Prediction Of Thunderstorms Research Articles

Transfer Learning in Weather Prediction: Why, How, and What Should

Transfer learning (TL) is a popular phrase in deep learning (DL) domain. It is one of the latest artificial intelligence (AI) technologies that has a significant impact on big data analysis. Methods of traditional machine learning (ML) require the availability of an adequate quantity of training data as well as similarity of characteristics among the feature spaces corresponding to training and test data while performing supervised learning tasks. However, in real-life analytical problems, data scarcity often arises. In such scenarios, the TL approach has shown effectiveness in transferring knowledge from the source tasks that had large training data to a target task that has less training data. Basically, in TL, a model that has been trained on one task is essentially applied to a second related (but not exact) task. In this way, the issue of distribution mismatch can also be addressed. TL is not like conventional machine learning algorithms that try to learn each task starting from the beginning. Meteorological research is such an example of big data analysis which often faces the data scarcity issue. The current study addresses the contemporary challenges in weather forecasting that can be solved (or better dealt with) using TL methods. It presents a brief review of earlier research, followed by potential applications of TL algorithms to several key challenges in weather prediction, which includes the prediction of air quality, thunderstorms, precipitation, visibility, and cyclones, among others. Special emphasis is given to high-impact weather (HIW) prediction. These high-impact events are extremely difficult to predict, and they can cause enormous property damage and fatalities around the world. TL techniques have shown advantages in predicting HIW. Various challenging issues in implementing TL technology are then discussed. Finally, we address various prospects associated with TL, propose new research directions, and more importantly mention some concerns for beginners in DL-TL research. An extensive list of references is also provided.

Using Machine Learning to Predict Convection-Allowing Ensemble Forecast Skill: Evaluation with the NSSL Warn-on-Forecast System

Abstract Forecasters routinely calibrate their confidence in model forecasts. Ensembles inherently estimate forecast confidence but are often underdispersive, and ensemble spread does not strongly correlate with ensemble-mean error. The misalignment between ensemble spread and skill motivates new methods for “forecasting forecast skill” so that forecasters can better utilize ensemble guidance. We have trained logistic regression and random forest models to predict the skill of composite reflectivity forecasts from the NSSL Warn-on-Forecast System (WoFS), a 3-km ensemble that generates rapidly updating forecast guidance for 0–6-h lead times. The forecast skill predictions are valid at 1-, 2-, or 3-h lead times within localized regions determined by the observed storm locations at analysis time. We use WoFS analysis and forecast output and NSSL Multi-Radar/Multi-Sensor composite reflectivity for 106 cases from the 2017 to 2021 NOAA Hazardous Weather Testbed Spring Forecasting Experiments. We frame the prediction task as a multiclassification problem, where the forecast skill labels are determined by averaging the extended fraction skill scores (eFSSs) for several reflectivity thresholds and verification neighborhoods and then converting to one of three classes based on where the average eFSS ranks within the entire dataset: POOR (bottom 20%), FAIR (middle 60%), or GOOD (top 20%). Initial machine learning (ML) models are trained on 323 predictors; reducing to 10 or 15 predictors in the final models only modestly reduces skill. The final models substantially outperform carefully developed persistence- and spread-based models and are reasonably explainable. The results suggest that ML can be a valuable tool for guiding user confidence in convection-allowing (and larger-scale) ensemble forecasts. Significance Statement Some numerical weather prediction (NWP) forecasts are more likely to verify than others. Forecasters often recognize situations where NWP output should be trusted more or less than usual, but objective methods for “forecasting forecast skill” are notably lacking for thunderstorm-scale models. Better estimates of forecast skill can benefit society through more accurate forecasts of high-impact weather. Machine learning (ML) provides a powerful framework for relating forecast skill to the characteristics of model forecasts and available observations over many previous cases. ML models can leverage these relationships to predict forecast skill for new cases in real time. We demonstrate the effectiveness of this approach to forecasting forecast skill using a cutting-edge thunderstorm prediction system and logistic regression and random forest models. Based on this success, we recommend the adoption of similar ML-based methods for other prediction models.

BiLSTM‐based thunderstorm prediction for IoT applications

AbstractAlthough the market demand for smart devices (SDs) in the Internet of Things (IoT) era is surging, the corresponding thunderstorm protection measures have rarely attracted attention. This paper presents a thunderstorm prediction method with elevation correction, to reduce the thunderstorm damage to SDs by visually tracking thunderstorm activities. First, a self‐made three‐dimensional atmospheric electric field apparatus (3DAEFA) deployed in IoT is developed to collect real‐time AEF data. A 3DAEFA‐based localization model is established, and the localization formula after correction is derived. AEF data predicted by the bi‐directional long short‐term memory (BiLSTM) model are input to this formula to obtain thunderstorm point charge localization results. Then, the localization skill is evaluated. Finally, the proposed method is assessed in experiments, under single and multiple point charge conditions. There are significant reductions of at least 33.1% and 8.8% in ranging and elevation angle errors, respectively. Particularly, this post‐prediction correction reduces the deviation of fitted point charge moving paths by at most 0.189 km, demonstrating excellent application effects. Comparisons with radar charts and existing methods testify that this method can effectively predict thunderstorms.

A Systematic Review of the Potential Influence of Urbanization on the Regional Thunderstorm Process and Lightning Activity

In the context of global climate change, lightning disasters have emerged as a serious environmental factor that restricts the sustainable development of megacities. This paper provides a review of the research on the impact of urbanization on thunderstorm processes and lightning activity, exploring various aspects, such as aerosols, urban thermal effects, urban dynamic effects, and building morphology. Despite numerous significant achievements in the study of the impact of air pollutants on lightning activity, there is no consensus on whether aerosols serve to enhance or inhibit lightning activity. The temperature difference between the urban underlying surface and the natural underlying surface could sustain and promote the occurrence and development of convective systems, thus enhancing lightning activity. In terms of urban dynamics, the barrier effect has led to the maximum center of lightning appearing at the edge of a built-up area, which might be associated with factors, such as urban heat island (UHI) intensity, wind speed, synoptic background, and city size. Additionally, the size of a city and the height of the buildings was also an influencing factor on lightning activity. In summary, scholars have made progress in understanding the characteristics and drivers of urban lightning activity in recent years, but there are still some urgent problems that need to be solved: (1) How to analyze, comprehensively, the spatiotemporal patterns of urban lightning activity under different thunderstorm intensity backgrounds? (2) How to conduct analysis to investigate the influence of alterations in the boundary layer structure, water–heat energy balance, and water vapor circulation processes on urban lightning activity in the context of urbanization? (3) How to couple numerical models of different scales to enhance the understanding of the impact of complex underlying surfaces on urban lightning activity? Future studies could investigate the relationship between urbanization and thunderstorm/lightning activity using a combination of observational data, numerical modeling, and laboratory experiments, which holds promise for providing valuable theoretical insights and technical support to enhance the prediction, nowcasting, early warning, and risk assessment of thunderstorms and lightning in urban areas.

Improving Afternoon Thunderstorm Prediction over Complex Terrain with the Assimilation of Dense Ground-Based Observations: Four Cases in the Taipei Basin

Abstract In this study, we investigate the impact of assimilating densely distributed Global Navigation Satellite System (GNSS) zenith total delay (ZTD) and surface station (SFC) data on the prediction of very short-term heavy rainfall associated with afternoon thunderstorm (AT) events in the Taipei basin. Under weak synoptic-scale conditions, four cases characterized by different rainfall features are chosen for investigation. Experiments are conducted with a 3-h assimilation period, followed by 3-h forecasts. Also, various experiments are performed to explore the sensitivity of AT initialization. Data assimilation experiments are conducted with a convective-scale Weather Research and Forecasting–local ensemble transform Kalman filter (WRF-LETKF) system. The results show that ZTD assimilation can provide effective moisture corrections. Assimilating SFC wind and temperature data could additionally improve the near-surface convergence and cold bias, further increasing the impact of ZTD assimilation. Frequently assimilating SFC data every 10 min provides the best forecast performance especially for rainfall intensity predictions. Such a benefit could still be identified in the earlier forecast initialized 2 h before the start of the event. Detailed analysis of a case on 22 July 2019 reveals that frequent assimilation provides initial conditions that can lead to fast vertical expansion of the convection and trigger an intense AT. This study proposes a new metric using the fraction skill score to construct an informative diagram to evaluate the location and intensity of heavy rainfall forecast and display a clear characteristic of different cases. Issues of how assimilation strategies affect the impact of ground-based observations in a convective ensemble data assimilation system and AT development are also discussed. Significance Statement In this study, we investigate the impact of frequently assimilating densely distributed ground-based observations on predicting four afternoon thunderstorm events in the Taipei basin. While assimilating GNSS-ZTD data can improve the moisture fields for initializing convection, assimilating surface station data improves the prediction of rainfall location and intensity, particularly when surface data are assimilated at a very high frequency of 10 min.

Enhanced prediction of premonsoon thunderstorms over eastern India through assimilation of INSAT‐3D sounding data

AbstractThe present study assesses the impact of Indian National Satellite 3D (INSAT‐3D)‐derived atmospheric (temperature and relative humidity) profiles on the simulation of severe thunderstorms over eastern India during the premonsoon season. Two sets of numerical experiments, without assimilation (CNTL) and with assimilation of INSAT‐3D profiles (INSAT) have been conducted using the three‐dimensional variational data assimilation system of the Weather Research and Forecasting model for 11 thunderstorm events. Analysis increments are positive with INSAT profile assimilation in the lower to middle atmosphere in the thunderstorm region. The errors in the initial fields have been reduced by approx. 50% after the assimilation. After assimilation, the predictions of the surface fields (2‐m temperature, 2‐m relative humidity, and 10‐m wind speed) have been improved at the mature stage (by 37%, 26%, and 15%) compared to the CNTL. Similar improvements are noticed in the vertical profiles of relative humidity and wind speed in the INSAT experiment. The time and magnitude of updrafts and downdrafts improved in INSAT. After assimilation, rainfall intensity improved in nine cases, and time improved in four cases out of 11. The Critical Success Index (False Alarm Rate) of the INSAT ranges from approx. 0.6–0.15 (∼0.37–0.48) for all thresholds, and are higher (lower) than the CNTL. The rainfall area is improved by 25% for light rainfall, 11% for moderate rainfall, and 2% for heavy rainfall in the INSAT over the CNTL experiment. Overall thunderstorm prediction has been improved with the assimilation of INSAT‐3D data. Thus, the study highlights the viability of INSAT‐3D profiles in improving thunderstorm predictions over eastern India.

Warn-on-Forecast System: From Vision to Reality

Abstract In 2009, advancements in NWP and computing power inspired a vision to advance hazardous weather warnings from a warn-on-detection to a warn-on-forecast paradigm. This vision would require not only the prediction of individual thunderstorms and their attributes but the likelihood of their occurrence in time and space. During the last decade, the warn-on-forecast research team at the NOAA National Severe Storms Laboratory met this challenge through the research and development of 1) an ensemble of high-resolution convection-allowing models; 2) ensemble- and variational-based assimilation of weather radar, satellite, and conventional observations; and 3) unique postprocessing and verification techniques, culminating in the experimental Warn-on-Forecast System (WoFS). Since 2017, we have directly engaged users in the testing, evaluation, and visualization of this system to ensure that WoFS guidance is usable and useful to operational forecasters at NOAA national centers and local offices responsible for forecasting severe weather, tornadoes, and flash floods across the watch-to-warning continuum. Although an experimental WoFS is now a reality, we close by discussing many of the exciting opportunities remaining, including folding this system into the Unified Forecast System, transitioning WoFS into NWS operations, and pursuing next-decade science goals for further advancing storm-scale prediction. Significance Statement The purpose of this research is to develop an experimental prediction system that forecasts the probability for severe weather hazards associated with individual thunderstorms up to 6 h in advance. This capability is important because some people and organizations, like those living in mobile homes, caring for patients in hospitals, or managing large outdoor events, require extended lead time to protect themselves and others from potential severe weather hazards. Our results demonstrate a prediction system that enables forecasters, for the first time, to message probabilistic hazard information associated with individual severe storms between the watch-to-warning time frame within the United States.

Thunderstorm prediction during pre-tactical air-traffic-flow management using convolutional neural networks

Thunderstorms can be a large source of disruption for European air-traffic management causing a chaotic state of operation within the airspace system. In current practice, air-traffic managers are provided with imprecise forecasts which limit their ability to plan strategically. As a result, weather mitigation is performed using tactical measures with a time horizon of three hours. Increasing the lead time of thunderstorm predictions to the day before operations could help air-traffic managers plan around weather and improve the efficiency of air-traffic-management operations. Emerging techniques based on machine learning have provided promising results, partly attributed to reduced human bias and improved capacity in predicting thunderstorms purely from numerical weather prediction data. In this paper, we expand on our previous work on thunderstorm forecasting, by applying convolutional neural networks (CNNs) to exploit the spatial characteristics embedded in the weather data. The learning task of predicting convection is formulated as a binary-classification problem based on satellite data. The performance of multiple CNN-based architectures, including a fully-convolutional neural network (FCN), a CNN-based encoder–decoder, a U-Net, and a pyramid-scene parsing network (PSPNet) are compared against a multi-layer-perceptron (MLP) network. Our work indicates that CNN-based architectures improve the performance of point-prediction models, with a fully-convolutional neural-network architecture having the best performance. Results show that CNN-based architectures can be used to increase the prediction lead time of thunderstorms. Lastly, a case study illustrating the applications of convection-prediction models in an air-traffic-management setting is presented.

Analysis of the Gálvez–Davison Index for the Forecasting Formation and Evolution of Convective Clouds in the Tropics: Western Cuba

The Gálvez–Davison Index (GDI) is an atmospheric stability index recently developed to improve the prediction of thunderstorms and shallower types of moist convection in the tropics. Because of its novelty, its use for tropical regions remains largely unexplored. Cuba is a region that suffers extreme weather events, such as tropical storms and hurricanes, some of them worsened by climate change. This research analyzes the effectiveness of the GDI in detecting the potential for convective cloud development, using forecast data from the Weather Research and Forecasting (WRF) model for Western Cuba. To accomplish this, here, we evaluated the performance of the GDI in ten study cases from the dry and wet seasons. As part of our study, we researched how GDI correlates with brightness temperatures (BTs) measured using GOES-16. In addition, the GDI results with the WRF model are compared with results using the Global Forecast System (GFS). Our results show a high correlation between the GDI and BT, concluding that the GDI is a robust tool for forecasting both synoptic and mesoscale convective phenomena over the region studied. In addition, the GDI is able to adequately forecast stability conditions. Finally, the GDI values computed from the WRF model perform much better than those from the GFS, probably because of the greater horizontal resolution in the WRF model.

Impacts of Mountain Topography and Background Flow Conditions on the Predictability of Thermally Induced Thunderstorms and the Associated Error Growth

Abstract Thermally induced thunderstorm simulations were conducted with the Weather Research and Forecasting (WRF) Model in an idealized configuration to investigate the associated error growth and predictability. We conducted identical twin experiments with different topography and background winds to assess the impacts of these factors. The results showed that mountain topography restrains error growth at the early stage of convection development. This topographic effect is sensitive to mountain geometry and background winds: it was more noticeable in cases with higher and narrower mountains and difficult to see without background wind. The topographic effect and its sensitivity resulted from the different natures of convection initiation. However, the topographic effect became less apparent when moist convection continued growing and triggered rapid error growth. The predictability of thunderstorms is then limited at the timing after the convective activity reached its maximum. A smaller initial error or starting a simulation at a later time did not break this timing of predictability limit. Mountain topography also limitedly affected the timing of the maximum convective activity and the predictability limit. In contrast, background flows changed the convection evolution and the following predictability. The predictability limit assessed by rainfall suggested other aspects of the topographic effect. The domain-scale rainfall distribution and the intense accumulated rainfall can be adequately captured in the presence of mountains.

Insights in Hailstorm Dynamics through Polarimetric High-Resolution X-Band and Operational C-Band Radar: A Case Study for Vienna, Austria

Abstract Data from a dual-polarized, solid-state X-band radar and an operational C-band weather radar are used for high-resolution analyses of two hailstorms in the Vienna, Austria, region. The combination of both radars provides rapid-update (1 min) polarimetric data paired with wind field data of a dual-Doppler analysis. This is the first time that such an advanced setup is used to examine severe storm dynamics at the eastern Alpine fringe, where the influence of local topography is particularly challenging for thunderstorm prediction. We investigate two storms transitioning from the pre-Alps into the Vienna basin with different characteristics: 1) A rapidly evolving multicell storm producing large hail (5 cm), with observations of an intense ZDR column preceding hail formation and the rapid development of multiple pulses of hail; and 2) a cold pool–driven squall line with small hail, for which we find that the updraft location inhibited the formation of larger hailstones. For both cases, we analyzed the evolution of different ZDR column metrics as well as updraft speed and size and found that (i) the 90th percentile of ZDR within the ZDR column was highest for the cell later producing large hail, (ii) the peak 90th percentile of ZDR preceded large hailfall by 20 min and highest updraft size and speed by 10 min, and (iii) sudden drops of the 90th percentile of ZH within the ZDR column indicated imminent hailfall. Significance Statement Thunderstorm evolution on the transition from complex terrain into the Vienna basin in northeastern Austria varies strongly. In some instances, thunderstorm cells intensify once they reach flat terrain, while in most cases there is a weakening tendency. To improve our process understanding and short-term forecasting methods, we analyze two representative cases of hail-bearing storms transitioning into the Vienna basin. We mainly build our study on data from a new, cost-efficient weather radar, complemented by an operational radar, lightning observations, and ground reports. Our results show which radar variables could be well suited for early detection of intensification, and how they relate to thunderstorm updraft speeds and lightning activity.

Thunderstorms Prediction Using Satellite Images

Human life and property are put at risk by lightning, heavy hail, and strong winds caused by thunderstorms. Accurate prediction of thunderstorms has long been a difficult problem. In difference from conventional methods, the modern methods for predicting thunderstorms utilize a combination of computer models, observation (via the use of radars and satellite photos), and modern knowledge and patterns employed by thunderstorms stations. Observations: Using these methods, it is possible to make reasonably accurate predictions. Prediction models based on complex formulas run the majority of forecasting computer models. Artificial neural networks, machine learning, deep learning, and statistics are all examples of adaptive systems that alter their structure in response to information flowing through the network. In this reviewed paper, different Artificial intelligence techniques are used such as machine learning, deep learning, and ANN (Artificial neural network). Different statistical models were presented to predict thunderstorm systems. Various researchers achieved technical milestones by presenting statistical methods on thunderstorm systems using satellite data.

The Role of Instability Indices in Forecasting Thunderstorm and Non-Thunderstorm Days across Six Cities in India

Thunderstorms are one of the most damaging natural hazards demanding in-depth understanding and prediction. These convective systems form in an unstable environment which is quantitatively expressed in terms of instability indices. These indices are studied over six locations across the Indian landmass in an attempt to predict thunderstorm activity on any given day. A combination of multiple regression, logistic regression, and range analysis provides new insight into the prediction of these storms. A supervised machine learning-based logistic regression model is developed in this study for thunderstorm prediction over Patna and can be further extended for operational forecasting of Thunderstorms over the region. Critical thresholds for the instability indices are determined over the considered locations providing valuable insight into the domain of Thunderstorm prediction

3DAEFA-Based Thunderstorm Prediction System With Higher Performance

This article presents a thunderstorm prediction system with a 3-D atmospheric electric field (AEF) apparatus (3DAEFA), wherein the data source is the high-resolution 3-D AEF (3DAEF) values. The system is mainly composed of 3DAEF sensor design and calibration, 3DAEF acquisition and processing, and thunderstorm prediction, localization, and imaging. First, a self-made single-axis rotary vane 3DAEFA is employed to measure 3DAEFs. Second, a calibration device is developed to calibrate 3DAEF directions while completing intensity calibration. Considering that the AEF signal (AEFS) is disturbed by low-frequency noises, full-frequency-domain AEFS denoising is conducted through Savitzky–Golay (SG) filtering, and baseline estimation and denoising with sparsity (BEADS). A prediction model is built based on the bidirectional long short-term memory (BiLSTM) network. After inputting denoised AEFS spatial features into the model, which is extracted by a convolutional neural network (CNN), a CNN-BiLSTM model is formed. The proposed system is assessed in different weathers. There is a significant improvement in AEFS’s signal-to-noise ratio (SNR) measured by this system, especially in thunderstorm weather, compared with the original AEFS (OAEFS). Meanwhile, determining coefficients are more than 95%, showing better prediction effects. The deviation between predicted values and real values is smaller than that of OAEFS, which makes it easier to analyze AEFS characteristics. Comparisons with radar charts demonstrate that the proposed system can effectively predict thunderstorms.

Improvement and Application of a GAN Model for Time Series Image Prediction—A Case Study of Time Series Satellite Cloud Images

Predicting the shape evolution and movement of remote sensing satellite cloud images is a difficult task requiring the effective monitoring and rapid prediction of thunderstorms, gales, rainstorms, and other disastrous weather conditions. We proposed a generative adversarial network (GAN) model for time series satellite cloud image prediction in this research. Taking time series information as the constraint condition and abandoning the assumption of linear and stable changes in cloud clusters in traditional methods, the GAN model is used to automatically learn the data feature distribution of satellite cloud images and predict time series cloud images in the future. Through comparative experiments and analysis, the Mish activation function is selected for integration into the model. On this basis, three improvement measures are proposed: (1) The Wasserstein distance is used to ensure the normal update of the GAN model parameters; (2) establish a multiscale network structure to improve the long-term performance of model prediction; (3) combined image gradient difference loss (GDL) to improve the sharpness of prediction cloud images. The experimental results showed that for the prediction cloud images of the next four times, compared with the unimproved Mish-GAN model, the improved GDL-GAN model improves the PSNR and SSIM by 0.44 and 0.02 on average, and decreases the MAE and RMSE by 18.84% and 7.60% on average. It is proven that the improved GDL-GAN model can maintain good visualization effects while keeping the overall changes and movement trends of the prediction cloud images relatively accurate, which is helpful to achieve more accurate weather forecast. The cooperation ability of satellite cloud images in disastrous weather forecasting and early warning is enhanced.

Can We Integrate Spatial Verification Methods into Neural Network Loss Functions for Atmospheric Science?

Abstract In the last decade, much work in atmospheric science has focused on spatial verification (SV) methods for gridded prediction, which overcome serious disadvantages of pixelwise verification. However, neural networks (NN) in atmospheric science are almost always trained to optimize pixelwise loss functions, even when ultimately assessed with SV methods. This establishes a disconnect between model verification during versus after training. To address this issue, we develop spatially enhanced loss functions (SELF) and demonstrate their use for a real-world problem: predicting the occurrence of thunderstorms (henceforth, “convection”) with NNs. In each SELF we use either a neighborhood filter, which highlights convection at scales larger than a threshold, or a spectral filter (employing Fourier or wavelet decomposition), which is more flexible and highlights convection at scales between two thresholds. We use these filters to spatially enhance common verification scores, such as the Brier score. We train each NN with a different SELF and compare their performance at many scales of convection, from discrete storm cells to tropical cyclones. Among our many findings are that (i) for a low or high risk threshold, the ideal SELF focuses on small or large scales, respectively; (ii) models trained with a pixelwise loss function perform surprisingly well; and (iii) nevertheless, models trained with a spectral filter produce much better-calibrated probabilities than a pixelwise model. We provide a general guide to using SELFs, including technical challenges and the final Python code, as well as demonstrating their use for the convection problem. To our knowledge this is the most in-depth guide to SELFs in the geosciences. Significance Statement Gridded predictions, in which a quantity is predicted at every pixel in space, should be verified with spatially aware methods rather than pixel by pixel. Neural networks (NN), which are often used for gridded prediction, are trained to minimize an error value called the loss function. NN loss functions in atmospheric science are almost always pixelwise, which causes the predictions to miss rare events and contain unrealistic spatial patterns. We use spatial filters to enhance NN loss functions, and we test our novel spatially enhanced loss functions (SELF) on thunderstorm prediction. We find that different SELFs work better for different scales (i.e., different-sized thunderstorm complexes) and that spectral filters, one of the two filter types, produce unexpectedly well calibrated thunderstorm probabilities.

ON THE POSSIBILITY OF USING MAPS OF METEOROLOGICAL EVENTS FROM RESEARCH CENTRE «PLANET» FOR STUDYING AND PREDICTING THUNDERSTORMS ON THE URAL TERRITORY

The diagnosis and prediction of convective events are a difficult and challenging process. One of the most promising methods used for such tasks are hydrodynamic modelling and remote sensing. The aim of this research is to make a comparative analysis of meteorological events maps and data from observational network to find the possibility of using the maps from research center «Planet» for studying and predicting thunderstorms on the territory of the Ural. The spatiotemporal features of thunderstorms were identified. The repeatability and success of detection of this event were also determined, depending on the barium field, synoptic situation and specific characteristics of thunderstorms. During the study it was found that about 20 % of thunderstorms observed by weather stations were not predicted on weather maps; the high probability prediction was better than the medium and low probability forecasts of thunderstorms; spatial and temporal shifts were common in the prediction of thunderstorms. For a more detailed and objective analysis, it is recommended to work with a larger sample and use data from other sources like meteorological radiolocators, stormscopes, and others.

Thunderstorm Prediction During Pre-Tactical Air-Traffic-Flow Management Using Convolutional Neural Networks

Thunderstorm Prediction During Pre-Tactical Air-Traffic-Flow Management Using Convolutional Neural Networks

Prediction of Thunderstorms based on Atmospheric Instability Indices over Bangladesh using WRF-ARW Model

In this study an attempt has been made to inspect the forecasting of thunderstorms based on two cases (1st case: 17th May, 2019 and 2nd case: 31st March, 2019) over Dhaka using WRF Model. The model is run for 72 hours with 03 nested domain of 09 km, 03 km and 01 km horizontal resolutions using 0.25º X 0.25º six hourly global data assimilation system. For model simulation, Milbrandt-Yau Double-Moment 7-class scheme (9) has been used as microphysics scheme in this study. The model performance is evaluated by calculating hourly instability indices (VTI, TTI, KI, CTI, MCAPE, MCIN, BRN, LI, SI, SWI) value and have been compared with the threshold value of indices. Different meteorological parameters such as MSLP, temperature, winds at upper (300 hPa) and lower (925 hPa) level, relative humidity along with vertical cross section are also studied by the model and compared with the favorable conditions for forming of thunderstorms. Area rage rainfall (hourly) value has been also calculated and compared with indices value to comprehend the nature of thunderstorms. Observing the indices value it is seen that all indices value increase sharply 5-6 hours before of thunderstorm occurring and MCAPE is giving more reliable result. Moreover, this study shows that inner two domains (3 and 1 km resolution) are giving better results than outer one and which indices are more probable in forecasting of thunderstorm for our country as well as giving less Root Mean square Error. From the simulated and validated results, it can be concluded that the model performance of instability indices can be used as forecasting of thunderstorms over Bangladesh.

Statistical and dynamical based thunderstorm prediction over southeast India

Thunderstorms, associated with lightning and heavy rain, are a weather hazard causing human deaths, urban floods and damage to crops. Current work attempted to study the thunderstorms over Andhra Pradesh, coastal state in southeast India, using multiple satellite datasets, gridded rainfall, Doppler Radar Images and Advanced Research Weather Research and Forecasting (ARW) model simulations during the pre-monsoon seasons of 2017 and 2018. Thermodynamic stability indices computed using INSAT-3D/3DR satellite data were used to identify precursors and lead time of prediction. India Meteorological Department (IMD) daily gridded rainfall data were used to identify the thunderstorm occurrence days, and Doppler Radar Images and INSAT imagery were conjointly used to fix the location. Eight severe thunderstorm cases were analyzed to assess the precursors and the predictability. Further, ARW model predictions for two thunderstorm cases were performed and stability indices computed using model output were compared with satellite-based indices for evaluation. Statistical metrics had shown good agreement of ARW model-based stability indices with satellite-based stability indices. Model had simulated rainfall and cloud properties associated with thunderstorm activity. The results illustrated the predictability of the location and intensity of thunderstorms with 3–4 hrs lead time, which would find usefulness in the real-time prediction of thunderstorms.