The study of summer rainfall characteristics in the middle of the Qilian Mountains can provide weather background support for artificial precipitation enhancement by terrain clouds. Based on hourly surface meteorological observations and Temperature of Black Body (TBB) from FY-2G satellite during 2015-2017, the characteristics of rainfall and convection in summer and relationship between diurnal variation of rainfall and local circulation are investigated. The results show that the total rainfall amount of meteorological stations with an altitude less than 3.5 km in the middle of the Qilian Mountains increases linearly with the altitude, and in the eastern part (east of 99.2°E) it changes more sharply with the altitude than that in the western part. The diurnal variation of valley wind circulation and rainfall at Yeniugou and Qilian stations in the valley is more obvious than that at Gangcha Station near the Qinghai Lake. The maximum and sub-peak of rainfall at the two stations are in the evening and early morning, respectively, corresponding to the high frequency periods of the two main topographic clouds (cumulonimbus and stracumulus) in the region, and the average rainfall intensity is 2.0-2.3 and 1.0-1.3 mm·h^-1, respectively. Compared with Zhangye Station in Hexi Corridor, the peak value of TBB probability distribution at Qilian, Yeniugou and Gangcha stations in the Qilian Mountains changed from -22--12 ℃ to -32--22 ℃ under two kinds of rainfall intensity (less than 1.5 and more than or equal to 1.5 mm·h^-1). When the threshold of TBB <-32 ℃ is used to identify rainfall cloud, the coverage rate of rainfall cloud in the Qilian Mountains is lower than that in Hexi Corridor. The threshold of TBB <-22 ℃ is more suitable for the identification of weak convective rainfall clouds in the Qilian Mountains. In the study area, the high value areas of deep and shallow convection are distributed in the direction of north-south and northwest to southeast, respectively. The diurnal variation of shallow convective frequency based on TBB data can reflect some characteristics of diurnal variation of rainfall in this region.

To improve the forecast and early warning ability for the low visibility along the expressway, the hourly observation data of 10 traffic meteorological stations along Shaanxi section of the Lian-Huo expressway and the hourly reanalysis data of the European Center for Medium-Range Weather Forecasts are employed to analyze the distribution characteristics of low visibility and to explore the relationship of low visibility with other meteorological factors. The results show that along Shaanxi section of the Lian-Huo expressway, January has the most low visibility, while February has the least. During a day, the low visibility mostly happens from 00:00 to 10:00, and the low visibility of 0-50 m mainly occurs from 05:00 to 06:00. The low visibility duration is short with the majority of within 2 hours and the longest of 17 hours. The low visibility weathers occur frequently from Xingping to Changxing and Chencang section, where attention should be paid in daily traffic meteorological service. Accordingly to the analysis of the relationship between low visibility and other meteorological factors, the low visibility generally occurs under the conditions of air temperature from 0 to 5 ℃, relative humidity above 90%, wind speed less than 1.0 m·s^-1 and northeast to east winds. The low visibility weather in summer and winter is mostly associated with precipitation, which usually occurs during or after the precipitation and is accompanied by weather systems always. Compared with the low visibility caused by radiation cooling, the low visibility associated with precipitation exhibits longer duration and the wider range. The low visibility weathers occur at different relative humilities in different seasons, which in winter, summer and autumn are high, while in spring are relatively low.

Based on the hail data from eleven ground meteorological observation stations in the eastern agricultural region of Qinghai Province from 1961 to 2020, the temporal and spatial distribution of hail days, hail diameter, duration and disaster risk characteristics were analyzed by using statistical methods. The main conclusions are as follows: (1) The hail days in the agricultural area of eastern Qinghai Province decreased with climate tendency rate of 11.6 d·(10 a)^-1 in the past 60 years, which passed the significance test of \alpha=0.05, and after 1995, the anomaly of the total number of hail days changed from the positive to the negative. Hualong was the area with the largest number of hail days, and the number of hail days in Jianzha was the least. (2) Hail occurred mainly from May to September each year with seasonal differences. The diurnal variation of hail was obvious, and the peak occurred at 16:00 BST in the afternoon. (3) The number of hail days was positively correlated with altitude of stations and the correlation coefficient between them was as high as 0.97.(4) In the past 60 years, the hail processes with hail diameter less than 6 mm and the duration less than 9 min accounted for 58.33% and 73.55% of the total number of hail processes, respectively. (5) Xunhua was a low-risk area for hail, Hualong, Huangzhong and Huangyuan were medium-risk areas for hail, and Ledu was a high-risk or extremely high-risk area for hail, which was basically consistent with historical hail disasters.

Based on the daily minimum temperature observation data in southeast of Gansu Province during 1969-2018, the variation of extreme low temperature events was analyzed, then 74 circulation characteristic quantities from National Climate Center were used to research the circulation system affecting the extreme low temperature events. The results are as follows: (1) The frequency of annual extreme low temperature events in southeast of Gansu Province reduced obviously with a rate of 2.3 d·(10 a)-1 in the last 50 years, and the reduction was most significant in summer and slowest in winter. There was an abrupt change of extreme low temperature days in 1987, after the abrupt change, the frequency of extreme low temperature events was relatively lower. (2) Compared with the climatic mean, the intensity of extreme low temperature events tended to increase, before 1987, the low temperature intensity anomaly increased with a rate of 0.2 ℃·(10 a)-1, while after 1987, the extrem low temperature intensity anomaly increased with a rate of 1.2 ℃·(10 a)-1. (3) The intensity of extreme low temperature in spring, summer, autumn and winter mainly ranged from -5.0～5.0 ℃, 10.0～15.0 ℃, -5.0～10.0 ℃, -20.0～-10.0 ℃, and the occurrence frequencies were 61.9%, 90.1%, 73.4% and 73.1%, respectively. (4) There was a positive correlation between extreme low temperature events and Eurasian meridional circulation in southeast of Gansu Province. The extreme low temperature events was related with cold air, the area index of western Pacific subtropical high and the intensity of polar vorticity center in the northern hemisphere in winter, while it was related with subtropical high northern boundary of South China sea, subtropical high northern boundary of the western Pacific and polar vorticity intensity of the Pacific in spring. The composited analysis of geopotential height fields indicated that Mongolia was a cold high pressure center on ground, and polar vorticity center was located in the eastern hemisphere from 500 hPa to 100 hPa, the westerly belt prevailed meridional circulation, and east Asian large trough in the westerly belt was deep and westward, the region of southeast of Gansu was controlled by strong northwest airflow after the trough, which was favourable to forming extreme low temperature events.