Figure - available from: Journal of Geophysical Research: Atmospheres
This content is subject to copyright. Terms and conditions apply.
Ten features of storm cells. (a) The features shown are based on the height. (b) The features shown are based on different reflectivity and volume values. (c) The features shown are based on the overhang structure and the gradient.
Source publication
This study investigates the characteristics of hail storms and cumulonimbus storms in China from 2005 to 2016. Ten features are proposed to identify storm cells that can produce hail, especially in the early stage of hail formation. These features describe hail storms based on three factors: the height and thickness of the cell core, the radar echo...
Citations
... Relying solely on a single feature parameter for hail identification can easily introduce biases. Therefore, the adoption of multiple indicators concurrently in the identification process is an inevitable trend in its development [18][19][20]. Lopez et al. [21] combined VIL, maximum reflectivity, and height of the maximum reflectivity to construct a logistic model, which yielded promising test results in probability form. Blair et al. [22] also emphasized the significance of echo heights at 50 dBZ and 60 dBZ in the identification of large hailstones. ...
... Leveraging commonly used feature construction methods in the field of image processing alongside dual-polarization radar data, this study establishes pixel-level features of hail cells. Hypothesis testing is utilized for validity analysis [24], while PCA and fisher linear discriminant analysis are applied for feature synthesis [20], ultimately providing a novel five-dimensional representation of the mechanism features of hail clouds. During experimental validation, a hail identification model was developed using the support vector machine (SVM) machine learning approach [1,24,30], confirming the effectiveness of the new features. ...
... Furthermore, the use of the 10-dimensional PCA comprehensive feature scheme is more effective than directly building models in the 490-dimensional feature space. This aligns with our expectations and previous research findings [19,20,42]; (2) The Fisher feature comprehensive scheme, overall, outperforms the PCA feature comprehensive scheme. This is because PCA's "maximum variance criterion" may not necessarily align with the classification objective, while Fisher's "intra-class cohesion, inter-class separation [52]" criterion aligns with the goals of a classification function. ...
Hail is a type of severe convective weather disaster characterized by abundant water vapor and strong updrafts, resulting in intense and high reflectivity echoes in hail clouds, often accompanied by an overhanging form. Although hail research has made great progress, it is still challenging to achieve accurate identification of hail. Compared with traditional radar, dual-polarization radar can output a variety of polarization parameters and provide information about the shape and phase of precipitation particles, which is conducive to the identification of hail particles. In this study, dual-polarization radar data are used to explore more hail features from various perspectives, starting with the morphological characteristics of hail clouds and using common feature extraction methods in the field of image processing. A comprehensive approach to high-dimensional features is developed. Using machine learning methods, hail identification models are constructed in both the traditional mechanism feature space and the new feature space constructed in this study. Experimental results strongly confirm the significant effectiveness of the five-dimensional new mechanism features developed in this paper for hail identification.
... These incidents have detrimental effects on the economy of predominantly agrarian countries (Timothy et al. 2021;Willemse 1995;Zhang et al. 2008), restricting production (Vogel et al. 2017;Lyubchich 2019;Berthet et al. 2011) and the yield (Eccel et al. 2012;McMaster 1999;Sanchez 1996;Mohr et al. 2015;Childs et al. 2020). The phenomenon is widespread around the world and has become a potential threat (Munich 2016;Saltikoff et al. 2010;Nisi et al. 2016;Wang et al. 2018;Ortega 2018). Hailstorms have been significantly damaging crops and infrastructure resulting in huge losses (Brian et al. 2019;Gunturi and Tippett 2017;Punge et al. 2017;Smith 2020). ...
... In the field of meteorology, radar is a very effective piece of equipment [1][2][3][4][5]. It can provide data including reflectivity intensity, spectral width and velocity, which can provide powerful data support for weather prediction [6][7][8][9][10]. ...
Radar reflectivity images have the potential to provide vital information on the development of convective cloud interiors, which can play a critical role in precipitation prediction. However, traditional prediction methods face challenges in preserving the high-frequency component, leading to blurred prediction results. To address this issue and accurately estimate radar reflectivity intensity, this paper proposes a novel reflectivity image prediction approach based on the Spatial Memory in Memory (Spatial MIM) networks and the Pix2Pix networks. Firstly, a rough radar reflectivity image prediction is made using the Spatial MIM network. Secondly, the prediction results from the Spatial MIM network are fed into the Pix2pix network, which improves the high-frequency component of the predicted image and solves the image blurring issue. Finally, the proposed approach is evaluated using data from Oklahoma in the United States during the second and third quarters of 2021. The experimental results demonstrate that the proposed approach yields more accurate radar reflectivity prediction images.
... These incidents have detrimental effects on the economy of predominantly agrarian countries (Timothy et al. 2021;Willemse 1995;Zhang et al. 2008), restricting production (Vogel et al. 2017;Lyubchich 2019;Berthet et al. 2011) and the yield (Eccel et al. 2012;McMaster 1999;Sanchez 1996;Mohr et al. 2015;Childs et al. 2020). The phenomenon is widespread around the world and has become a potential threat (Munich 2016;Saltikoff et al. 2010;Nisi et al. 2016;Wang et al. 2018;Ortega 2018). Hailstorms have been significantly damaging crops and infrastructure resulting in huge losses (Brian et al. 2019;Gunturi and Tippett 2017;Punge et al. 2017;Smith 2020). ...
The North-Western Himalayas are becoming the hotspots for hydro-meteorological catastrophes due to climate change implications. Present study aims to analyse spatio-temporal dynamics of hailstorms, posing a threat to livelihood security in the Valley of Kashmir on account of significant land use shift from agriculture to horticulture which is highly susceptible to hail hazard. The mean maximum and mean minimum temperature and annual rainfall for Qazigund, Kupwara and Srinagar weather stations were analysed through statistical treatments like Mann Kendall and Sen’s slope estimator. To establish the land use change scenario, area under horticulture and under Apple was calculated and demonstrated with the help of regression analysis. Hailstorm incidents were calculated through the archival newspapers from 2007 to 2022 and mapped spatio-temporally in GIS environment. Additionally, to forecast the near-term hail scenario, a trendline was established by using Linear Regression Equation for a period of 2022–2040. Furthermore, a 4 Point Likert scale survey for evaluating the perception of people regarding the changing climatic scenarios and intensification of hailstorm activity was carried out throughout the valley. A total of 203 hailstorm events have occurred from year 2007 to 2022. The findings reveal that the hail storm intensity and frequency has increased with the corresponding increase in temperature and decline in rainfall on account of changing Climatic scenarios across the Kashmir Valley. There has been an increase in hailstorms from 2 events in the year 2007 to 27 events in 2022. North-Kashmir districts, namely Baramulla and Kupwara are emerging as the hail storm hot spots during the spring seasons due to their location around the entrance corridor of western disturbance. The southern districts of Pulwama, Shopian and Anantnag have witnessed enhanced summer hailstorm activity probably under the influence of south-west monsoons. The forecast indicated a significant increase in the hailstorm occurrence with an R² value of 0.83. The study also concludes that hailstorms are responsible for (30–70%) loss in the productivity of horticulture in the affected areas, thereby threatening livelihood of millions of farmers. Furthermore, the study indicates that the region does not have adequate adaptation and mitigation strategies in place as only 0.06% of the horticulturists are having Anti-hail net protection while as the crop insurance-cover is almost non-existent. The study shall be helpful in developing effective mitigation and adaptation strategies to combat the hail hazard for securing livelihoods by promoting sustainable horticulture in the region.
... Hailstorms and their accompanying hail, gales, short-term heavy rainfall, and tornadoes pose huge threats to human life and property. Weather radar plays an important role in observing hailstorms and the early warning of these severe weather events [1][2][3][4][5][6]. Storm identification and tracking algorithms, which use single-polarization radar observations to identify storms, calculate structural storm characteristics-such as centroid position, vertically integrated liquid water (VIL), maximum reflectivity, top height, area, and so on-and track and forecast the location and strength of storms, constitute a key part of severe weather warning operations [2,3]. ...
... Storm identification and tracking algorithms, which use single-polarization radar observations to identify storms, calculate structural storm characteristics-such as centroid position, vertically integrated liquid water (VIL), maximum reflectivity, top height, area, and so on-and track and forecast the location and strength of storms, constitute a key part of severe weather warning operations [2,3]. On the one hand, these output storm structural characteristics can also serve as input to other reflectivity-based severe weather detection algorithms, such as hail detection algorithms [4][5][6]. On the other hand, the algorithms that provide the temporal trends of structural storm characteristics are also suitable for studying the physical mechanisms of storm evolution and, thus, provide an important tool for nowcasting [2,3]. However, better hail nowcasting also requires the inclusion of microphysical (e.g., hail and graupel) and dynamical (e.g., mesocyclones) characteristics of storms. ...
To further enhance the application of dual-polarization radar in hail nowcasting, we develop an integrated convective characteristic extraction (ICCE) algorithm based on the storm cell identification and tracking (SCIT) algorithm using dual-polarization radar data and its secondary product (hydrometeor classification data and mesocyclone data). The ICCE identifies and tracks not storm cells but convective systems, and it adds other storm characteristics, such as storm microphysics (hail- and graupel-related) and storm dynamics (mesocyclone-related), to the original storm characteristics, such as storm structure (reflectivity-related) and storm tracking (motion-related). The data of four mesocyclonic hailstorms observed by the two S-band dual-polarization radars in Guangdong Province, China, are utilized, from which we draw the following conclusions: (1) ICCE excels in identifying, characterizing, matching, and tracking convective systems; and (2) the newly added storm microphysics and dynamics characteristics can more accurately quantify the relationship between mesocyclone development, hail growth, and convective system enhancement throughout the evolution of the convective system.
... Weather radar, as an important remote sensing technique, has also been used to investigate and predict the climatology and characteristics of hailstorms. With S-band radar data, Wang et al. (2018) investigated the characteristics of hail storms and cumulonimbus storms in China from 2005 to 2016. ...
Hail hazards have caused severe losses and threatened the safety of residents in Beijing, the Chinese capital city. A refined analysis of the spatial and temporal distribution of hailstorms in Beijing can help to evaluate the risk of hail and guide the operations of hail prevention. The distribution of weather stations is relatively sparse to figure out the fine distribution of hail events. Therefore, a quality-controlled dataset of disaster information data reported from information reporters is used to analyze the fine temporal and spatial distributions of hail days and events in Beijing from 1980 to 2021 in this study. Hail events and hail days show an obvious downward trend with years from 1981 to 2010, while hail events show a strong upward trend from 2011 to 2021. The seasonal pattern of hail events shows a unimodal distribution from March to October, and the peak appears in June. Most of the hail events occurred from 14:00 to 21:00, while the highest counts appeared from 15:00 to 17:00. More and larger hails occurred in the northwestern mountains rather than southeastern plains in Beijing, highly correlated with the topography. Both total and severe hails hit the mountain area statistically earlier than the plain area. The most frequent hours of hails in the northwestern area and southeastern area were concentrated in the range of 13:00—17:00 and 16:00—20:00 CST, respectively. This time delay is due to the initiation location and movement direction of the convective storms. The influence of ENSO on warm season hails is positive in Beijing, which has a lag of 3 months or longer. The arctic oscillation has a negative correlation with hail days in each month from May to September.
... In recent years, machine learning (ML) methods have been widely used in various fields of meteorology and have achieved remarkable results [2], [29]- [34]. ML methods could effectively represent complex nonlinear meteorological processes from the statistical view by modeling large historical data. ...
... The reason for this "blur" could be explained from two perspectives: (1) MSE is sensitive to outliers, which makes the model more inclined to mean prediction [43], [44]. (2) The predictability of radar echo is related to the echo scale, and small-scale high-frequency echo details usually have low predictability [24]- [27], [45]. For the radar extrapolation task, this blur phenomenon will produce two adverse effects on extrapolations: (1) The reflectivity intensity of the extrapolations are usually underestimated. ...
... For the radar extrapolation task, this blur phenomenon will produce two adverse effects on extrapolations: (1) The reflectivity intensity of the extrapolations are usually underestimated. (2) The echo small-scale details in the extrapolations are lost, and the convective systems' structure cannot be represented correctly. These make it difficult for ML methods to provide sufficient information for hazard forecasting, especially for convective hazards. ...
Radar echo extrapolation is a basic but essential task in meteorological services. It could provide radar echo prediction results with high spatiotemporal resolution in a computationally efficient way, and effectively enhance the operational system's forecasting capability for meteorological hazards. Traditional methods perform extrapolation by estimating echo motions between contiguous radar data. This strategy is difficult to characterize complex nonlinear meteorological processes effectively, and it is difficult to benefit from large historical data. Recently, machine learning (ML) models have been used for radar echo extrapolation. These methods have effectively improved extrapolation quality in a data-driven way and from the statistical perspective. Although the ML-based methods show excellent performance, they usually produce blurry extrapolations. This leads to underestimating radar echo intensity and making echo lack small-scale details. Moreover, it makes models difficult to predict severe convective hazards. To solve this problem, a two-stage extrapolation model based on 3D Convolutional Neural Network (3D-CNN) and Conditional Generative Adversarial Network (CGAN) is proposed. These two models form the ‘'pre-extrapolation" and "post-processing" paradigm. The pre-extrapolation model is trained in the traditional way and performs rough extrapolation. The post-processing model uses the pre-extrapolation result as input and is trained with the adversarial strategy. It could correct the echo intensity and increase the echo’s details. In the experiment, our model could provide more precise radar echo extrapolations than other methods, especially for intense echoes and convective systems, in the data of North China from 2015 to 2016.
... Notable SP radar methods that have been shown to perform well are the vertically integrated liquid water (VIL) density (Amburn and Wolf 1997) and the maximum expected size of hail (MESH; Witt et al. 1998a), both of which are based on vertical integration of radar reflectivity at horizontal polarization Z H . These and other similar Z H -based methods have been cross evaluated and/or verified against hail reports in multiple studies during the past two decades (e.g., Edwards and Thompson 1998;Holleman et al. 2000;Marzban and Witt 2001;Ortega et al. 2005;San Ambrosio et al. 2007;Donavon and Jungbluth 2007;López and Sanchez 2009;Liu and Heckman 2010;Saltikoff et al. 2010;Cintineo et al. 2012;Skripniková and Rezá cová 2014; Nisi et al. 2016;Lukach et al. 2017;Capozzi et al. 2018;Wang et al. 2018;Ortega 2018). ...
Severe hail days account for the vast majority of severe weather–induced property losses in the United States each year. In the United States, real-time detection of severe storms is largely conducted using ground-based radar observations, mostly using the operational Next Generation Weather Radar network (NEXRAD), which provides three-dimensional information on the physics and dynamics of storms at ~5-min time intervals. Recent NEXRAD upgrades to higher resolution and to dual-polarization capabilities have provided improved hydrometeor discrimination in real time. New geostationary satellite platforms have also led to significant changes in observing capabilities over the United States beginning in 2016, with spatiotemporal resolution that is comparable to that of NEXRAD. Given these recent improvements, a thorough assessment of their ability to identify hailstorms and hail occurrence and to discriminate between hail sizes is needed. This study provides a comprehensive comparative analysis of existing observational radar and satellite products from more than 10 000 storms objectively identified via radar echo-top tracking and nearly 6000 hail reports during 30 recent severe weather days (2013–present). It is found that radar observations provide the most skillful discrimination between severe and nonsevere hailstorms and identification of individual hail occurrence. Single-polarization and dual-polarization radar observations perform similarly at these tasks, with the greatest skill found from combining both single- and dual-polarization metrics. In addition, revisions to the “maximum expected size of hail” (MESH) metric are proposed and are shown to improve spatiotemporal comparisons between reported hail sizes and radar-based estimates for the cases studied.
Hail, an intense convective catastrophic weather, is seriously hazardous to people’s lives and properties. This article proposes a multi-step cyclone hail weather recognition model, called long short-term memory (LSTM)-C3D, based on radar images, integrating attention mechanism and network voting optimization characteristics to achieve intelligent recognition and accurate classification of hailstorm weather based on long short-term memory networks. Based on radar echo data in the strong-echo region, LSTM-C3D can selectively fuse the long short-term time feature information of hail meteorological images and effectively focus on the significant features to achieve intelligent recognition of hail disaster weather. The meteorological scans of 11 Doppler weather radars deployed in various regions of the Hunan Province of China are used as the specific experimental and application objects for extensive validation and comparison experiments. The results show that the proposed method can realize the automatic extraction of radar reflectivity image features, and the accuracy of hail identification in the strong-echo region reaches 91.3%. It can also effectively realize the prediction of convective storm movement trends, laying the theoretical foundation for reducing the misjudgment of extreme disaster weather.
The magnitude of damage caused by hail depends on its size; however, direct observation or indirect estimation of hail size remains a significant challenge. One primary reason for estimations by proxy, such as through remote sensing methods, is that empirical relationships or statistical models established in one region may not apply to other areas. This study employs a machine learning method to build a hail size estimation model without assuming relations in advance. It uses FY-4A AGRI data to provide cloud-top information and ERA5 data to add vertical environment information. Before training the model, we conducted a principal component analysis (PCA) to analyze the highly influential factors on hail sizes. A total of 18 features, composed of four groups, namely brightness temperature (BT), the difference in BT (BTD), thermodynamics, and dynamics groups, were chosen from 29 original features. Dynamic and BTD features show superior performance in identifying large hail. Although the selected features are more closely correlated to hail sizes than unselected ones, the relationships are complicated and nonlinear. As a result, a two-layer regression back propagation neural network (BPNN) model with powerful fitting ability is trained with selected features to predict maximum hail diameter (MHD). The linear fitting R2 between predicted and observed MHDs is 0.52 on the test set, which signifies that our model performs well compared with other hail size estimation models. We also examine the model concerning all three hail cases in Shanghai, China, between 2019 and 2021. The model attained more satisfactory results than the radar-based maximum estimated hail size (MEHS) method, which overestimates the MHDs, thus further supporting the operational applications of our model.
