The Aircraft Meteorological Data Relay (AMDAR) data has the characteristics of high sampling frequency and dense detection levels, which can effectively supplement the lack of spatial and temporal resolution of high-altitude data. In this paper, the quality analysis and control of AMDAR observation data at different heights were carried out by using the China regional AMDAR data provided by the National Meteorological Information Center and the Global Forecast System (GFS) data provided by National Centers for Environmental Prediction (NCEP) from March 1 to May 31, 2020 (spring). The influence of the assimilation of AMDAR data after quality control on the analysis field and the forecast field was tested by a two-week cycle assimilation comparison test. The results show that the root mean square error (RMSE) between AMDAR wind speed and GFS data shows an obvious increasing trend with the increase of height, and the bias of wind speed and temperature also shows an increasing trend with the increase of height. After quality control of AMDAR data at different heights, the RMSE and bias between AMDAR data and GFS data have been improved, and the observation minus background (OMB) between AMDAR data and GFS data was more consistent with Gaussian distribution. The assimilation of AMDAR data after quality control has a certain improvement on the analysis field of wind, temperature and geopotential height, and the improvement could affect the 12-hour or even 24-hour forecast field. The assimilation of AMDAR data after quality control could improve the precipitation forecasting skills, especially for medium precipitation forecast.

Based on the drought disaster data of the Heng-Shao drought corridor from 1961 to 2018, the meteorological drought composite index (MCI) was used to study the drought monitoring and evaluation methods of the Heng-Shao drought corridor. The results are as follows: (1) During the peak period of crop water demand (from June to October), the drought events with weighted mean of regional MCI (DI) less than or equal to -0.5 and the process duration greater than or equal to 16 days were included in the statistics, and the three elements such as the extreme intensity, cumulative intensity and duration of DI index were the best factors for annual regional drought assessment. Furthermore, based on the three elements of DI index, the annual assessment index of regional drought (MCI_e) calculated by using the TOPSIS method was the best. (2) Based on the MCI_e index, the combined grading method of average value and standard deviation was used to obtain the threshold of regional drought degree for normal, drought, severe drought, and extreme drought years. It was found that the MCI_e index had a strong ability to assess extreme drought years and normal years in the Heng-Shao district, and had good assessment ability for 2019 and 2020. Furthermore, the extreme disaster year in 2013 was simulated, it was found that the MCI_e index could better capture the change of drought in the Heng-Shao drought corridor. Therefore, the MCI_e index could support the rapid assessment and early warning of drought in the Heng-Shao drought corridor to some extent.

Based on the merged data of CloudSat from June to August during 2007-2010, the occurrence probability, vertical structure, type, microphysical structure of single-layer and multi-layer clouds and their environmental temperature and humidity over Liaoning Province were investigated. The results are as follows: (1) The clouds were mainly composed of single-layer cloud over Liaoning Province in summer from 2007 to 2010, and then followed by double-layer cloud. The single-layer clouds were mainly consisted of cirrus and altostratus, and then followed by altocumulus and deep convection. The cirrus mainly distributed at upper layer of multi-layer clouds. The altocumulus, altostratus and stratocumulus mainly distributed at lower layer of double-layer clouds, while the altostratus and altocumulus mainly dominated at middle layer and the stratocumulus mainly dominated at lower layer of triple-layer clouds. (2) The echo intensity and cloud thickness for each layer of multi-layer clouds were less than those of single-layer cloud. (3) The vertical profiles of microphysical quantities (particle number density, effective radius and cloud water content) for each layer of multi-level clouds were similar to single-layer cloud, while their average values were relatively smaller, and they decreased gradually from top to bottom, which were related to cloud types, namely the values of deep convection were the largest and cirrus were the smallest. (4) Compared with clear sky, the specific humidity on cloudy weather was larger, which indicated that the water vapour of atmosphere on cloudy weather was more abundant. The average near-surface temperature was lower under the cloudy weather during the daytime, while it was opposite at night.

Based on the monthly average temperature and precipitation at Shenyang station during 1951-2016, the multi-scale temporal characteristics of annual average temperature and precipitation anomaly were analyzed by using the ensemble empirical mode decomposition (EEMD) method, firstly. Then, the periods of main intrinsic mode functions (IMFs) of two elements were discussed by using power spectrum analysis method. And on this basis, the series of two elements based on the components of IMFs were rebuilt and compared. The results show that the variation of annual mean temperature was mainly caused by the oscillation of the first two components with high frequency and the trend component of IMFs in Shenyang during 1951-2016, which reflected the periodic changes with quasi-5-year, quasi-7-year, and long-term upward tendency, respectively. The impact of the third component of IMFs with quasi-14-year period on the annual average temperature wasn’t ignorable. In addition, the variations of the fourth and fifth components of IMFs with longer interdecadal time scales were consistent with the trend component after the 1980s, which indicated the obvious increase of temperature in Shenyang after the 1980s. The variation of annual precipitation series was mainly caused by the high-frequency oscillations of the first two components of IMFs with quasi-3-year and quasi-5-year periods, respectively, while the trend component with quasi-64-year period of annual precipitation reflected the trend change with the decrease initially and followed by the increase before and after the 1980s in general. Compared with the temperature series, the contribution of interdecadal and long-term trend change to the precipitation series was limited.