snowfall in a storm, but the environment within winter storms is complex. This study attempts to identify the environments that favor different precipitation structures. Although regions with a higher frequency of large, band-like objects tend to coincide with stronger midlevel forcing for ascent, weak stability, and some weak low-level vertical shear, the results are not statistically significant. Thus, this motivates more research to better understand how the environment changes as objects grow during their life cycle.
