The 10 most-cited articles in Monthly Weather Review, 2019

Abstract Deep learning models, such as convolutional neural networks, utilize multiple specialized layers to encode spatial patterns at different scales. In this study, deep learning models are compared with standard machine learning approaches on the task of predicting the probability of severe hail based on upper-air dynamic and thermodynamic fields from a convection-allowing numerical weather prediction model. The data for this study come from patches surrounding storms identified in NCAR convection-allowing ensemble runs from 3 May to 3 June 2016. The machine learning models are trained to predict whether the simulated surface hail size from the Thompson hail size diagnostic exceeds 25 mm over the hour following storm detection. A convolutional neural network is compared with logistic regressions using input variables derived from either the spatial means of each field or principal component analysis. The convolutional neural network statistically significantly outperforms all other methods in terms of Brier skill score and area under the receiver operator characteristic curve.…

Abstract By 31 December 2018, 39 537 quality-controlled reports of large hail had been submitted to the European Severe Weather Database (ESWD) by volunteers and ESSL. This dataset and the NatCatSERVICE Database of Munich RE jointly allowed us to study the hail hazard and its impacts across Europe over a period spanning multiple decades. We present a spatiotemporal climatology of the ESWD reports, diurnal and annual cycles of large hail, and indicate where and how they may be affected by reporting biases across Europe. We also discuss which hailstorms caused the most injuries and present the only case with hail fatalities in recent times. Additionally, we address our findings on the relation between hail size to the type of impacts that were reported. For instance, the probability of reported hail damage to roofs, windows, and vehicles strongly increases as hail size exceeds 5 cm, while damage to crops, trees, and greenhouses is typically reported with hailstone diameters of 2–3 cm. Injuries to humans are usually reported with hail 4 cm in diameter and larger, and number of injuries…

Abstract A deep learning convolutional neural network model is used to explore the possibilities of estimating tropical cyclone (TC) intensity from satellite images in the 37- and 85–92-GHz bands. The model, called “DeepMicroNet,” has unique properties such as a probabilistic output, the ability to operate from partial scans, and resiliency to imprecise TC center fixes. The 85–92-GHz band is the more influential data source in the model, with 37 GHz adding a marginal benefit. Training the model on global best track intensities produces model estimates precise enough to replicate known best track intensity biases when compared to aircraft reconnaissance observations. Model root-mean-square error (RMSE) is 14.3 kt (1 kt ≈ 0.5144 m s−1) compared to two years of independent best track records, but this improves to an RMSE of 10.6 kt when compared to the higher-standard aircraft reconnaissance-aided best track dataset, and to 9.6 kt compared to the reconnaissance-aided best track when using the higher-resolution TRMM TMI andAquaAMSR-E microwave observations only. A shortage of training…

Abstract Warm-sector heavy rainfall in southern China refers to the heavy rainfall that occurs within the warm sector hundreds of kilometers south of a front or without a front during April–June, characterized by poor predictability and a close relationship with low-level jets (LLJs). Based on 45 warm-sector heavy rainfall episodes in 2013 and 2014 in southern China, this study examines their general characteristics and evaluates the performance of convection-permitting WRF Model simulations from an LLJ perspective. The results show that 64% of the warm-sector heavy rainfall episodes are associated with an LLJ (LLJ type) and 36% are not (no-LLJ type). The LLJ type is distinct from the no-LLJ type, with large rainfall accumulation along the coastal area. It is more common for LLJs to occur at both 800 and 925 hPa in the LLJ type, where there is a wide 800-hPa LLJ west of Guangdong Province and two 925-hPa LLJs over Beibu Gulf and the South China Sea (SCS). The coastal convergence associated with the terminus of the LLJ on 925 hPa is conducive to the coastal rainfall. WRF generally…

Abstract The value of ensemble forecasts is well documented. However, postprocessing by statistical methods is usually required to make forecasts reliable in a probabilistic sense. In this work a flexible statistical method for making probabilistic forecasts in terms of quantile functions is proposed. The quantile functions are specified by linear combinations of Bernstein basis polynomials, and their coefficients are assumed to be related to ensemble forecasts by means of a highly adaptable neural network. This leads to many parameters to estimate, but a recent learning algorithm often applied to deep-learning problems makes this feasible and provides robust estimates. The method is applied to ~2 yr of ensemble wind speed forecasting data at 125 Norwegian stations for lead time +60 h. An intercomparison with two quantile regression methods shows improvements in quantile skill score of nearly 1%. The most appealing feature of the method is arguably its versatility.

Abstract Atmospheric blocking has been shown to be a phenomenon that models struggle to predict accurately, particularly the onset of a blocked state following a more zonal flow. This struggle is, in part, due to the lack of a complete dynamical theory for block onset and maintenance. Here, we evaluate the impact cyclone representation had on the forecast of block onset in two case studies from the North Atlantic Waveguide and Downstream Impact Experiment field campaign and the 20 most unpredictable block onsets over the Euro-Atlantic region in medium-range forecasts from the ECMWF. The 6-day forecast of block onset in the case studies is sensitive to changes in the forecast location and intensity of upstream cyclones (one cyclone for one case and two for the other case) in the days preceding the onset. Ensemble sensitivity analysis reveals that this is often the case in unpredictable block onset cases: a one standard deviation change in 1000-hPa geopotential height near an upstream cyclone, or 320-K potential vorticity near the tropopause, two or three days prior to block onset is…

Abstract Two diagnostics based on potential vorticity and the envelope of Rossby waves are used to investigate upscale error growth from a dynamical perspective. The diagnostics are applied to several cases of global, real-case ensemble simulations, in which the only difference between the ensemble members lies in the random seed of the stochastic convection scheme. Based on a tendency equation for the enstrophy error, the relative importance of individual processes to enstrophy-error growth near the tropopause is quantified. After the enstrophy error is saturated on the synoptic scale, the envelope diagnostic is used to investigate error growth up to the planetary scale. The diagnostics reveal distinct stages of the error growth: in the first 12 h, error growth is dominated by differences in the convection scheme. Differences in the upper-tropospheric divergent wind then project these diabatic errors into the tropopause region (day 0.5–2). The subsequent error growth (day 2–14.5) is governed by differences in the nonlinear near-tropopause dynamics. A fourth stage of the error…

Abstract During the 2015 Plains Elevated Convection at Night (PECAN) field campaign, several nocturnal low-level jets (NLLJs) were observed with integrated boundary layer profiling systems at multiple sites. This paper gives an overview of selected PECAN NLLJ cases and presents a comparison of high-resolution observations with numerical simulations using the Weather Research and Forecasting (WRF) Model. Analyses suggest that simulated NLLJs typically form earlier than the observed NLLJs. They are stronger than the observed counterparts early in the event, but weaker than the observed NLLJs later in the night. However, sudden variations in the boundary layer winds, height of the NLLJ maximum and core region, and potential temperature fields are well captured by the WRF Model. Simulated three-dimensional fields are used for a more focused analysis of PECAN NLLJ cases. While previous studies often related changes in the thermal structure of the nocturnal boundary layer and sudden mixing events to local features, we hypothesize that NLLJ spatial evolution plays an important role in such…

Abstract Coupled mesoscale–microscale simulations are required to provide time-varying weather-dependent inflow and forcing for large-eddy simulations under general flow conditions. Such coupling necessarily spans a wide range of spatial scales (i.e., ~10 m to ~10 km). Herein, we use simulations that involve multiple nested domains with horizontal grid spacings in the terra incognita (i.e., km) that may affect simulated conditions in both the outer and inner domains. We examine the impact on simulated wind speed and turbulence associated with forcing provided by a terrain with grid spacing in the terra incognita. We perform a suite of simulations that use combinations of varying horizontal grid spacings and turbulence parameterization/modeling using the Weather Research and Forecasting (WRF) Model using a combination of planetary boundary layer (PBL) and large-eddy simulation subgrid-scale (LES-SGS) models. The results are analyzed in terms of spectral energy, turbulence kinetic energy, and proper orthogonal decomposition (POD) energy. The results show that the output from the…

Abstract The Weather Research and Forecasting Model with Aerosol–Cloud Interactions (WRF-ACI) is developed for studying aerosol effects on gridscale and subgrid-scale clouds using common aerosol activation and ice nucleation formulations and double-moment cloud microphysics in a scale-aware subgrid-scale parameterization scheme. Comparisons of both the standard WRF and WRF-ACI models’ results for a summer season against satellite and reanalysis estimates show that the WRF-ACI system improves the simulation of cloud liquid and ice water paths. Correlation coefficients for nearly all evaluated parameters are improved, while other variables show slight degradation. Results indicate a strong cloud lifetime effect from current climatological aerosols increasing domain average cloud liquid water path and reducing domain average precipitation as compared to a simulation with aerosols reduced by 90%. Increased cloud-top heights indicate a thermodynamic invigoration effect, but the impact of thermodynamic invigoration on precipitation is overwhelmed by the cloud lifetime effect. A…