Based on the hourly precipitation analysis data of the China Meteorological Administration multi-source merged precipitation analysis system (CMPAS) and the hourly precipitation data predicted by the China Meteorological Administration mesoscale weather forecast system (CMA-MESO)， the distribution characteristics of precipitation in the Beijing-Tianjin-Hebei region under special topographic conditions from June to September 2021 were analyzed, and the prediction performance of CMA-MESO was discussed. The results are as follows: (1) The observational maximum centers of mean hourly precipitation in the Beijing-Tianjin-Hebei region were primarily located in 100-600 m altitude on the windward slopes of the eastern Taihang Mountains and the southern foothills of the Yanshan Mountains, while the maximum centers predicted by the CMA-MESO were located on the side of the windward slope leaning towards the plain in front of the mountains. The observational hourly precipitation frequency and intensity were similar to precipitation amount, but the maximum center of hourly precipitation frequency was located on the windward slopes of the Taihang Mountains, leaning towards the mountainous side, while the maximum center of precipitation intensity was mainly distributed on the windward slopes in front of the mountains and the plain areas of the eastern Beijing-Tianjin-Hebei region. (2) The observational regional average hourly precipitation amount on the windward slopes in front of mountains of the Beijing-Tianjin-Hebei region exhibited a bimodal diurnal pattern, with the primary peak occurring from afternoon to evening and the secondary peak in the early morning. The primary peak predicted by the CMA-MESO was colse to observations, but the regional average hourly precipitation amount was significantly overestimated. (3) On the windward slopes in front of the mountains, the peak period of precipitation above 10 mm·h^-1 occurred from the afternoon to the early morning and the early hours of the next day. The CMA-MESO forecast indicated that the precipitation above 10 mm·h^-1 in the afternoon to evening period was slightly higher, while the precipitation in the early hours of the next day was slightly lower. (4) Precipitation events on the windward slopes from afternoon to early nighttime were mainly short-term precipitation events within 3 hours. The CMA-MESO taked the characteristic, but the amount of short-term precipitation events predicted by it was relatively high. (5) The CMA-MESO successfully forecasted the topographic enhancement of precipitation on the windward side of the mountains. However, the specific humidity below 850 hPa was underestimated, and the convective available potential energy value at 14:00 (Beijing Time) was significantly underestimated. These biases contributed to the existence of a negative precipitation bias center over the windward slopes.

Inner Mongolia is one of significant seasonal snow-covered regions in China. Snow identification and snow depth inversion are crucial for agricultural production, ecological assessment, and research on spring floods and snow-related disasters. In order to improve the accuracy of local snow identification, a direct comparison snow identification method based on Normalized Difference Snow Index (NDSI) is proposed in this paper, the method involves applying the NDSI difference operation between the snow map to be identified from the Himawari-8 satellite images and the snow-free base map from the current autumn to identify snow, and it is compared with the snow identification methods used in routine business. The results indicate that the Snow Mapping (SNOMAP) algorithm, based on the Normalized Difference Vegetation Index (NDVI), tends to miss some thin snow pixels, while the Fractional Snow Cover (FSC) algorithm can be affected by water bodies in snow identification and ultimately affect its accuracy. In the non-forest areas of Inner Mongolia, the accuracy of NDSI direct comparison was 3.88% higher than SNOMAP and 0.52% higher than FSC. The difference between the accuracy of NDSI direct comparison and FSC in non-forest areas was small. In forest areas, compared with FSC algorithm, NDSI direct comparison method significantly improved the identification accuracy, while the error rate decreased, indicating that NDSI direct comparison method is more suitable for snow identification in forest areas of Inner Mongolia.

To improve the fine-grained forecasting and early warning level of high-impact weather in cities, this paper utilizes 7 categories of data from the HMB-KPS millimeter-wave cloud radar at Rongcheng meteorological station of Xiong’an, including radar reflectivity factor, radial velocity, etc and surface meteorological observations, as well as weather analysis, radar echo image morphology analysis and other methods to analyze the major high-impact weathers in Xiong’an from January 2020 to December 2023， such as torrential rains，strong winds，fog，haze，and sand-dust weathers.The results are as follows: 1) The millimeter-wave cloud radar clearly and stereoscopically detects the structural distribution of cloud systems in high-impact weather including precipitation. In spring, summer and autumn, when clouds with echoes intensity equal to or more than 21 dBZ (15 dBZ in winter) and vertical extension height equal to or more than 5 km (2 km in winter) appear, and the lowest cloud base reaches 150 m, it can be considered as the characteristic index of start time of surface precipitation, and the average time advance is about 20 minutes. Conversely，when echo intensity weakens and cloud base height rises, precipitation tends to end. Among all levels of precipitation events, torrential rains have the maximum vertical extensions, liquid water content, etc. 2) When the extension height of the echo top reaches more than 8 km and then the height of the cloud top falls rapidly to less than 2 km, the ground is prone to strong winds of more than 8 magnitudes. The echo characteristic of precipitation accompanied by strong winds is that obvious U or V notch is formed in the strong echo area, and the maximum radial velocity in the notch area is 12.5 m·s^-1. 3) Cloud radar inversion droplet spectrum data products can better reflect the particle distribution characteristics of precipitation with different levels. The Vertically Integrated Liquid Water (VIL) has an early warning significance for precipitation and wind, and the advance is 0-30 min or more. 4) For non-precipitation high-impact weather (fog, haze, sand-dust), the millimeter-wave radar echoes are overall weaker, with indistinct transitions between weather phenomena. However, during weather transitions, the radar can rapidly and clearly detect the precipitation clouds, improving fine-grained forecasting and early warning capabilities of precipitation.

In order to improve the ability of the hydro-meteorological service, the hydro-meteorological service system of Hanjiang Basin in southern Shaanxi was established based on the refined weather forecast technology and hydro-meteorological model as well as related meteorological observations and hydrological information. The system was designed and developed according to the developing of automatic hydro-meteorological monitoring network, refined hydro-meteorology forecast, as well as the preliminary application of hydro-meteorological service and exploring of 3D simulation. Meanwhile, an ‘ABC’ technology including “artificial intelligence”, “big data” and “cloud computing” had been put into use in the system. As a result, the system could provide the refined forecast and service products and the scientific emergency strategies for the weather monitoring and forecasting, the meteorological service in flood prevention, especially the flood prevention scheduling in Hanjiang Basing in southern Shaanxi.

Based on the data obtained from“Beijing 2006WeatherServices Public Survey”, the qualitative and quantitative statistical analysiswere carried out from three aspects: the general public’swilling to pay for the yearlyweather services, the general public’s attention to themeteorological products and the issues for the further improvement ofweather services. The general public needswere analyzed. It’s necessary to give publicity about theweather service for the generalpublic to realize the value of theweather service. It is found that the generalpublic needs ofmeteorological information have shown certain characteristics, which can bewell explained by theMaslow’sHierarchy ofNeed Theory and theAvailabilityTheory in behavioral economics. Based on the results obtained from this study, some suggestionswere given for improving themeteorological services in the future.

The spatial distribution of land surface temperature (LST) in different seasons in the Qilian Mountain area were derived based on Becker - L i (1990) ’s algorithm. The results show that the distribution of instantaneous LSTwas in general accord with macroscop ical variation of coinstantaneous air temperature data, and the diverse topography and different cover types of underlying surfacedecided spatial distribution patterns of LST. The vertical distribution of LST was also discussed. The results show when the Terra -MOD IS collects data, the LST lap se rate varies from 4. 52 - 6. 56 ℃/km, and the minimum value happens in the direction of sunshine, which app roximates to 4. 52 ℃/km on the right south side of the mountains; the maximal value happens at the opposite side of sunshine, which is 6. 56 ℃/km at the north side of the mountains. The LST lap se rate in high altitude regions is higher than that in low altitude regions.

 Based on the climatic; data of  ten years(1991一2000)China are analyzed.  The result show that the hail distribution is the spatial一temporal distribution(harac;teristi<、of hail in Northwest in belt patterns and the main hail zone is in the middle part of Tibetan plateau Qilian mountain and the west part of Tianshan mountain the more frequent、entre is in the south side along the east一WPSt mountains than that and the east side along the north一、outh mountains } the date o<(urrin} hails in mOthe south part of the plateau and the hail duration is lone but the hail days are the north part of Tibetan Plateau is earlier fewer than that in the south part of the plateau from the (urves of dec;adal and pentad evolution of hail days