Flood Disasters and the Development of Early Warning Model for Railway

Cryptosporidium spp. cause enteric diseases in a wide range of animals, including dairy cattle. However, limited information is available regarding prevalence and molecular characterization of Cryptosporidium spp. in dairy cattle in Gansu province and Ningxia Hui Autonomous Region (NXHAR), northwest China. A total of 2945 dairy feces samples (1257 from Gansu province and 1688 from NXHAR) were collected between December 2012 and March 2014 and were tested by PCR amplification of the small subunit (SSU) rRNA gene. A total of 150 (5.09%, 58 from Gansu and 92 from NXHAR) samples were PCR-positive for Cryptosporidium, and the prevalence is associated with the region and age of dairy cattle. Species identification showed Cryptosporidium andersoni in 36 samples (24.00%, 19 from NXHAR and 17 from Gansu), Cryptosporidium ryanae in 24 samples (16.00%, 13 from NXHAR and 11 from Gansu), Cryptosporidium bovis in 70 samples (46.67%, 41 from NXHAR and 29 from Gansu), and Cryptosporidium parvum in 20 samples (13.33%, 19 from NXHAR and 1 from Gansu). A DNA sequence analysis of the gp60 gene suggested that all the 20 C. parvum isolates represented subtype IIdA15G1. These findings indicated the presence of zoonotic Cryptosporidium in Gansu and NXHAR. This is the first report of four species of Cryptosporidium (C. andersoni, C. ryanae, C. bovis, and C. parvum) infection in dairy cattle in Gansu province. This is also the first report of C. ryanae infection in dairy cattle in NXHAR. Effective control strategies should be implemented to prevent and control Cryptosporidium infection in dairy cattle and humans.

More than 240 items of historical records containing climatic information were re- trieved from official historical books, local chronicles, annals and regional meteorological disaster yearbooks. By using moisture index and flood/drought (F/D) index obtained from the above information, the historical climate change, namely wet-dry conditions in borderland of Shaanxi Province, Gansu Province and Ningxia Hui Autonomous Region (BSGN, mainly in- cluded Ningxialu, Hezhoulu, Gongchanglu, Fengyuanlu and Yan'anlu in the Yuan Dynasty) was studied. The results showed that the climate of the region was generally dry and the ratio between drought and flood disasters was 85/38 during the period of 1208-1369. According to the frequencies of drought-flood disasters, the whole period could be divided into three phases. (1) 1208-1240: drought dominated the phase with occasional flood disasters. (2) 1240-1320: long-time drought disasters and extreme drought events happened frequently. (3) 1320-1369: drought disasters were less severe when flood and drought disasters happened alternately. Besides, the reconstructed wet-dry change curve revealed obvious transition and periodicity in the Mongol-Yuan Period. The transitions occurred in 1230 and 1325. The wet-dry change revealed 10- and 23-year quasi-periods which were consistent with solar cycles, indicating that solar activity had affected the wet-dry conditions of the study region in the Mongol-Yuan Period. The reconstructed results were consistent with two other study results reconstructed from natural evidences, and were similar to another study results from historical documents. All the above results showed that the climate in BSGN was character- ized by long-time dry condition with frequent severe drought disasters during 1258 to 1308. Thus, these aspects of climatic changes might have profound impacts on local vegetation and socio-economic system.

Tree-line contradiction refers to the conflict between planting trees in the power line erection space and ensuring the safe and stable operation of power lines. In severe convective weather, the tree-line contradiction is further intensified. Once the tree-barrier grounding fault is triggered, the power supply reliability of the rural distribution network will be seriously affected. Aiming at this, a zonal early warning mechanism for tree-line contradiction of rural distribution networks considering severe convective weather is proposed in this paper. Firstly, due to the imbalance of tree-barrier grounding fault records, the Synthetic Minority Oversampling Technique (SMOTE) algorithm is used to replace part of the majority class samples with minority class samples, and the data preprocessing is realized based on keeping the scale of the data set unchanged. Secondly, combined with six meteorological monitoring indexes of severe convective weather, a mapping model of severe convective weather and tree barrier grounding risk is established based on Extreme Learning Machine (ELM). Finally, the data of a rural distribution network in China is used for example analysis. The results show that the proposed model can effectively reflect the mapping relationship between severe convective weather and tree-barrier grounding faults, and accurately realize the zonal warning of tree-barrier grounding risk, which has good robustness and scalability.

Severe convective weather, characterized by short-term intense precipitation, thunderstorms, and strong winds, poses significant threats to human life and property. Therefore, accurate and efficient prediction of severe convective weather is crucial for disaster prevention. Currently, utilizing deep learning for radar echo extrapolation stands as the primary method for forecasting severe convective weather. We propose a predictive recurrent neural network model that integrates a self-attention mechanism, specifically designed for radar echo extrapolation in severe convective weather forecasting. The self-attention mechanism offers the advantage of being lightweight, as it does not substantially increase the model parameters. Additionally, it facilitates global attention extraction, thereby enhancing the model’s accuracy to some extent. By utilizing radar echo images from the previous hour as input, the model undergoes self-learning to achieve the best forecast for radar echo extrapolation in the subsequent two hours. Research findings demonstrate that our model outperforms other models in accurately predicting severe convective weather within this two-hour timeframe. Received: 30 November 2023| Revised: 19 April 2024 | Accepted: 16 May 2024 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement The data used in this study is a competition dataset, which is not an open source at this time. It is available from the corresponding author upon reasonable request. Author Contribution Statement Qiongying Xue: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Writing - original draft, Writing - review & editing, Supervision, Project administration. Fei Fang: Conceptualization, Methodology, Investigation, Resources, Data curation, Writing - original draft, Writing - review & editing, Supervision, Project administration, Funding acquisition.

The strongest El Nino event that has been ever seen since observation records started in 1951 occurred in 2015/2016. This paper analyzes the characteristics of the severe convective weather and extreme rainfalls that were experienced from January to August, 2016, which are then compared to those found for the corresponding periods since 1981. The findings are summarized as follows. (1) Severe convective weather, especially thunderstorm gales and short-time rainstorms, occurred much more in 2016 than normal, resulting in the most serious severe convective disasters in the recent years and the most extreme precipitation events since 2000. (2) The interannual variation of the severe convective weather and extreme rainfalls since 1981 reveals that in the years (1982/1983, 1997/1998, 2015/2016) when super El Nino events ended the extreme rainfalls and severe convective weather occurred much more than those during the normal El Nino years, when there are not very obvious features of high frequency of such weather events. (3) In 1983, thunderstorm gales and hails occurred frequently, but the extreme rainfall days were similar with those in normal years except in the Yangtze River valley. (4) In 1998, extreme precipitation days over the whole country were the most since 1981, and the frequency of short-time heavy precipitation ( ³ 20 mm/h) was only second to that in 2016 while the disaster-inducing short-time rainstorms ( ³ 50 mm/h) was the most since 1981. (5) In 2016, severe convective weather was mainly reflected by frequent thunderstorm gales and short-time rainstorms, which were more than the climatic averages over the Yangtze River valley and its south regions. Especially, the fierce thunderstorm gales stronger than 25 m/s were the most in the recent five years, about 3–4 times of those in the other years. However, hail events were still in the decreasing trend as the result of the remarkable decrease in hail over the northern part of China. Comparatively, the days of extreme rainfalls in North China were the second most since 1981, of which the nationwide rainfall amount in the 18–20 July severe precipitation event with the North China region as the heavy rainfall center exceeded the precipitation in the corresponding period in history. (6) In the years when the three super El Nino events ended, extreme rainfalls occurred more frequently than normal, but difference existed in different regions. In 2016, the frequency ratio of extreme rainfalls was the highest since 2000 no matter whether it was for the nationwide, or the Yangtze River valley and North China. The extreme rainfall days that appeared nationwide and the Yangtze River valley in 1998 was the highest since 1981 while such days in 1983 were found more only in the Yangtze River valley.

In this study, we propose a model called CNN-BiLSTM-AM that utilizes deep learning techniques to forecast severe convective weather based on ERA5 hourly data and observations. The model integrates a CNN with a Bidirectional Long Short-Term Memory (BiLSTM) system and an Attention Mechanism (AM). The CNN is tasked with extracting features from the input data, while the BiLSTM effectively captures temporal dependencies. The AM enhances the results by considering the impact of past feature states on severe weather phenomena. Additionally, we assess the performance of our model in comparison to traditional network architectures, including ConvLSTM, Predrnn++, CNN, FC-LSTM, and LSTM. Our results indicate that the CNN-BiLSTM-AM model exhibits superior accuracy in precipitation forecasting. Especially with the extension of the forecast time, the model performs well across multiple evaluation metrics. Furthermore, an interpretability analysis of the convective weather mechanisms utilizing machine learning highlights the critical role of total precipitable water (PWAT) in short-term heavy precipitation forecasts. It also emphasizes the significant impact of regional variables on convective weather patterns and the role of convective available potential energy (CAPE) in fostering conditions conducive to convection. These findings not only confirm the effectiveness of deep learning in the automatic identification of severe weather features but also validate the suitability of the sample dataset employed. Given its remarkable performance and robustness, we advocate for the adoption of this model to enhance the forecast of severe convective weather across various business applications.

Prevalence of antibodies to Toxoplasma gondii and risk factors with infection were assessed in dairy cattle from Gansu Province and Ningxia Hui Autonomous Region (NXHAR), northwest China. In total, 1657 serum samples were collected and assayed by the modified agglutination test. The overall seroprevalence was 4.83% at a 1:100 cut-off, with titers of 1:100 in 72, 1:200 in 4, 1:400 in 4. Among the risk factors examined, no statistically significant difference was observed between T. gondii seroprevalence and regions or age of dairy cattle in the logistic regression analysis (P>0.05) and left out of the final model. However, numbers of pregnancies of dairy cattle was considered as main risk factor associated with T. gondii infection. Dairy cattle in nulliparity group (8.89%) had 6 times (OR=6.31, 95% CI, 2.69-14.83, P<0.001) higher risk of being seropositive compared to dairy cattle in 3 or above 3 pregnancies group (1.52%), followed by 1 pregnancy group (4.27%) had nearly 3 times (OR=2.89, 95% CI, 1.11-7.52, P = 0.03) higher risk of being seropositive compared to dairy cattle in 3 or above 3 pregnancies group, although no statistical difference was found between 2 pregnancies group and 3 or above 3 pregnancies group (P = 0.70). The results of this survey indicated the presence of T. gondii infection in dairy cattle in Gansu Province and NXHAR, which enriches the epidemiological data of T. gondii infection in dairy cattle in China, and is helpful to strengthen prevention and control of T. gondii infection in dairy cattle in these two regions.

In this study, we investigate the most severe East Asian dust storm in the past decade that occurred on 14–16 March 2021 based on the Weather Research and Forecasting model coupled with chemistry (WRF-Chem) and a variety of site measurements and satellite retrievals. The dust emissions from the Gobi Desert, especially over Mongolia on March 14, are the dominant sources of this intense dust event. The maximal hourly accumulated dust emissions over Mongolian and Chinese areas reached 1490.18 kt at 07:00 UTC on 14 March and 821.70 kt at 2:00 UTC on 15 March, respectively. During this dust event, the accumulated dust emissions in coarse modes (i.e., bin 4 and bin 5) account for 64.1% of the total dust emission mass, and the accumulated dust emissions in fine modes (i.e., bin 1) are the least, accounting for 7.6% of the total dust emission mass. Because the coarse mode bins of dust dominate the emissions, the downwind transported coarse mode particles can affect the North China Plain, while the fine particles can only affect the desert source and its surrounding regions such as the Gansu and Ningxia provinces. Due to the dust emissions and the dust transport path, the high AOD areas are located in the Gobi Desert and Northwest China and the vertical spatial distributions of aerosol extinction coefficients have the same characteristics. We also found the model drawback of overestimating simulated wind speeds, which leads to the overestimations of dust emissions and concentrations, indicating the urgency of improving the simulated wind field.

Severe convective weather (SCW) forecasting has emerged as a cutting-edge research focus due to its complex dynamical characteristics and significant socioeconomic impacts. However, the advent of deep learning (DL), with its capacity for extracting massive information and modeling non-linear relationships, offers a promising alternative for enhancing the accuracy of SCW forecasts. This study reviews advances in DL applications to SCW forecasting. It begins by summarizing current mainstream DL techniques and analyzes the similarities between SCW tasks and typical DL problems. Subsequently, it surveys the utilization in forecasting four types of SCW events: rainstorms, hail, thunderstorm winds and tornadoes, while identifying existing challenges and outlines future directions. DL effectively improved SCW forecasting accuracy and forecast lead times for integrating multi-source observational data and analyzing both observational and numerical weather prediction (NWP) datasets. Current challenges include lack of physical constraints, weak interpretability and insufficient high-quality samples. Future efforts will leverage higher-resolution and fully-integrated multi-source observation data to deepen the understanding of the small-scale structure and stage characteristics of SCW. Physical-based mechanism understanding, numerical prediction, and DL technologies are being continuously advanced in an integrated manner, aiming to build cognitively capable forecasting foundation models that can better support forecasters’ operational expertise and drive SCW forecasting toward greater intelligence and interpretability.

Convective weather, often associated with heavy precipitation, hail, lightning, and other hazardous phenomena, is highly unpredictable, short-lived, and localized, making forecasting and early warning particularly challenging. The formation of lightning is closely tied to the thermodynamic and microphysical processes within severe convective weather systems (e.g., Qie et al., 2021). Not only does it pose a significant threat to human life and properties, but it has also been recognized by the International Electrotechnical Commission (IEC) as a major hazard to power systems, communication networks, buildings, and electronic devices.Since the mid-20th century, Doppler weather radars have been widely used to monitor hazardous weather by identifying precipitation, storm structures, and movement. Advances in radar technology, especially the introduction of array weather radar, have further enhanced the precision and timeliness of severe weather nowcasting. Unlike traditional single-antenna radars, array radars use multiple small antennas to form a large, flexible antenna array for rapid and precise beam control. This distributed phased-array system excels in detecting fine-scale flow and intensity fields, offering powerful tools for studying small-scale convective phenomena (e.g., Adachi et al., 2016).This study utilizes array radar data from Foshan, Guangdong, China, high-precision lightning location data, and ground-based meteorological observation data to identify, track, and forecast severe convective weather. Based on a radar dual-threshold convective storm tracking and identification algorithm (e.g., Tian et al., 2019), combined with a lightning jump algorithm (e.g., Schultz et al., 2017), this nowcasting method monitors the lightning variation characteristics within strong convective cells (CCs), providing indices for severe convective weather. By comparing results with observations and optimizing algorithm parameters, the method improves hit rates, reduces false alarms, and achieves an average lead time of ~22 minutes with a hit rate over 80%, as demonstrated by case studies. This method can be effectively applied to enhance the monitoring and early warning capabilities for severe convective weather, thereby mitigating the impact of lightning and reducing lightning-related disasters for critical infrastructure, particularly power systems.&amp;#160;AcknowledgmentThis work was jointly supported by the KERAUNIC project (ref: NIA2_NGET0055, National Grid Electricity Transmission, 2024) under the Network Innovation Allowance (NIA), the Arctic Pavilion Open Research Fund of Nanjing Joint Institute for Atmospheric Sciences under Grant BJG202410 and the China Scholarship Council program under Grant 202305330027.&amp;#160;ReferencesAdachi, T., Kusunoki, K., Yoshida, S., et al. (2016). High-speed volumetric observation of a wet microburst using X-band phased array weather radar in Japan.&amp;#160;Monthly Weather Review,&amp;#160;144(10), 3749-3765.National Grid Electricity Transmission. (2024). Knowledge Elicitation of Risks to Assets Under LightNing Impulse Conditions (KERAUnIC). https://smarter.energynetworks.org/projects/nia2_nget0055Qie, X., Yuan, S., Chen, Z., et al. (2021). Understanding the dynamical-microphysical-electrical processes associated with severe thunderstorms over the Beijing metropolitan region. Science China Earth Sciences, 64, 10-26.Schultz, C. J., Carey, L. D., Schultz, E. V., &amp;amp; Blakeslee, R. J. (2017). Kinematic and microphysical significance of lightning jumps versus nonjump increases in total flash rate.&amp;#160;Weather and forecasting,&amp;#160;32(1), 275-288.Tian, Y., Qie, X., Sun, Y., et al. (2019). Total lightning signatures of thunderstorms and lightning jumps in hailfall nowcasting in the Beijing area.&amp;#160;Atmospheric Research,&amp;#160;230, 104646.

Severe convective weather can lead to a variety of disasters, but they are still difficult to be pre-warned and forecasted in the meteorological operation. This study generates a model based on the light gradient boosting machine (LightGBM) algorithm using C-band radar echo products and ground observations, to identify and classify three major types of severe convective weather (i.e., hail, short-term heavy rain (STHR), convective gust (CG)). The model evaluations show the LightGBM model performs well in the training set (2011-2017) and the testing set (2018) with the overall false identification ratio (FIR) of only 4.9% and 7.0%, respectively. Furthermore, the average probability of detection (POD), critical success index (CSI) and false alarm ratio (FAR) for the three types of severe convective weather in two sample sets are over 85%, 65% and lower than 30%, respectively. The LightGBM model and the storm cell identification and tracking (SCIT) product are then used to forecast the severe convective weather 15 - 60 minutes in advance. The average POD, CSI and FAR for the forecasts of the three types of severe convective weather are 57.4%, 54.7% and 38.4%, respectively, which are significantly higher than those of the manual work. Among the three types of severe convective weather, the STHR has the highest POD and CSI and the lowest FAR, while the skill scores for the hail and CG are similar. Therefore, the LightGBM model constructed in this paper is able to identify, classify and forecast the three major types of severe convective weather automatically with relatively high accuracy, and has a broad application prospect in the future automatic meteorological operation.

Based on observational data, ERA5 hourly reanalysis data, and s‐band dual‐polarization radar data, the extremely severe convective weather happening in Shandong Peninsula on 1 October 2021 is analyzed. The results show that the process was affected by the upper‐level cold vortex, and extremely severe convective weather happened from west to east in the northern area of Shandong Peninsula. Extra‐large hail occurred in this process. The upper‐level cold vortex provided the synoptic‐scale dynamic forcing and unstable thermal stratification conditions, which were conducive to the occurrence of severe convective weather. About 2400 J · kg−1 sufficient convective available potential energy (CAPE) and 30.2 m · s−1 strong vertical wind shear between 0 and 6 km altitude prompted this severe convective weather to arise in October. Due to the special coastal geographical environment of Shandong Peninsula, the surface sea‐breeze front triggered this severe convection process, and the movement of the surface convergence line stimulated the occurrence and development of convection nearby upstream and downstream. The three‐body scatter spike (TBSS) and bounded weak echo region (BWER) of the storm were very significant. The value of ZDR is lower below 0°C layer than that above 0°C layer; meanwhile, the values of CC and KDP are higher below 0°C layer than that above 0°C layer. There were two ZDR columns on the east and west sides in the supercell storm, and the height (HT) of the east ZDR column top exceeds that of the −20°C layer. The thickness of the east ZDR column (about 3.2 km) was thicker than that of the west ZDR column (about 2.2 km). These all indicate that the storm had a deep updraft, which was conducive to the formation and growth of hail. There were also two KDP columns on the east and west sides of the BWER, which had almost the same thickness about 3.2 km, and both extended beyond the HT of the −20°C layer. The thickness of the strong reflectivity above 65 dBZ reaches 7 km, and the thickness of the ZDR column and KDP column reaches 2–3 km can be used as a reference for predicting extra‐large hail in autumn short‐term forecasting.

Severe convective weather events—tornadoes, hailstorms, high winds, flash floods—are inherently mesoscale phenomena. While the large-scale flow establishes environmental conditions favorable for severe weather, processes on the mesoscale initiate such storms, affect their evolution, and influence their environment. A rich variety of mesocale processes are involved in severe weather, ranging from environmental preconditioning to storm initiation to feedback of convection on the environment. In the space available, it is not possible to treat all of these processes in detail. Rather, we will introduce several general classifications of mesoscale processes relating to severe weather and give illustrative examples. Although processes on the mesoscale are often intimately linked with those on smaller and larger scales, we will exclude from discussion those that obviously lie outside the mesoscale domain (e.g., baroclinic waves on the synoptic scale or charge separation in clouds on the microscale).

Convective-scale transport of mineral dust in a severe weather setting is investigated with the approach of three-dimensional cloud-resolving simulations coupled with a dust emission-transport modeling. The simulations are intended to explicitly represent convective- and cloud-scale processes (such as updraft/downdraft, surface cold pool, precipitation) in a squall-line-type convective system, and are performed in an idealized setup in order to focus the primary mechanisms for convective-scale transport of dust within a squall-line system. Initialized based on an observation in a severe duststorm case over the Gobi Desert in China, the cloud model well simulates an observational feature of the squall line and the associated duststorm in spite of a simplified model setup.Dust is emitted by strong surface winds associated with a well-developed surface cold pool, and is contained and mixed within the cold pool: a high dust concentration of greater than 10 mg m−3 is induced. Owing to a high subgrid-turbulence mixing at the leading edge of the cold pool, the contained dust is transferred out of the cold pool and is entrained into the updraft region at the cold pool edge. Dust is then transported upward by the convective updraft which is continuously regenerated at the cold-pool leading edge, and spreads laterally in the cross-line directions at upper levels by system-scale circulation. Rearward dust transport relative to the leading edge of the system is pronounced at upper levels, according to the prevalent front-to-rear flow typically found in the squall-line systems. This study suggests that the representations of convective-scale transport processes should be adequately updated in order to improve the accuracy of the regional-scale to global-scale numerical predictions.

Based on meteorological observations and products of a GRAPES and an ECMWF model from March to April 2014, some indexes and parameters with good relevancy were selected as predictors. Through analyzing the spatial distributions and the binary logistic regressions of the indexes, estimated values of the predictors and severe convective weather diagnostic prediction equations were established to get a severe weather predictor P for forecasting severe convective weather for the next 12 hours in Guangdong province. The equations were tested and analyzed, respectively, with the two models as well as the radiosonde data. The results indicated that the severe weather forecasts’ CSI by the predictor P was obviously higher than by any single index. The TT error between the models and the soundings was small, while the K index of the models was more discrete than the soundings. The index MDPIs were 1 greater than the soundings, but their trends of change were consistent with the soundings.