Abstract:A fullautomatic cloud detection algorithm based on deep semantic segmentation is proposed to improve the accuracy of cloud detection for the remote sensing imagery of FY2E satellites. Firstly, to train and evaluate, a sample data set is created by the data of FY2E L1 matched with the cloud detection results with high accuracy. Secondly, a deep semantic segmentation network is designed. A loss function is improved to extract the cloud’s boundary effectively for a severe imbalance between positive and negative samples in the train data set. Finally, FY2E and MODIS data, taken as train and label samples, respectively, are used for training networks, resulting in four classification models for detecting FY2E L1 imagery. The test results show that the proposed method’s accuracy and the Kappa coefficient are 75% and about 0.53 in four classification tests, respectively. Compared with the existing multichannel threshold method in two classification tests, the proposed method can improve the accuracy of about 90% of the samples and the accuracy of some samples by more than 20%. In addition, the proposed method has a strong recognition ability for cloud edges, broken clouds and other details. It has a certain degree of robustness, which is less affected by the misclassification categories in train samples. Furthermore, by expanding the data set and optimizing the network, the proposed method will improve the data quality of the entire disk imagery of FY2.

Abstract:The time sequence design of the Lband Upperair Sounding System is analyzed. GTS1 Sonde and RS92 Sonde are compared to determine the time difference between the sensing meteorological elements and the time of the ground equipment caused by the Lband Upperair Sounding System will cause the uncertainly of the measurement results. The influence of time difference on pressure is larger, the temperature is smaller, and humidity is more complex. The uncertainty of air pressure will affect the accuracy of sounding application files. The sensing time is always delayed relative to the measuring element, which is about 1.2 s after the measuring element senses the meteorological elements. The time difference is about 0.98 s. The maximum pressure error is up to 0.86 hPa. The BDSGPS Dualmode Navigation Wind Sonde is designed to eliminate the time difference and significantly reduce the random error of the measurements. It has a good effect on the Yangjiang test conducted by the China Meteorological Administration in December 2013, which proposed an improved method.

Abstract:Based on the testing example of visibility measurement and testing laboratory (Hefei), the measurement uncertainty evaluation method of visibility measurement and testing results with a forward scattering visibility meter are presented. Firstly, the composition and technical requirements of the visibility detection system are introduced. Secondly, the measurement model of the indication error of the forward scattering visibility meter is established, and the source of measurement uncertainty is analyzed. Finally, the measurement uncertainty component is evaluated according to different evaluation methods. The results show that the uncertainty components of the forward scattering visibility meter mainly include the visibility meter to be measured, the measuring standard and the standard uncertainty introduced by the test chamber. At 50 m and 200 m detection points, the indication error’s extended uncertainty U (k=2) is 4.9 m and 14.3 m, respectively. The relative extended uncertainty U (k=2) is between 8.5% and 11.0% at the detection point from 500 m to 10,000 m.

Abstract:In the verification process of the humidity sensor, in addition to the uncertainty of verification results caused by the dewpoint meter and temperature and humidity box, the verification conditions and nonstandard operation will also bring additional errors to the verification results. In this paper, the standard deviation of the measured value at each humidity point of the dewpoint meter, the thermal effect of the dewpoint meter during operation, the flow change of the sensor skimming the dewpoint meter, and the optimum working environment of the dewpoint meter are tested. Combined with the historical data of the humidity sensor test, a comprehensive analysis is made to draw the following conclusions: (1) When the flow hole of the dewpoint sensor is facing the airflow direction, the dewpoint temperature measured at the same temperature is higher, and the relative humidity measured by the dewpoint sensor is smaller. Otherwise, it will be larger. (2) When the ambient temperature is 20 ℃, the standard deviation (discreteness) of the measurements of the dewpoint meter at each humidity point is minimum. (3) When the temperature sensor of the dewpoint meter in the suction type temperature and humidity box is installed at 4 cm the front end at 4 cm the back end, the measured value of its rear end is 0.45 ℃ higher than the maximum value installed at the front end, and the relative humidity is 0.5% higher. (4) The stability time at each humidity point (when the humidity in the box reaches the set value and tends to be stable) should be greater than or equal to 20 s, the standard value and measured value collected by the dewpoint meter and humidity sensor under inspection are true.

Abstract:This article focuses on a detailed introduction of the construction requirements and overall architecture of the Meteorological Big Data Cloud Platform (“Tianqing”) archive storage system, and an indepth analysis of several key technical issues of the platform implementation. “Tianqing” archive storage system is based on online storage, offline storage and tape drive management software. Through the hierarchical structure system, it establishes a dynamic and scalable archive cluster. The system implements multiprotocol data archiving and sensitive data archiving to meet various data archive scenarios. The system has data restoration, entire data process traceability, standardized tape management and other functions. In August 2020, the “Tianqing” archive storage system began to run operationally.

Abstract:To realize the standardization and automation of meteorological observation equipment’s test and experiment operations, the Test and Experiment System of Meteorological Observation Equipment was designed and implemented directing at conventional ground observation equipment. Based on Web technology and the Oracle database, several functional modules, including process management, data configuration, equipment operation and state management, expert analysis and calculation, data inquiry and display, generation and export of test report and system management, have been realized in the way of flexible configuration, hierarchical design and modular application. It standardized the management of test and experiment operations and improved the work efficiency as well. The results of calculation and evaluation are consistent with those of manual treatment. On this basis, further design and development will be reinforced of the test and experiment platform system appliable to radiosounding equipment, weather radar and other meteorological observation equipment.

Abstract:By stamping the Esignature on the electronic documents, it can meet the needs of information integrity, legitimacy and nonrepudiation in the process of electronic office documents of government departments. In 2005, China Meteorological Administration began to promote the Esignature application in the electronic official document system, which effectively accelerated the whole electronic process within the China Meteorological Administration and improved the management efficiency. With the development of independent and controllable national security and Internet Plus government services, the Esignature service platform based on domestic cryptography and domestic format documents is built under the overall framework of the meteorological administration information system. The system provides electronic signature services for electronic document systems and administrative approval electronic certification applications through cloud services. The system runs stably in operation and improves the security and scalability of the business significantly.

Abstract:The China Meteorological Administration has always had the tradition of weather discussion through video conferences, but due to the technical conditions, the meteorological video conference system can only be used via the meteorological intranet for many years. In order to meet the needs of users to access the video conference system based on internet, the National Meteorological Information Center has completed the cloud video conference system in 2019. Relying on domestic Internet video conference products, the system is deployed in the way of private cloud, and realizes the integration with the existing meteorological video conference system based on H.323 protocol. Setting up a friendly enterprise mode among multiple cloud consultation systems meets the requirements of system classification and localized management of meteorological departments. It also dramatically improves the simplicity of interdepartmental consultation and linkage. In this paper, the system’s design scheme and key technologies are described, and the application effect of the system in meteorological services for significant events and the process of meteorological departments fighting the COVID19 epidemic is introduced. The development direction of the cloud consultation service is also proposed.

Abstract:This platform implements the stationlevel monitoring of the new generation weather radar ROSE system through the deployment of Zabbix. The entire monitoring architecture is divided into three layers: the provincial level, the ROSE platform, and the server of the bureau. The process is mainly divided into several steps, including environment deployment, host configuration, template configuration, alarm module configuration, and log analysis. The form added by Zabbix templates (Templates) can easily complete conventional monitoring items such as host performance, network connection status, service operation status, etc., and the monitoring and alarming of streaming can be realized by constructing polling function monitoring items and associated trigger functions. The entire monitoring platform has a featurerich frontend interface, which visually displays the operating status of the ROSE system. The system test shows that the Zabbix mail alarm module functions normally; the monitoring platform operates stably during the trial operation; and the ROSE system operation status evaluation is consistent with the actual situation.

Abstract:Weather Elements Nowcasting based on machine learning (WEN) has the characteristics of high release frequency, high time resolution, and complex forecast model based on climate and time. Using Python libraries such as multidimensional tag array, machine learning tools, parallel computing framework aiming at “fast computing”, a testing subsystem is established. It takes the time range covered by the “prediction model” as the statistical test time boundary. It objectively gives the prediction performance, which provides a basis for evaluating model optimization effect and the operation of products.

Abstract:Using GRAPESGFS forecast data and temperature observation data of Guangxi regional automatic weather stations during 2017-2018, errors of the 2 m temperature forecast of the GRAPESGFS model over Guangxi are analyzed. It is found that the 2 m temperature forecast of the GRAPESGFS model is lower than the observation in Guangxi. Forecast errors increase with the forecast time and regularly appear in the mountain areas in the northern Guangxi, Zuojiang, Youjiang river valley, and coastal areas. The temperature forecast error at noon is systematic in spring, summer and autumn but the errors at noon in winter and that at night in all seasons are random. To develop the optimal timewindow of the moving average method, we compare different moving average solutions with the unfixed timewindows and verify its improvement with the trial correction products of the optimal timewindow moving average method during 2020. Results show that the moving average solutions of fixed timewindow, optimal seasonal timewindow, and monthly optimal time window are all effective in spring, summer, and autumn. The correction skill is higher at noon than that at night. Among all the solutions, fixed long timewindow (15 to 60 d) solution, seasonal optimal timewindow solution and monthly optimal timewindow solution are more effective. Running optimal timewindow method based on different moving average solutions can steadily improve the 2 m temperature forecast.

Abstract:Taking Yangpu Port of Hainan as an example, using wind data from 2015 to 2019, the circulation characteristics, inducing mechanism and forecasting indicators of different kinds of High Wind Processes (HWPs) are investigated in this paper. Results suggest that: (1) HWPs of Yangpu Port can be divided into four types, namely coldair type, shearline type, tropicalcyclone (TC) type and thermallow type, according to their influencing systems. (2) The coldair HWP is mainly due to the advection of cold air. When the decreasing of the temperature in the critical area at 925 hPa is greater than 6 ℃ and the magnitude of north wind is greater than 11.5 m/s, it indicates that the coldair HWP will occur in 6 hours; the shearlinear HWP is mainly caused by thunderstorm gales and squall lines induced by severe convection; when favorable configuration appears in synopticscale systems such as 925 hPa low shearline and 500 hPa south trough, CAPE≥1500 J·kg-1 with appropriate CIN value, the shearlinear HWP is prone to occur; the thermallow HWP is closely related to the sea breeze front caused by the difference of sealand heat; when the difference of the increasing temperature between northwest inland and offshore sea surface in 6 hours is greater than or equal to 3.5 ℃, CAPE≥1500 J·kg-1, CIN≤20 J·kg-1, the sea breeze front is prone to trigger severe convection such as thunderstorms and gales; the TC HWP mainly occurs in the isobaric dense belt near the TC center and the mesoscale convective systems in the peripheral spiral rain belt. The wind scale can be determined synthetically based on the numerical forecast and combined with the statistical law, satellite, radar, and other monitoring data in the operational forecast.

Abstract:From 12 to 17 December 2019, a heavy pollution process with a long duration and extensive influence range of PM2.5 appeared in Hunan Province. This paper comprehensively applies the AQI hourly primary pollutant data and various meteorological observation data released by Hunan environmental monitoring stations. Changsha is selected as the representative station to deeply explore the meteorological conditions, evolution characteristics, and boundary layer characteristics of this heavy pollution weather process and explore the sources of pollutants and the moving path of the external source air mass. The results show that: this process was an inputsuperimposed local compound pollution event. In the pollution accumulation stage, the weak cold air opened up the pollution transmission channel, which was conducive to the pollutants from the north to the south. The pollution outbreak stage was mainly caused by the local static and stable weather superimposed by the upstream external sources. In the pollution removal stage, the ground in Hunan was mainly affected by strong cold air, the wind increased, and the pollutants were rapidly diluted and diffused. The boundary layer height was negatively correlated with AQI. The existence of nearsurface inversion was an essential condition for the pollution to continue and erupt. The backward trajectory shows that the process in the southern Hunan was the result of the heavy pollution weather process and the superposition of external and local sources.

Abstract:Based on the NCEP/NCAR reanalysis data and the AWOS data, the weather situation, water vapour conditions and boundary layer structure of two radiation fog processes at Zhaotong Airport on January 3 and December 7, 2019, are compared and analyzed. The commonness and differences of the formation and maintenance conditions of the two radiation fogs are discussed. The results show that: (1) The two fog events both occurred after the rain and snow weather turned clear, and the airport was controlled by the weak pressure field behind the highpressure bottom. However, the water vapour content and dynamic conditions of the “1.3” process were significantly better than those of the “12.7” process, and the fog concentration and duration were both greater than those of the “12.7” process. (2) The process of “1.3” was a typical radiation fog, which formed in the early morning when the weather was clear, the breeze was light, and the radiation temperature droped most strongly. In the process of the fog, if the water vapour was abundant and the dynamic conditions were suitable, it would be conducive to the formation and development of the fog. (3) The radiation fog in the process of 12.7 had the characteristics of local explosive development. Before the burst of the fog, only a light fog formed at the airport due to poor water vapour and dynamic conditions. After sunrise, due to the fog layer in the southwest direction moved northward, the airport’s low temperature and high humidity environmental conditions were conducive to the maintenance of the fog. Moreover, the nearground wind speed was conducive to the mixing of the fog layer, which led to the burst development of the fog in the airport area.

Abstract:The effect of typhoon circulation on remote rainstorms is complicated, and extreme precipitation is easy to occur, so it is difficult to forecast accurately the level and falling area. On August 8th, 2014, there were heavy rains from the east of Jiangsu to the north of Zhejiang. Using the numerical forecast situation, the NCEP final operational Global Data and actual observation data, we found collections between the heavy rains and the typhoon “XiaoLong”, 1300 km away on the western Pacific Ocean. The mesoscale vortex mainly caused the rainstorm in the eastern part of Jiangsu Province. The warm and moist air was transmitted to the rainstorm area by the easterly flow of typhoon north side, which was conducive to the formation and maintenance of the rainstorm. Then the rainstorm in the northern part of Zhejiang was caused by the combined effect of the mesoscale vortex and the typhoon circulation. In the process of typhoon moving northward, the distance between the typhoon centre and the vortex shortened, and then the typhoon circulation produced traction on the vortex, which changed the flow field situation. The northwest airflow brought weak cold air to the northern part of Zhejiang and met the warm air in lowlevel in the inshore region, causing convergence uplift and the rainstorm. Then, the convective instability energy broke out, resulting in the mesoscaleγ cyclone, which caused the extreme heavy precipitation.

Abstract:To accurately evaluate wind shear characteristics in the plain and lake areas of Hubei Province, the paper uses hourly wind data from 27 wind towers with the height between 120 m to 150 m and researches mainly on the temporal and spatial variations. The results show that: (1) The wind shear indexes have prominent characteristics in seasonal and diurnal changes. The wind shear indexes are generally minimum in autumn and winter, larger in summer, larger at night and smaller in the day. (2) The annual average wind shear index of 70 to 120 m is 0.27 in plain and lake areas and 0.12 in mountain areas. The wind shear indexes of each tower decrease gradually from 70 m to 150 m. (3) As for the wind shear indexes of 6 typical wind towers, the frequency of the relatively stable condition is the highest, which is 59% to 75%, and the lowest under unstable conditions, which is no more than 5%. (4) As for the wind speed ranges of 0 to 3.0 m/s, 3.0 to 10.0 m/s and above 10.0 m/s, the wind shear indexes gradually increase. The wind shear indexes in the wind speed range from 3.0 to 10.0 m/s are closest to the annual mean wind shear index, and the wind shear indexes above 10.0 m/s have the stronger dispersion. (5) The measured and simulated wind shear indexes are calculated, respectively. The measured values in the central plain are significantly higher than the simulated values, with the overall deviation ranging from -0.06 to 0.14. The conclusions of this paper can be used to calculate or revise the wind speed at different heights in the low wind speed area in the surface layer to improve the accuracy of wind energy resource evaluation, development planning, wind farm site selection and wind power prediction in this area.

Abstract:Based on the radiosonde data of Dunhuang, Jiuquan, Zhangye and Minqin stations from 2014 to 2019, the characteristics of cloud vertical structure in the northern Qilian Mountains are analyzed. The results show that the occurrence frequency of clouds in the northern Qilian Mountains is 20% to 40% in the whole year, and the occurrence frequency of singlelayer clouds is higher than that of multilayer clouds, which are mainly 2layer clouds; the occurrence frequency of clouds is higher in summer and lower in winter, the frequencies of singlelayer and 2layer clouds in summer are close, and the occurrence frequency of singlelayer clouds is much higher than those of 2layer clouds and 3layer clouds in spring, autumn and winter; the thickness of the lower layer clouds of 2layer clouds is obviously greater than that of the upper layer clouds, and the thickness of the clear air interlayer between the bottom layer and the middle layer of the 3layer clouds is greater than that between the middle layer and the top layer; cloud heights in the northern Qilian Mountains change significantly in different seasons, higher in summer and lower in winter.

Abstract:Precipitation over the lake surface is one of the essential factors affecting the water quantity and quality of the lake. This article uses precipitation observation data from 11 meteorological stations around Erhai Lake to analyze the precipitation characteristics around the Erhai Lake and based on the natural neighbour interpolation method, the precipitation over Erhai Lake is interpolated by the 0.01° resolution grid. The results show that the distribution of precipitation on the Erhai Lake has a seasonal feature. The precipitation is largest in summer, the second in autumn, and the smallest in winter. There is also a significantly uneven distribution in space. The highvalue centre exists in the central part of the Erhai Lake, nearby the west bank, and the lowvalue centre exists in the lake area near Haidong Station in the summer and autumn seasons. The spatial and temporal differences in precipitation on the lake surface are significant. This study has a specific reference value for the regional sensitivity study of the impact of precipitation on Erhai water quality and quantity. It offers the technical support for the precipitation prediction on water surface and the research and management of the water environment for the lake.

Abstract:Based on the fifthgeneration global atmospheric reanalysis data (ERA5) of the European Centre for Mediumrange Weather Forecasts (ECMWF) and the CWRMEM method of cloud water resources monitoring and evaluation published by the Center for Weather Modification of China Meteorological Administration, the cloud water resources in the Guangxi region are evaluated from 2009 to 2018. The results show that Guangxi’s average annual cloud water resources are about 5.1×1014kg, including 1.42×1014kg of retaining cloud water. The cloud water is mainly generated in the local region, and the average annual net output cloud water resources are about 1.97×1013kg. The cloud water resources of Guangxi are characterized by seasonal variation, showing a single peak distribution, with the highest value in June and the lowest value in February. However, the retain cloud water is not characterized by seasonal variation. The maximum value of the horizontal distribution of cloud water resources appears in the northeast of Guangxi and gradually decreases to the west and south. Cloud water mainly distributes in the lower layer of the atmosphere, about 925 to 600 hPa, mainly warm stratiform clouds composed of liquid droplets; however, in summer, which appears mainly in a higher layer, about 600 to 400 hPa of atmosphere, mainly mixed cloud water composed of supercooled liquid droplets and ice particles, and the cloud water significantly decreases in a lower layer.

Abstract:Based on the analysis of forest health indicators such as temperature, humidity, altitude, greening degree and air quality in the Panxi area, the evaluation system of forest health index is constructed using literature research method, expert scoring method and analytic hierarchy process. The results show that: temperature and air quality accounted for the highest proportion in all indicators of forest health. The Panxi area has fewer air pollutants, better air quality, higher greening degree, comfortable temperature and humidity, and the altitude is the best living height for human beings. The ArcGIS software is used to classify the health care services in the Panxi area. The results show that the best health care areas are located in the East, West, Renhe, Yanbian, Miyi of Panzhihua and the valley or low altitude areas of Dechang, Ningnan, Xichang, Huili, Huidong, Yanyuan and Muli of Liangshan Prefecture. These areas are the first choice for oldage care, rehabilitation and convalescence. Compared with other cities in China, such as Kunming and Sanya, this area has more advantages in health care. The findings of this study provide a necessary scientific basis for the sustainable development of the forest health industry in the Panxi area.