Hail Climate Characteristics and Influencing Factors in Longdong Region from 1978 to 2023

ZHANG Kexin, ZHAO Yujuan, LI Meiyu
(Qingyang Meteorological Bureau, Xifeng 745000, Gansu, China)

Abstract

Hail is an extreme weather phenomenon caused by intense convective activity, characterized by rapid development and substantial damage, particularly threatening to the Longdong region where agriculture constitutes the economic backbone. In the context of climate warming, it is imperative to investigate the climatic characteristics of hail in Longdong and identify its influencing factors. Utilizing hail observation and disaster records from 15 regional meteorological stations spanning from 1978 to 2023, along with ERA5 reanalysis data from the European Centre for Medium-Range Weather Forecasts, this study employs linear trend estimation, the Mann-Kendall test, and Morlet wavelet analysis to examine the spatial-temporal distribution of hail and its key drivers. The results indicate the following: (1) Hail days exhibit uneven spatial distribution, with higher frequencies in the northwest and southeast, and lower frequencies in the central and southern regions. Hail-prone zones are predominantly situated on downslope terrains, leeward mountain slopes, and the Ziwuling Mountains, while hail-scarce areas are concentrated in the flat loess plateau and southern Liupan Mountains. (2) Over the past 46 years, hail days have declined, with the sharpest decrease observed in spring. Hail predominantly occurs between May and August, accounting for 81.5% of annual hail events. (3) The diurnal variation of hail follows a single-peak pattern, with peak occurrences between 15:00 and 18:00. Hail events of short duration (0–9 min) and medium diameter are most frequent. Localized hail occurs more frequently than regional hail, though the incidence of the latter is increasing significantly. (4) A primary oscillation period of 3 years, and secondary cycles of 14 and 35 years, characterize hail frequency. (5) The principal meteorological drivers of hail vary seasonally, with convective available potential energy (CAPE) and the 0°C isotherm height being the most influential. These findings provide a scientific basis for understanding hail occurrence patterns in Longdong and serve as a reference for enhancing forecasting and early warning systems, as well as for guiding artificial hail suppression strategies.

Keywords: hail; climatic characteristics; influencing factors; climate warming; Longdong region

Hail is solid precipitation in the form of hard spheres, cones, or irregular shapes that falls from intensely developed cumulonimbus clouds. It is a disastrous weather phenomenon with distinct seasonality, strong locality, rapid onset, short duration, and primarily mechanical damage, often causing significant harm to industry, agriculture, and people's lives and property. Hail has long been a research focus for meteorologists both domestically and internationally. Research on hail extends beyond short-term forecasting and physical mechanisms to include climatic characteristics, which help identify occurrence patterns and features at specific locations and are equally important for hail forecasting and suppression efforts. Hail climatic characteristics vary across regions, as do the topographical and meteorological factors influencing hail formation.

Studies have shown that in southwestern Germany, the annual average number of thunderstorm days remained essentially unchanged within a certain period, while hail days increased significantly. In China, since the 1980s, hail frequency in northern and northwestern regions has decreased noticeably. Zhang et al. analyzed the spatiotemporal characteristics of hail across China, noting that hail events in most regions occur primarily between 15:00–20:00 local time, while events in Guizhou and Hubei provinces often occur at night. Many domestic scholars have conducted climate analyses for specific regions. Feng et al. found that hail frequency on the Qinghai Plateau decreased significantly from 1980 to 2018, with high-altitude areas in the south being hail-prone zones where hail duration was longer, and large hail mainly fell in eastern low-altitude areas with fewer hail events. In Shandong Province, hail primarily occurs in central and northern regions, featuring large convective available potential energy (CAPE), moderate to strong vertical wind shear, significant conditional instability, and suitable characteristic layer heights. Guizhou's hail is mainly distributed in the central-western region, with annual hail frequency positively correlated with altitude, and hail events concentrated between 14:00–02:00, with daytime frequency decreasing from northwest to southeast.

The Longdong region is located on the Loess Plateau with complex terrain, crisscrossing mountains, hills, and gullies, large elevation differences, and significant temperature variations, resulting in high incidence of strong convective weather. The Loess Plateau has loose soil and fragile ecology. Hail occurs mainly in spring and summer, coinciding with critical growth periods for winter wheat, apples, corn, and other major economic crops, easily causing significant economic losses and threatening lives. Many scholars have conducted research on the causes of hail weather in Longdong, radar characteristics, and forecasting indicators, providing a basis for short-term forecasting and early warning. Wu et al. analyzed hail climate characteristics in Longdong from 1971 to 2000, concluding that hail concentrated in May–September, mainly between 13:00–19:00. Wang et al. analyzed the spatiotemporal distribution, circulation background, and radar echo characteristics of hail in Pingliang City. However, these conclusions were based on data before 2010, focusing on basic spatiotemporal distribution and forecasting summaries. With significant warming and humidification in Northwest China, research on hail climate in Longdong, especially analysis of influencing factors, is particularly necessary.

This study uses daily hail observation data from 15 national meteorological stations in Longdong from 1978 to 2023, combined with ERA5 reanalysis data from the European Centre for Medium-Range Weather Forecasts, to analyze hail climate change characteristics and influencing factors. The aim is to reveal spatiotemporal variation features and occurrence patterns of hail in Longdong, providing scientific references for constructing local hail forecasting and early warning models and optimizing artificial hail suppression mechanisms.

1.1 Study Area Overview

According to Li et al.'s climate division of Gansu Province, the province is divided into Hexi, Longzhong, Longdong, Longnan, and Gannan regions. This study focuses on the Longdong region, including Pingliang City and Qingyang City, totaling 15 counties (districts). Pingliang comprises 7 counties (districts): Kongtong District, Jingchuan County, Lingtai County, Chongxin County, Zhuanglang County, Jingning County, and Huating City. Qingyang comprises 8 counties (districts): Xifeng District, Huan County, Huachi County, Qingcheng County, Zhenyuan County, Heshui County, Ning County, and Zhengning County. Geographically, Longdong is located between 105°20′–108°45′E and 34°54′–37°10′N, at the intersection of Gansu, Shaanxi, and Ningxia provinces, belonging to the Loess Plateau region. Influenced by the uplift of Liupan Mountain, Long Mountain, Huajia Ridge, and Ziwuling Mountain, the region forms eastern loess hilly areas, northern loess hilly-gully areas, southwestern mid-mountain areas, and central-southern Loess Plateau gully areas. Elevations range from 885 to 2857 m. The region is affected by northwesterly winds, with southeasterly winds prevailing in summer and rainfall decreasing from south to north. Longdong is a major agricultural area in Gansu, known as the "Granary of Longdong." Due to complex terrain and large elevation differences, it is one of Gansu's hail-prone areas, posing serious threats to agricultural production and people's lives and property.

1.2 Data Sources

This study uses hail disaster data from 15 counties (districts) in the National Comprehensive Disaster Risk Survey, important weather reports (CODE 939) from national basic stations, and hail disaster information reported by information officers to analyze spatiotemporal distribution, hail diameter, duration, extent, and periodic variation in Longdong. To ensure data validity, stations with complete records during the study period were selected, resulting in 15 stations with records including station number, altitude, latitude/longitude, hail time, maximum hail diameter, hail frequency, and start/end times (precise to minutes). Atmospheric physical quantity data were obtained from ERA5 monthly mean data from the European Centre for Medium-Range Weather Forecasts, with spatial resolution of 0.1°×0.1°. Selected parameters include 0°C isotherm height, dew point temperature, K-index, -20°C layer height, cloud base height, total precipitation, and convective precipitation. This dataset shares the same source as the ECMWF data most frequently used by forecasters and is commonly employed in meteorological analysis and mechanism research. Correlation analysis between atmospheric physical parameters from ERA5 and hail days can test the applicability of these parameters for hail forecasting in Longdong and provide scientific references for forecasters.

Following relevant studies, a hail day is defined as a day when hail occurs at one or more stations across the region within a day. During a hail day, multiple hail events at a single station are counted as one station occurrence. The day boundary follows meteorological observation standards from 20:00 to 19:59. Hail duration is taken as the integer minute value from start to end time; if less than 1 minute, it is recorded as 1 minute. Hail diameter (D) is classified according to national standards: small hail (D < 5 mm), medium hail (5 mm ≤ D < 20 mm), large hail (20 mm ≤ D < 50 mm), and extremely large hail (D ≥ 50 mm). Hail extent is divided into regional and localized hail. A regional hail process involves hail in 3 or more counties (districts) in one day, while localized hail involves fewer than 3 counties (districts). Seasons are defined as spring (March–May), summer (June–August), autumn (September–November), and winter (December–February).

1.3 Research Methods

Inverse distance weighting interpolation in the ArcGIS spatial analysis module was used to map the spatial distribution of hail days. Linear trend estimation was applied to calculate climate tendency rates, and the Mann-Kendall test was used for trend analysis and mutation point detection. Morlet wavelet analysis was employed to analyze periodic variations in hail days. Python programming was used to process ERA5 reanalysis data and calculate correlations between atmospheric physical parameters and hail days. The study period spans 1978–2023, with significance levels of P < 0.001 (|r| > 0.469), P < 0.01 (|r| > 0.376), and P < 0.05 (|r| > 0.291).

2.1 Spatial Distribution Characteristics

From the spatial distribution of hail days in Longdong from 1978 to 2023 [FIGURE:2], Huan County in the north had the most hail days (150 d), followed by Zhenyuan County (131 d), while Qingcheng County had the fewest (53 d). The difference between maximum and minimum reached 2.8 times, showing extremely uneven distribution with an overall pattern of more hail in the northwest and southeast, and less in the central and southern regions. Huating, Chongxin, Zhuanglang, and Lingtai counties are located on the southern slopes of the Liupan Mountains (north-south orientation) and have relatively fewer hail days, consistent with findings by Zhou et al.

To investigate the causes of this spatial distribution, we conducted linear fitting between station altitude and hail days [FIGURE:231]. The correlation was weak, likely because complex terrain in Longdong means meteorological stations are often built on flat loess plateau and river valleys whose elevations cannot objectively represent entire counties. Further analysis revealed that hail-prone areas are influenced by two factors: topography and hail cloud source/movement paths. Huan County, the most hail-prone, is in northern Longdong with high northwest and low southeast terrain (maximum elevation 2089 m, minimum 1136 m). Large elevation differences and low vegetation coverage provide favorable conditions for hail. Most strong convective clouds affecting this area originate from Daluo Mountain in Ningxia and move southeastward; as they cross the downslope region, mountain airflow promotes convective activity at wave peaks, creating a hail-prone zone. Kongtong District and Zhenyuan County are located on the eastern foothills of southern Liupan Mountain, where strong convective clouds mostly originate from Haiyuan County in Ningxia and intensify after crossing Liupan Mountain, forming another hail-prone area. Zhengning County is in the Ziwuling Mountain area with complex terrain and abundant moisture, making it a major hail source affecting the Shaanxi-Gansu border region, thus forming the third hail-prone zone. Qingcheng, Xifeng, and Heshui counties are on the flat loess plateau with high vegetation coverage, which is unfavorable for hail cloud development, resulting in fewer hail days.

In summary, hail distribution in Longdong is extremely uneven, showing a pattern of more hail in the northwest and southeast, and less in the central and southern regions. Hail-prone areas can be divided into three zones: northwestern Huan County; the eastern foothills of Liupan Mountain (Kongtong District and Zhenyuan County); and the Ziwuling Mountain area. Hail-scarce areas include two zones: the flat loess plateau centered on Xifeng, Qingcheng, and Heshui counties; and southern Liupan Mountain including Huating, Chongxin, Zhuanglang, and Lingtai counties. This spatial distribution is closely related to Longdong's topography, with hail-prone areas on downslope terrain, leeward slopes, and mountainous regions, while hail-scarce areas are on relatively flat loess plateau and southern Liupan Mountain.

2.2.1 Interannual Variation

From 1978 to 2023, Longdong experienced 649 hail days, with large interannual fluctuations and alternating high and low years. The year with the most hail days was 1979 (23 d). The overall trend shows a decline of 0.426 d/10a, though not statistically significant. In terms of decadal variation, the 1980s had the most hail days (16.1 d annually), while the 2010s had the fewest (12.5 d annually). Spring and summer had no-hail years concentrated in the 2010s–2020s, while autumn showed alternating no-hail and hail years.

2.2.2 Mann-Kendall Mutation Test of Interannual Variation

The Mann-Kendall test was applied to analyze mutation characteristics of hail day interannual variation. Results show that annual hail days [FIGURE:4] exhibited an overall decreasing trend. The forward sequence (UF) showed decreasing characteristics, with the decreasing trend exceeding the confidence interval after 1990, indicating a significant reduction in hail days after 1990. The backward sequence (UB) showed an increasing trend from 1978–1990 and decreasing trend from 1990–2000. The intersection of UF and UB curves in 1990 represents a mutation point.

Spring hail days [FIGURE:4] showed an overall decreasing trend, passing the P < 0.01 significance test. The UF curve exhibited an increase-decrease pattern, with both increasing and decreasing trends exceeding the confidence interval, indicating significant increases before 1990 and significant decreases after 1990. Summer hail days [FIGURE:4] showed a decreasing trend, with the UF curve showing a decrease-increase-decrease pattern and the decreasing trend exceeding the confidence interval after 2000, indicating significant reduction after 2000. Autumn hail days [FIGURE:4] showed a relatively consistent decreasing trend, with the UF curve exceeding the confidence interval after 1998, indicating a very significant decreasing trend.

2.2.3 Monthly Variation

Analysis of monthly hail day distribution from 1978 to 2023 [FIGURE:5] shows distinct seasonal characteristics in Longdong. Hail occurs in spring, summer, and autumn (excluding winter), with summer being the high-incidence period and autumn having few events. Monthly variation shows a single-peak pattern, with July having the highest frequency (188 d, 29.0% of annual total), followed by June (144 d, 22.2%). Winter months have extremely low occurrence, with only 1 event each in December and February.

The seasonal and monthly variation characteristics are mainly due to rising surface temperatures after March, with the region located ahead of a surface warm high-pressure system while upper-level northwesterly flows bring frequent cold air activity. Increased atmospheric instability provides favorable conditions for hail formation. The earliest hail days in Longdong were March 30 (Zhengning County) and April 1 (Zhuanglang and Lingtai counties), all in southern Longdong where temperatures are relatively higher, demonstrating that thermal conditions significantly influence hail onset timing.

2.2.4 Diurnal Variation

Statistics of hail occurrence times from 1978 to 2023 [FIGURE:6] (7,870 station occurrences) show a single-peak diurnal pattern. Hail station occurrences increase significantly after 12:00, with 15:00–18:00 being the main hail period (3,998 station occurrences, 50.8% of total). The peak occurs at 17:00–18:00 (1,438 station occurrences, 18.3%). Four time periods had no hail occurrences: 20:00–21:00, 02:00–03:00, 04:00–05:00, and 06:00–07:00, all in early morning hours. Hail occurrences decrease significantly after 20:00.

Time-period statistics show: afternoon (12:00–20:00) accounts for 91.6% of hail; night (20:00–08:00) accounts for 7.6%; morning (08:00–12:00) accounts for only 0.8%. This indicates hail mainly occurs in afternoon to evening when temperatures are highest. Enhanced thermal contrast between lower and upper atmosphere favors formation of thermally unstable stratification, promoting hail development.

2.3 Hail Duration

Statistics of single hail event duration from 1978 to 2023 [FIGURE:7] (7,870 station occurrences) show that hail duration in Longdong is generally short. Events lasting 0–9 min are most frequent (6,419 occurrences, 81.5%). As duration increases, occurrences decrease rapidly. Only 2 events lasted ≥40 min: Huachi County on June 4, 2005 (60 min) and Jingning County on June 23, 2014 (48 min). All hail days except these two had fewer than 3 station occurrences, indicating higher probability of longer duration for multi-station hail events, consistent with Sui et al.'s findings.

2.4 Hail Diameter

Statistics of maximum hail diameter from 1978 to 2023 [FIGURE:8] (7,870 station occurrences) show that medium-diameter hail (5–20 mm) occurs most frequently (4,981 occurrences, 63.2%), followed by large hail (20–50 mm, 1,456 occurrences, 18.5%). The maximum recorded diameter was 10 cm (Zhenyuan County, July 23, 1979), reaching the extremely large hail standard. According to Sun Jisong's definition of severe convective weather, D ≥ 5 mm is severe convection and D ≥ 20 mm is extreme severe convection. In Longdong, 25.7% of hail events reached severe convection standards, and 6.5% reached extreme severe convection standards.

The China Meteorological Disaster Encyclopedia: Gansu Volume records that on July 23, 1979, Huan County experienced extremely large hail (egg-sized) causing casualties and crop failure over 1,172.4 hm². On July 23, 2013, Pingliang City experienced historically rare hail disasters in 7 townships with egg-sized hail, some as large as bricks or basins, affecting 7,870 hm² of crops. Although extreme severe convection hail has low frequency, its occurrence causes catastrophic damage in Longdong.

2.5 Hail Extent

From 1978 to 2023, Longdong experienced 157 hail events, with localized hail predominating (145 events, 92.4%) and regional hail accounting for only 12 events (7.6%). Localized hail occurred in all months, peaking in July (45 events). Regional hail occurred from May to August, peaking in July (5 events). Interannual variation [FIGURE:9] shows regional hail increasing significantly (P < 0.01), with the highest frequency in 2013 (3 events, 25% of annual regional hail events). Localized hail decreased gradually over time. There were 19 years with no regional hail.

Comparison of hail diameter and extent shows most large and extremely large hail events were associated with regional hail processes. For example, the July 23, 1979 event affected 3 counties with maximum diameter 10 cm; the July 23, 2013 event affected 7 counties with maximum diameter 75 mm; and the June 4, 2005 event affected 8 counties with maximum diameter 72 mm. This indicates regional hail processes in Longdong are more likely to produce large-diameter hail.

2.6 Hail Periodic Variation

Morlet wavelet analysis was applied to study periodic variation of hail days in Longdong from 1978 to 2023 [FIGURE:10]. In the wavelet coefficient contour map, warm colors indicate positive values (hail-prone years) and cool colors indicate negative values (hail-scarce years). The results show strong oscillations at three time scales: 3–5 years, 13–15 years, and 34–35 years, with alternating positive and negative centers. Wavelet variance shows peaks at 3 years, 14 years, and 35 years, indicating these are the primary and secondary oscillation periods. Since Longdong was in a high-incidence period of the 14-year cycle around 2023, we can infer that the region will maintain a hail-prone trend in the coming years.

2.7 Correlation Between Atmospheric Physical Parameters and Hail Days

Hail formation requires favorable synoptic conditions, low-level moisture convergence, atmospheric instability, and triggering mechanisms. Cloud physics mechanisms, vertical wind shear, and characteristic layer heights such as the -20°C level also play crucial roles. Based on these physical conditions, we selected atmospheric parameters from ERA5 data—dew point temperature, CAPE, K-index, 0°C isotherm height, -20°C layer height, cloud base height, total precipitation, and convective precipitation—and calculated their correlations with hail days [TABLE:1].

Spring: Correlations with all parameters passed significance tests. Hail days are positively correlated with CAPE, K-index, convective precipitation, total precipitation, dew point temperature, and cloud base height, and negatively correlated with 0°C and -20°C isotherm heights. The strongest correlations are with CAPE (r = 0.632, P < 0.001) and K-index (r = 0.581, P < 0.001), followed by 0°C isotherm height (r = -0.479, P < 0.001) and dew point temperature (r = 0.421, P < 0.01). This shows spring hail formation is complex, with instability energy factors showing higher correlations due to lower overall atmospheric temperatures in spring—stronger instability energy can lead to frequent hail when combined with other factors.

Summer: Hail days show significant negative correlation only with 0°C isotherm height (r = -0.312, P < 0.05). Lower 0°C height corresponds to more frequent hail. In summer, high temperatures and abundant moisture create favorable conditions, with terrain uplift and local thermal instability significantly increasing severe convection. Higher 0°C heights in summer mean lower heights favor increased hail days.

Autumn: Hail days show significant positive correlation with CAPE (r = 0.521, P < 0.001). Autumn in Longdong is cool and rainy with generally suitable moisture conditions and characteristic layer heights, but weakened thermal instability. Therefore, stronger instability energy plays a dominant role in hail occurrence.

In summary, atmospheric physical parameters show strong correlations with spring hail days, with CAPE, K-index, 0°C isotherm height, and cloud base height being dominant factors. The 0°C isotherm height dominates summer hail days, while CAPE dominates autumn hail days.

3 Conclusions

(1) Hail days in Longdong exhibit uneven spatial distribution, with an overall pattern of more hail in the northwest and southeast, and less in the central and southern regions. The spatial distribution is closely related to topography. Hail-prone areas can be divided into three zones: northwestern Huan County; the eastern foothills of Liupan Mountain (Kongtong District and Zhenyuan County); and the Ziwuling Mountain area. Hail-scarce areas include two zones: the flat loess plateau centered on Xifeng, Qingcheng, and Heshui counties; and southern Liupan Mountain including Huating, Chongxin, Zhuanglang, and Lingtai counties.

(2) Interannual variation shows a declining trend in hail days, with the most significant decrease in spring. Mann-Kendall tests reveal that annual and summer hail days show a decrease-increase-decrease pattern, with significant decreases after 1990 and 2000, respectively. Spring shows an increase-decrease pattern with significant increases before 1990 and decreases after 1990. Autumn shows a consistently significant decreasing trend. Hail exhibits clear seasonality, with May–August being the high-incidence period (81.5% of annual total). Diurnal variation shows a single-peak pattern, with peak occurrence between 15:00–18:00. Short-duration (0–9 min) and medium-diameter hail are most frequent. Localized hail far exceeds regional hail, though regional hail shows a significant increasing trend.

(3) Hail days exhibit a primary oscillation period of 3 years and secondary periods of 14 and 35 years. Longdong was in a high-incidence period of the 14-year cycle around 2023, indicating the region will likely maintain a hail-prone trend in coming years.

(4) Correlation analysis shows seasonal differences in dominant factors. In spring, CAPE, K-index, 0°C isotherm height, and cloud base height dominate hail days. In summer, 0°C isotherm height is dominant. In autumn, CAPE is dominant.

(5) Hail data relies primarily on manual observations, limiting precision to the county level. This study examined topographic effects and selected atmospheric physical parameters, but the analysis is not comprehensive. Future research should integrate multi-source data for deeper investigation of hail climate influencing factors.

References

[1] Feng Xiaoli, Ma Zhanliang, Guan Qin, et al. Spatiotemporal characteristics of hail and its influence factors in Qinghai Plateau during 1980–2018[J]. Meteorological Monthly, 2021, 47(6): 717–726.

[2] Sun Jisong. Fundamental principles and technical methods for severe convective weather forecasting[M]. Beijing: China Meteorological Press, 2014: 3–6.

[3] Yu Huaizheng, Long Dianli, Zhu Xiuhong, et al. Analysis of spatial-temporal distribution characteristics of hails in Shandong Province and classified forecast indexes[J]. Meteorological Science and Technology, 2023, 51(2): 254–261.

[4] Zhang Qiang. Research on hail in northwest China[M]. Beijing: China Meteorological Press, 2005: 56–68.

[5] Yu Xiaoding, Wang Xiuming, Li Wanli, et al. Nowcasting of thunderstorms and severe convective weather[M]. Beijing: China Meteorological Press, 2020: 65–75.

[6] Liu Tao, Xia Xiaoling, Peng Yuxiang, et al. Climatic characteristic and causes of hail in Guizhou: 1961–2021[J]. Journal of Tropical Meteorology, 2023, 39(6): 872–882.

[7] Kang F, Zhang Q, Lu S. Validation and development of a new hailstone formation theory: Numerical simulations of a strong hailstorm occurring over the Qinghai Tibetan Plateau[J]. Journal of Geophysical Research: Atmospheres, 2007, 112(D2): 1–13.

[8] Punge H, Kunz M. Hail observations and hailstorm characteristics in Europe: A review[J]. Atmospheric Research, 2016(1): 176–193.

[9] Li M X, Zhang Q H, Zhang F Q. Hail day frequency trends and associated atmospheric circulation patterns over China during 1960–2012[J]. Journal of Climate, 2016, 29: 7027–7044.

[10] Zhang C, Zhang Q, Wang Y. Climatology of hail in China: 1961–2005[J]. Journal of Applied Meteorology & Climatology, 2008, 47(3): 795–804.

[11] Zhang Qiang, Yang Jinghu, Wang Pengling, et al. Progress and prospect on climate warming and humidification in northwest China[J]. Chinese Science Bulletin, 2023, 68(14): 1814–1828.

[12] Zhang Fanghua, Gao Hui. Temporal and spatial features of hail days in China[J]. Journal of Nanjing Institute of Meteorology, 2008, 31(5): 687–693.

[13] National Meteorological Center. Hail grade GB/T 27957-2011[S]. Beijing: Standards Press of China, 2011: 3–6.

[14] Wen Kegang. China meteorological disaster encyclopedia: Gansu[M]. Beijing: China Meteorological Press, 2005: 169–197.

[15] Wu Aimin, Zhang Tianfeng, Bai Aijun. Climatic characteristics and forecast of hail on the Loess Plateau of eastern Gansu[J]. Journal of Catastrophology, 2005, 20(2): 37–40.

[16] Wang Ruosheng, Zhang Tong, Fan Xiaochun, et al. Analysis of the climate features and radar echo of the hail weather in Pingliang of Gansu Province[J]. Journal of Arid Meteorology, 2013, 31(2): 373–377.

[17] Yang Lijie, Cheng Peng, Liu Xiang, et al. Analysis of climatic characteristics of severe convective weather in Longdong Area in recent 50 years[J]. Desert and Oasis Meteorology, 2023, 17(1): 25–31.

[18] Lu Yaqi, Jiao Meiling, Li Xiangke, et al. Comparative analysis and prediction method of short time heavy precipitation and hail weather in eastern Gansu Province[J]. Arid Land Geography, 2016, 39(2): 317–326.

[19] Lu Yaqi, Cao Yanchao, Zhang Feng, et al. The prediction and forewarning system as well as weather characteristics analyses of hail in the east of Gansu[J]. Plateau Meteorology, 2016, 35(6): 1565–1576.

[20] Lei Yu, Huang Wubin, Li Qian, et al. Characteristics of Doppler radar products of strong hails under different weather classification in Hedong region of Gansu Province[J]. Journal of Arid Meteorology, 2022, 40(2): 234–243.

[21] Zhou Wei, Zhang Qiang, Kang Fengqin. Recent advancements of hail research in northwest China[J]. Advances in Earth Science, 2005, 20(9): 1029–1036.

[22] Huang Yuxia, Wang Baojian, Wang Yanfeng, et al. Variation characteristics of hail days in Gansu Province during 1974–2013[J]. Meteorological Monthly, 2017, 43(4): 450–459.

[23] Ding Yihui, Liu Yanju, Xu Ying, et al. Regional responses to global climate change: Progress and prospects for trend, causes, and projection of climatic warming and wetting in northwest China[J]. Advances in Earth Science, 2023, 38(6): 551–562.

[24] Li Dongliang, Liu Dexiang. Gansu climate[M]. Beijing: China Meteorological Press, 2000: 41–56.

[25] Guo Haiying, Zhang Moucao, Zhou Zhongwen, et al. The impact of warming and wetting of climate on production of winter wheat in eastern Gansu Province[J]. Chinese Agricultural Science Bulletin, 2023, 39(15): 127–131.

[26] Tang Yandong, Zang Cuiping, Yu Yunpeng, et al. Developmental characteristics and influencing factors of gullies based on fractal theory in eastern Gansu Province[J]. Arid Land Geography, 2025, 48(2): 223–233.

[27] Cao Yanchao, Han Jing, Lu Yaqi, et al. Activity characteristics and influence factors of two kinds of summer thunderstorms in Longdong[J]. Desert and Oasis Meteorology, 2021, 15(5): 64–70.

[28] Yi Nana, Su Lijuan, Zheng Xucheng, et al. Environmental parameters and forecast models of hail events[J]. Arid Zone Research, 2024, 41(1): 13–23.

[29] Sui Yuxiu, Yang Jingtai, Li Yuqian, et al. Climatic characteristics of hails in Dalian region of Liaoning Province from 1971 to 2018[J]. Journal of Arid Meteorology, 2020, 38(2): 249–255.

[30] Zhang Haiyao, Huang Yuxia, Wu Huiyan, et al. Comparative analysis of two consecutive hail weathers in complex terrain area of Loess Plateau[J]. Journal of Arid Meteorology, 2022, 40(4): 646–655.

[31] Xin Yue, Su Lijuan, Zheng Xucheng, et al. Hail characteristics and hail recognition method based on machine learning in Inner Mongolia[J]. Arid Land Geography, 2025, 48(1): 11–19.

[32] Ma Rongjun, Li Xiaofei, Mao Yanlin, et al. Characteristics and causes of climate change in hail occurrence in Shaanxi Province[J]. Arid Zone Research, 2022, 39(3): 767–773.

[33] Huang Yue, Sun Jianhua, Zhang Yuanchun, et al. Characteristics of temporal and spatial distribution of severe convective weather and its response to climate warming in Qinghai Province[J]. Journal of Arid Meteorology, 2011, 29(4): 439–445, 454.

[34] Wang Zhilan, Chen Luyuan, Shang Kezheng, et al. Initiation, organizational modes and environmental conditions of severe convective wind events during the warm season over north China[J]. Science China: Earth Sciences, 2024, 54(9): 2929–2947.

[35] Qiu Li, Jiang Zongzu, Wang Lingjun, et al. Study on characteristics and influence path of hail disaster in Yan'an gully region[J]. Agricultural Disaster Research, 2023, 13(3): 184–186.

[36] Wang Zixin. Application of multi-source data fusion algorithm in a severe convective weather process[D]. Nanjing: Nanjing University of Information Science and Technology, 2021.

[37] Sun Xuying. Research on formation mechanism, characteristics and radar recognition technology of hail weather in Gansu Province[D]. Nanjing: Nanjing University of Information Science and Technology, 2007.

[38] Zhao Wenhui, Yao Zhanyu, Jia Shuo, et al. Characteristics of spatial and temporal distribution of hail duration in China during 1961–2015 and its possible influence factors[J]. Chinese Journal of Atmospheric Sciences, 2019, 43(3): 539–551.

[39] Zhang Hongfa, Gong Jiawei, et al. Observational investigation of characteristics of severe convective hook echo in Pingliang region[J]. Chinese Journal of Atmospheric Sciences, 1997, 21(4): 18–29.