Abstract:
keep_len="250">To meet the requirement of numerical weather forecast for local severe convective weather, especially disastrous weather and extreme weather events, based on CMA-BJ V2.0 system, many works have been implemented, including increasing the model vertical layer to 59 layers, testing different physical parameterization schemes, assimilating unconventional local dense data such as wind profile radar and the near surface data, developing rapid cycle technology, and applying the incremental analysis update initialization technique of large-scale dynamic hybrid scheme for forecast field. By integrating all the jobs mentioned, the rapid analysis and forecast system CMA-BJ V2.0 has been established and put into operational run since June 2021 with 1 h time interval. A large number of tests and evaluations on multiple versions of the CMA-BJ numerical forecast system have been carried out. It is confirmed that the forecast skills of the model are improving year by year. There are still some problems in the forecast, such as heavy precipitation, high percentage of effective precipitation hours, and large deviation in the forecast of weak precipitation. Based upon 24 h precipitation forecast and 24 h hourly precipitation forecast information in North China on each day of the 2021 flood season (June to September), the comprehensive performance of CMA-BJ V2.0 forecast system with different resolutions (3 km and 9 km) is carefully evaluated and analyzed in terms of accumulation, percentage of effective precipitation hours, precipitation intensity, and daily cycle characteristics. The results show that both 9 km and 3 km resolutions can forecast the precipitation level and the rainfall area well and capture the regional distribution characteristics of precipitation with daily average precipitation greater than 8 mm well, but the forecast of precipitation level is larger than the observation. The forecast of hourly precipitation and the daily cycle of percentage of effective precipitation hours in North China is generally consistent with the observation, but the forecast of the peak in the evening is strong. The hourly precipitation is overestimated due to false alarms. For 3 km resolution forecast, the trend of percentage of effective precipitation hours is more similar to the observation. The magnitudes are closer to the observation than 9 km resolution forecast. 9 km resolution forecasts have better forecasting ability for weak precipitation processes, while 3 km resolution forecast is better at strong precipitation processes. The forecast results of a typical precipitation case in North China on 21 July 2021 are consistent with the test results of the average of the whole flood season: The model of both resolutions can better forecast the precipitation process, but the amount and percentage of effective precipitation hours is overestimated. To meet the requirement of numerical weather forecast for local severe convective weather, especially disastrous weather and extreme weather events, based on CMA-BJ V2.0 system, many works have been implemented, including increasing the model vertical layer to 59 layers, testing different physical parameterization schemes, assimilating unconventional local dense data such as wind profile radar and the near surface data, developing rapid cycle technology, and applying the incremental analysis update initialization technique of large-scale dynamic hybrid scheme for forecast field. By integrating all the jobs mentioned, the rapid analysis and forecast system CMA-BJ V2.0 has been established and put into operational run since June 2021 with 1 h time interval. A large number of tests and evaluations on multiple versions of the CMA-BJ numerical forecast system have been carried out. It is confirmed that the forecast skills of the model are improving year by year. There are still some problems in the forecast, such as heavy precipitation, high percentage of effective precipitation hours, and large deviation in the forecast of weak precipitation. Based upon 24 h precipitation forecast and 24 h hourly precipitation forecast information in North China on each day of the 2021 flood season (June to September), the comprehensive performance of CMA-BJ V2.0 forecast system with different resolutions (3 km and 9 km) is carefully evaluated and analyzed in terms of accumulation, percentage of effective precipitation hours, precipitation intensity, and daily cycle characteristics. The results show that both 9 km and 3 km resolutions can forecast the precipitation level and the rainfall area well and capture the regional distribution characteristics of precipitation with daily average precipitation greater than 8 mm well, but the forecast of precipitation level is larger than the observation. The forecast of hourly precipitation and the daily cycle of percentage of effective precipitation hours in North China is generally consistent with the observation, but the forecast of the peak in the evening is strong. The hourly precipitation is overestimated due to false alarms. For 3 km resolution forecast, the trend of percentage of effective precipitation hours is more similar to the observation. The magnitudes are closer to the observation than 9 km resolution forecast. 9 km resolution forecasts have better forecasting ability for weak precipitation processes, while 3 km resolution forecast is better at strong precipitation processes. The forecast results of a typical precipitation case in North China on 21 July 2021 are consistent with the test results of the average of the whole flood season: The model of both resolutions can better forecast the precipitation process, but the amount and percentage of effective precipitation hours is overestimated.

Abstract:
keep_len="250">In order to provide more powerful support for forecasters in Sichuan and Chongqing with complicated terrain to carry out short-term (0-12 h) precipitation forecast, a convective scale ensemble forecast operational system is designed based on ensemble Kalman filter data assimilation method (31 ensemble samples) and WRF model with 3 km resolution (model domain: 24.5°-34.5°N, 99°-113°E) and lead time of 12 h, which is started by 3 h cycle. It is urgent to decide how many members should be used for the 12 h ensemble forecast to achieve the most representative probability distribution and optimal ensemble forecast skills. An ensemble forecast experiment is carried out on 16 heavy convective scale precipitation cases occurred in Sichuan and Chongqing with different amount of ensemble members, and the results are analyzed comprehensively. It is concluded that the precipitation forecast skills of the ensemble members for different magnitude of precipitation are roughly the same, so there is little difference in the totally averaged prediction skills of different ensemble size. Talagrand distribution becomes better with the increase of ensemble size first. However, when the ensemble size is larger than 17, the improvement by increasing ensemble size is no longer significant. Meanwhile, the forecast error probability becomes smaller with the increase of ensemble size first, but when the ensemble size reaches 16 to 18, the difference between the forecast error probability and the ideal value tends to be stable, indicating that the improvement by further increasing the ensemble size is no longer significant. The relative area of operational characteristic (AROC) score which represents the prediction probability forecast skills, improves gradually with the increase of ensemble size. However, when the ensemble size is large enough, the improvement by lager ensemble size is no longer significant and the AROC scores tend to be stable. The ensemble size required for stable AROC score increases with the magnitude of precipitation. Overall, when the AROC scores become stable, the ensemble size required for light rain, moderate rain, heavy rain, rainstorm (and heavy rainstorm) are 10, 14, 16 and 18, respectively. Based on the comparative analysis results and considering that there is generally little difference in forecasting skills when the number of members is different by 2, in order to achieve the most representative probability distribution and optimal ensemble forecast skills of precipitation, it is recommended to set the ensemble size of convective scale ensemble prediction system from 16 to 18. In order to provide more powerful support for forecasters in Sichuan and Chongqing with complicated terrain to carry out short-term (0-12 h) precipitation forecast, a convective scale ensemble forecast operational system is designed based on ensemble Kalman filter data assimilation method (31 ensemble samples) and WRF model with 3 km resolution (model domain: 24.5°-34.5°N, 99°-113°E) and lead time of 12 h, which is started by 3 h cycle. It is urgent to decide how many members should be used for the 12 h ensemble forecast to achieve the most representative probability distribution and optimal ensemble forecast skills. An ensemble forecast experiment is carried out on 16 heavy convective scale precipitation cases occurred in Sichuan and Chongqing with different amount of ensemble members, and the results are analyzed comprehensively. It is concluded that the precipitation forecast skills of the ensemble members for different magnitude of precipitation are roughly the same, so there is little difference in the totally averaged prediction skills of different ensemble size. Talagrand distribution becomes better with the increase of ensemble size first. However, when the ensemble size is larger than 17, the improvement by increasing ensemble size is no longer significant. Meanwhile, the forecast error probability becomes smaller with the increase of ensemble size first, but when the ensemble size reaches 16 to 18, the difference between the forecast error probability and the ideal value tends to be stable, indicating that the improvement by further increasing the ensemble size is no longer significant. The relative area of operational characteristic (AROC) score which represents the prediction probability forecast skills, improves gradually with the increase of ensemble size. However, when the ensemble size is large enough, the improvement by lager ensemble size is no longer significant and the AROC scores tend to be stable. The ensemble size required for stable AROC score increases with the magnitude of precipitation. Overall, when the AROC scores become stable, the ensemble size required for light rain, moderate rain, heavy rain, rainstorm (and heavy rainstorm) are 10, 14, 16 and 18, respectively. Based on the comparative analysis results and considering that there is generally little difference in forecasting skills when the number of members is different by 2, in order to achieve the most representative probability distribution and optimal ensemble forecast skills of precipitation, it is recommended to set the ensemble size of convective scale ensemble prediction system from 16 to 18.

Abstract:
keep_len="250">The Nenjiang is the north source of the Songhua River. Nenjiang Basin is an important commodity grain base in China. The change of water level in Nenjiang Basin during the flood season is closely related to the precipitation, especially the continuous rainstorm and heavy rainstorm are very easy to cause flood disaster. For example, Nenjiang Basin is affected by the continuous rainstorm and heavy rainstorm weather on 18 July 2021, the Yong'an Reservoir bursted, the Xin'an Reservoir collapsed, residents across the towns are hit by the flood disaster. The flood in Nenjiang Basin has great impacts on the national economy and people's lives. Therefore, in order to improve the accuracy of rainstorm prediction in Nenjiang Basin, the deviation between CMA-TYM and SCMOC precipitation products are analyzed from the aspects of rainstorm area and intensity, and the correction ability is improved, which has certain practical significance for agricultural production, reservoir storage, and water resource allocation in the basin. At the same time, it also provides a strong guarantee for forecast warning, people's lives and property security, and sustainable healthy development of social. Nine rainstorm days are selected in 2021, using merged precipitation, based on numerical model products by CMA-TYM and SCMOC, the contiguous rain area (CRA) technique is used to test 24 h precipitation predicted at 2000 BT. The results show that maximum precipitation position deviation of rainstorm days predicted by CMA-TYM and SCMOC are west and north, but precipitation location of rainstorm days tested by CRA technique are west, the former is north, the latter is slightly south. SCMOC prediction preforms better than CMA-TYM. Error analysis show that, it is smaller than the precipitation observation that maximum precipitation value and average precipitation of observed rainstorm area predicted by CMA-TYM and SCMOC, but the grid numbers and area are larger than the observation. On the whole, CMA_TYM forecast is closer to the observation. CRA technique shows that the intensity and pattern of precipitation location predicted by CMA-TYM, location and pattern of precipitation predicted by SCMOC are close to the observation, and it has certain instructive significance. The Nenjiang is the north source of the Songhua River. Nenjiang Basin is an important commodity grain base in China. The change of water level in Nenjiang Basin during the flood season is closely related to the precipitation, especially the continuous rainstorm and heavy rainstorm are very easy to cause flood disaster. For example, Nenjiang Basin is affected by the continuous rainstorm and heavy rainstorm weather on 18 July 2021, the Yong'an Reservoir bursted, the Xin'an Reservoir collapsed, residents across the towns are hit by the flood disaster. The flood in Nenjiang Basin has great impacts on the national economy and people's lives. Therefore, in order to improve the accuracy of rainstorm prediction in Nenjiang Basin, the deviation between CMA-TYM and SCMOC precipitation products are analyzed from the aspects of rainstorm area and intensity, and the correction ability is improved, which has certain practical significance for agricultural production, reservoir storage, and water resource allocation in the basin. At the same time, it also provides a strong guarantee for forecast warning, people's lives and property security, and sustainable healthy development of social. Nine rainstorm days are selected in 2021, using merged precipitation, based on numerical model products by CMA-TYM and SCMOC, the contiguous rain area (CRA) technique is used to test 24 h precipitation predicted at 2000 BT. The results show that maximum precipitation position deviation of rainstorm days predicted by CMA-TYM and SCMOC are west and north, but precipitation location of rainstorm days tested by CRA technique are west, the former is north, the latter is slightly south. SCMOC prediction preforms better than CMA-TYM. Error analysis show that, it is smaller than the precipitation observation that maximum precipitation value and average precipitation of observed rainstorm area predicted by CMA-TYM and SCMOC, but the grid numbers and area are larger than the observation. On the whole, CMA_TYM forecast is closer to the observation. CRA technique shows that the intensity and pattern of precipitation location predicted by CMA-TYM, location and pattern of precipitation predicted by SCMOC are close to the observation, and it has certain instructive significance.

Abstract:
keep_len="250">The plateau vortex and the tropical cyclone over the Bay of Bengal have common active periods but there are few studies on the influence of their interaction on the plateau precipitation. Therefore, a large scale heavy precipitation process over the Tibetan Plateau under the joint effects of a tropical cyclone over the Bay of Bengal and plateau vortex is analyzed which occurs during 26-31 May 2017, based on the Joint Typhoon Warning Center (JTWC) best-track data, hourly precipitation observation data and combined with hybrid single-particle lagrangian integrated trajectory (HYSPLIT) model. The results show that, with the cooperation of tropical cyclone over the Bay of Bengal and the India-Burma though, the water vapor transport jet from the Bay of Bengal to the southeast of the Tibetan Plateau is established, providing water vapor for the low vortex and shear line of the plateau. The cold air behind the India-Burma though forms a cold cushion on the steep slope of the southern Tibetan Plateau, and the warm water vapor from the Bay of Bengal first rises northward along the cold cushion, then sinks after ascending to the plateau, and rises northward again near the low vortex and shear line of the plateau, which increases the precipitable water between the plateau surface and the upper troposphere atmosphere. Meanwhile, frontogenesis is generated by the confluence of tropical cyclone southerly warm water vapor with the cold and dry air in the northern part of the plateau. In the process of frontogenesis, the atmospheric wet baroclinicity increases significantly and the wet isentropic line rises sharply, which promotes the sharp development of vertical vorticity and the enhancement of plateau low vortex. During the northward movement, the anticyclonic outflow from the upper layer of the tropical cyclone strengthens the southwest jet in front of the upper trough of the Tibetan Plateau, and the enhancement of divergence favors the development of the plateau shear line and the eastward movement of the plateau low vortex, resulting in a large-scale heavy precipitation. On the other hand, the positive feedback of the water vapor transportation of the tropical cyclone over the Bay of Bengal and the shear line of the plateau vortex provides continuous apparent heat source and apparent moisture sink for the precipitation area, which is favorable for the maintenance and development of the precipitation system on the plateau. The plateau vortex and the tropical cyclone over the Bay of Bengal have common active periods but there are few studies on the influence of their interaction on the plateau precipitation. Therefore, a large scale heavy precipitation process over the Tibetan Plateau under the joint effects of a tropical cyclone over the Bay of Bengal and plateau vortex is analyzed which occurs during 26-31 May 2017, based on the Joint Typhoon Warning Center (JTWC) best-track data, hourly precipitation observation data and combined with hybrid single-particle lagrangian integrated trajectory (HYSPLIT) model. The results show that, with the cooperation of tropical cyclone over the Bay of Bengal and the India-Burma though, the water vapor transport jet from the Bay of Bengal to the southeast of the Tibetan Plateau is established, providing water vapor for the low vortex and shear line of the plateau. The cold air behind the India-Burma though forms a cold cushion on the steep slope of the southern Tibetan Plateau, and the warm water vapor from the Bay of Bengal first rises northward along the cold cushion, then sinks after ascending to the plateau, and rises northward again near the low vortex and shear line of the plateau, which increases the precipitable water between the plateau surface and the upper troposphere atmosphere. Meanwhile, frontogenesis is generated by the confluence of tropical cyclone southerly warm water vapor with the cold and dry air in the northern part of the plateau. In the process of frontogenesis, the atmospheric wet baroclinicity increases significantly and the wet isentropic line rises sharply, which promotes the sharp development of vertical vorticity and the enhancement of plateau low vortex. During the northward movement, the anticyclonic outflow from the upper layer of the tropical cyclone strengthens the southwest jet in front of the upper trough of the Tibetan Plateau, and the enhancement of divergence favors the development of the plateau shear line and the eastward movement of the plateau low vortex, resulting in a large-scale heavy precipitation. On the other hand, the positive feedback of the water vapor transportation of the tropical cyclone over the Bay of Bengal and the shear line of the plateau vortex provides continuous apparent heat source and apparent moisture sink for the precipitation area, which is favorable for the maintenance and development of the precipitation system on the plateau.

Abstract:
keep_len="250">China Meteorological Administration launched China's Global atmospheric reanalysis program in November 2013, and the global atmospheric reanalysis product (CMA-RA) has been successfully developed. The performance of CMA-RA on describing the circulation background of tropical cyclone activity affecting the South China Sea is analyzed and compared with ERA5 and NCEP-Ⅰ, exploring the applicability of CMA-RA in tropical cyclone activity analysis, based on the tropical cyclone best track dataset compiled by Shanghai Typhoon Research Institute of China Meteorological Administration, CMA-RA, the fifth generation ECMWF atmospheric reanalysis dataset (ERA5) and the first generation atmospheric monthly reanalysis dataset of National Center for Environmental Prediction(NCEP) and National Center for Atmospheric Research(NCAR) from 1981 to 2020. The results are shown as follows. Three reanalysis datasets can basically depict the anomaly circulation characteristics of the key influence regions closely related to tropical cyclone activity affecting the South China Sea from July to October, including the Southern Oscillation, low-level zonal wind field in the Philippines to the eastern sea of the South China Sea, reverse distribution pattern of low-level zonal wind filed in the tropics, low-level vorticity from the Philippines to the central and eastern part of the South China Sea, environmental vertical wind shear in the tropical western Pacific, and mid-level humidity field from the South China Sea to the eastern sea of the Philippines. All datasets are highly similar in describing the Southern Oscillation, low-level zonal wind field and mid-level humidity field of key regions. CMA-RA and ERA5 have high agreement on the Southern Oscillation, low-level zonal wind characteristics and their relationship with tropical cyclone activity, which are closer than NCEP-Ⅰ. However, their characterization of low-level meridional wind field, relative vorticity and vertical wind shear are relatively different. Some circulations in the tropical Indian Ocean are relatively different with each other too. All datasets have similar ability to depict the key regions circulations in the extreme years of tropical cyclone activity, but they are different in area and intensity. They are highly consistent in the sea level pressure and low-level zonal wind characteristics, with CMA-RA and ERA5 being the most similar. The mid-level humidity of CMA-RA is consistent with ERA5, and they are both lower than NCEP-Ⅰ. But the characteristics of low-level relative vorticity and vertical wind shear are significantly different. CMA-RA has comparable performance with ERA5 and NCEP-Ⅰ in describing the circulation background of tropical cyclone activity affecting the South China Sea, and it's highly consistent with ERA5 on the whole. Therefore, it can provide an alternative atmospheric reanalysis dataset for the research of tropical cyclone activity in the South China Sea. China Meteorological Administration launched China's Global atmospheric reanalysis program in November 2013, and the global atmospheric reanalysis product (CMA-RA) has been successfully developed. The performance of CMA-RA on describing the circulation background of tropical cyclone activity affecting the South China Sea is analyzed and compared with ERA5 and NCEP-Ⅰ, exploring the applicability of CMA-RA in tropical cyclone activity analysis, based on the tropical cyclone best track dataset compiled by Shanghai Typhoon Research Institute of China Meteorological Administration, CMA-RA, the fifth generation ECMWF atmospheric reanalysis dataset (ERA5) and the first generation atmospheric monthly reanalysis dataset of National Center for Environmental Prediction(NCEP) and National Center for Atmospheric Research(NCAR) from 1981 to 2020. The results are shown as follows. Three reanalysis datasets can basically depict the anomaly circulation characteristics of the key influence regions closely related to tropical cyclone activity affecting the South China Sea from July to October, including the Southern Oscillation, low-level zonal wind field in the Philippines to the eastern sea of the South China Sea, reverse distribution pattern of low-level zonal wind filed in the tropics, low-level vorticity from the Philippines to the central and eastern part of the South China Sea, environmental vertical wind shear in the tropical western Pacific, and mid-level humidity field from the South China Sea to the eastern sea of the Philippines. All datasets are highly similar in describing the Southern Oscillation, low-level zonal wind field and mid-level humidity field of key regions. CMA-RA and ERA5 have high agreement on the Southern Oscillation, low-level zonal wind characteristics and their relationship with tropical cyclone activity, which are closer than NCEP-Ⅰ. However, their characterization of low-level meridional wind field, relative vorticity and vertical wind shear are relatively different. Some circulations in the tropical Indian Ocean are relatively different with each other too. All datasets have similar ability to depict the key regions circulations in the extreme years of tropical cyclone activity, but they are different in area and intensity. They are highly consistent in the sea level pressure and low-level zonal wind characteristics, with CMA-RA and ERA5 being the most similar. The mid-level humidity of CMA-RA is consistent with ERA5, and they are both lower than NCEP-Ⅰ. But the characteristics of low-level relative vorticity and vertical wind shear are significantly different. CMA-RA has comparable performance with ERA5 and NCEP-Ⅰ in describing the circulation background of tropical cyclone activity affecting the South China Sea, and it's highly consistent with ERA5 on the whole. Therefore, it can provide an alternative atmospheric reanalysis dataset for the research of tropical cyclone activity in the South China Sea.

Abstract:
keep_len="250">The concentration of negative oxygen ions in air is an important index to evaluate the freshness and cleanliness of air. In recent years, it has become one hot topic concerned by governments and the public. From 2018 to 2021, Fujian has set up a number of observation stations for negative oxygen ions and meteorological factors over the entire province including seashore, mountain, humanities landscape areas, with good representativeness, reliability and continuity. Using the local observations, the spatial and temporal variations of negative oxygen ions concentration in Fujian is analyzed, and the negative oxygen ions concentration and grade prediction models are established based on multiple linear regression method and LightGBM machine learning method. The results show that, negative oxygen ions in Fujian is very rich and is very good for human health. The annual average concentration is between 708-8315 cm^-3, which is highest in high altitude, next in low altitude, and the concentration in middle altitude is the smallest. Overall, the annual average concentration of negative oxygen ions of nearly 80% site is beyond the standard of fresh air defined by World Health Organization. The diurnal variation of the concentration of negative oxygen ions show the characteristics of a peak and a trough, with the peak value mainly occurring at 0400-0600 BT and the trough value at 1200-1300 BT. The seasonal variation of negative oxygen ions concentration is more complex. The seasonal variation in the middle altitude area is greater, the seasonal average concentration in descending order is spring, summer, winter and autumn, while the seasonal variation in the high and low altitude area is relatively small. The main meteorological factors affecting the concentration of negative oxygen ions are temperature, humidity, precipitation, wind speed, air pressure and visibility. The concentration of negative oxygen ions is significantly positively correlated with humidity, precipitation and visibility at different altitudes, while the concentration of negative oxygen ions is significantly correlated with air temperature, wind speed and air pressure, but the correlation is different at different altitudes. The comparisons indicate the effects of LightGBM machine learning model are better than those of the traditional multiple linear regression model at different altitudes. The overestimation of negative oxygen ions concentration prediction is significantly improved, and the prediction grade of negative oxygen ions concentration can be improved by up to 12%. The results of logistic regression show that the traditional logistic regression basically has no predictive ability for small samples, while the LightGBM method has good learning ability in the case of small samples or unbalanced samples. The concentration of negative oxygen ions in air is an important index to evaluate the freshness and cleanliness of air. In recent years, it has become one hot topic concerned by governments and the public. From 2018 to 2021, Fujian has set up a number of observation stations for negative oxygen ions and meteorological factors over the entire province including seashore, mountain, humanities landscape areas, with good representativeness, reliability and continuity. Using the local observations, the spatial and temporal variations of negative oxygen ions concentration in Fujian is analyzed, and the negative oxygen ions concentration and grade prediction models are established based on multiple linear regression method and LightGBM machine learning method. The results show that, negative oxygen ions in Fujian is very rich and is very good for human health. The annual average concentration is between 708-8315 cm^-3, which is highest in high altitude, next in low altitude, and the concentration in middle altitude is the smallest. Overall, the annual average concentration of negative oxygen ions of nearly 80% site is beyond the standard of fresh air defined by World Health Organization. The diurnal variation of the concentration of negative oxygen ions show the characteristics of a peak and a trough, with the peak value mainly occurring at 0400-0600 BT and the trough value at 1200-1300 BT. The seasonal variation of negative oxygen ions concentration is more complex. The seasonal variation in the middle altitude area is greater, the seasonal average concentration in descending order is spring, summer, winter and autumn, while the seasonal variation in the high and low altitude area is relatively small. The main meteorological factors affecting the concentration of negative oxygen ions are temperature, humidity, precipitation, wind speed, air pressure and visibility. The concentration of negative oxygen ions is significantly positively correlated with humidity, precipitation and visibility at different altitudes, while the concentration of negative oxygen ions is significantly correlated with air temperature, wind speed and air pressure, but the correlation is different at different altitudes. The comparisons indicate the effects of LightGBM machine learning model are better than those of the traditional multiple linear regression model at different altitudes. The overestimation of negative oxygen ions concentration prediction is significantly improved, and the prediction grade of negative oxygen ions concentration can be improved by up to 12%. The results of logistic regression show that the traditional logistic regression basically has no predictive ability for small samples, while the LightGBM method has good learning ability in the case of small samples or unbalanced samples.

Abstract:
keep_len="250">The knowledge of accurate cloud heights (including cloud base height and cloud top height) information and its variation is of great importance to elucidating synoptic variation and improving climate model and prediction precision. Utilizing the theory of variation continuity and first-order discontinuity of meteorological element in frontal zone, cloud front zone is defined as transitional zone between the cloud cluster and its adjacent area in vertical direction in order to solve the problem of cloud heights uncertainties observed by different equipments. According to the humidity, scattering and turbulence properties of cloud, using observation from L-band sounding, Ka-band millimeter wave cloud radar (MMCR) and the wind profiler, the variation characteristics of temperature, humidity, radar reflectivity and signal noise ratio (SNR) as well as their differences from the ambient atmosphere are studied. In addition, the differences between convective clouds and stratified clouds are studied in terms of the characteristics of element gradient variation inside and outside clouds. Finally, the identification for cloud front zone is verified by case study and the reasonable scope and identification criterion for cloud base height and cloud top height are concluded. The results show that the first-order and second-order derivative of temperature, humidity, and radar reflectivity are discontinuous in cloud front zone (they are not equal inside and outside the cloud front region), and the vertical gradient of SNR retrieved by wind profiler is also instable, which shows that the cloud boundary range with better spatial consistency can be obtained by different devices, based on the frontal theory. In addition, there are two indicators that can be utilized to distinguish the stratiform clouds from convective clouds. The first is the difference between the vertical gradient of temperature and humidity in clouds and that in ambient atmosphere, which is larger in convective clouds than that in stratiform clouds. The second is the fluctuation amplitude of the second-order derivative of reflectivity in clouds, which is also larger in convective clouds than that in stratiform clouds. The concept of cloud front zone can be used to comprehensively identify the common range of cloud height detected by different devices, indicating that there are consistent variation characteristics in a certain area near the cloud front zone for different devices. The similarity of cloud vertical structures retrieved by multi-source equipment observation are elucidated through the characteristics of cloud front zone, which is worth applying for collaborative observation of different devices. The knowledge of accurate cloud heights (including cloud base height and cloud top height) information and its variation is of great importance to elucidating synoptic variation and improving climate model and prediction precision. Utilizing the theory of variation continuity and first-order discontinuity of meteorological element in frontal zone, cloud front zone is defined as transitional zone between the cloud cluster and its adjacent area in vertical direction in order to solve the problem of cloud heights uncertainties observed by different equipments. According to the humidity, scattering and turbulence properties of cloud, using observation from L-band sounding, Ka-band millimeter wave cloud radar (MMCR) and the wind profiler, the variation characteristics of temperature, humidity, radar reflectivity and signal noise ratio (SNR) as well as their differences from the ambient atmosphere are studied. In addition, the differences between convective clouds and stratified clouds are studied in terms of the characteristics of element gradient variation inside and outside clouds. Finally, the identification for cloud front zone is verified by case study and the reasonable scope and identification criterion for cloud base height and cloud top height are concluded. The results show that the first-order and second-order derivative of temperature, humidity, and radar reflectivity are discontinuous in cloud front zone (they are not equal inside and outside the cloud front region), and the vertical gradient of SNR retrieved by wind profiler is also instable, which shows that the cloud boundary range with better spatial consistency can be obtained by different devices, based on the frontal theory. In addition, there are two indicators that can be utilized to distinguish the stratiform clouds from convective clouds. The first is the difference between the vertical gradient of temperature and humidity in clouds and that in ambient atmosphere, which is larger in convective clouds than that in stratiform clouds. The second is the fluctuation amplitude of the second-order derivative of reflectivity in clouds, which is also larger in convective clouds than that in stratiform clouds. The concept of cloud front zone can be used to comprehensively identify the common range of cloud height detected by different devices, indicating that there are consistent variation characteristics in a certain area near the cloud front zone for different devices. The similarity of cloud vertical structures retrieved by multi-source equipment observation are elucidated through the characteristics of cloud front zone, which is worth applying for collaborative observation of different devices.

Abstract:
keep_len="250">Satellite cloud images contain abundant information, which can reflect daily weather conditions. Nowcasting based on cloud images can strengthen the application of satellite data in the early warning and forecasting of severe weather. At present, the cloud images predicted by most nowcasting methods based on artificial intelligence are not accurate enough and the lead time is limited. Thus, it's necessary to improve the accuracy and period validity of cloud images in nowcasting. Using the infrared cloud image data of Fengyun-4A (FY-4A) and the generative adversarial network (GAN) method, an infrared cloud image extrapolation nowcasting model is proposed. The cloud images in the next 3 hours in East China are predicted by the proposed model, and the spatial resolution of predicted cloud images is 4 km and the temporal resolution is 1 hour. The results show that the evaluation values of SSIM (structural similarity), PSNR (peak signal to noise ratio) and RMSE (root mean square error) predicted by the proposed GAN-based cloud images extrapolation model are 0.75, 20.92 and 10.00 K, respectively. In addition, the MAE (mean absolute error), MSE (mean squared error), and SSIM are chosen as loss function and analyzed, aiming to verify the rationality of the loss function in the generator. Comparative experiments of different loss functions show that it is reasonable and effective to choose SSIM combined with MAE as the loss function. Furthermore, to verify the effectiveness of the proposed GAN-based model, the prediction results are compared with those of the optical flow method and the TrajGRU model with the GAN-based model. The experimental results show that the cloud image extrapolation model based on GAN has the superior prediction performance, with the highest SSIM and PSNR, and the lowest RMSE within 1-3 h of cloud images nowcasting. The observational examples show that the cloud images predicted by the proposed model can well describe the movement, development and dissipation trend of clouds. Meanwhile, the experiments obtain accurate prediction performance on the intensity, location and shape of clouds in the study region. It indicates that the cloud image extrapolation model based on GAN is rational and feasible, which can be effectively applied to the meteorological business to monitor the occurrence and movement of clouds and warn the occurrence of severe weather in advance, and can provide an important reference for weather forecasting. Satellite cloud images contain abundant information, which can reflect daily weather conditions. Nowcasting based on cloud images can strengthen the application of satellite data in the early warning and forecasting of severe weather. At present, the cloud images predicted by most nowcasting methods based on artificial intelligence are not accurate enough and the lead time is limited. Thus, it's necessary to improve the accuracy and period validity of cloud images in nowcasting. Using the infrared cloud image data of Fengyun-4A (FY-4A) and the generative adversarial network (GAN) method, an infrared cloud image extrapolation nowcasting model is proposed. The cloud images in the next 3 hours in East China are predicted by the proposed model, and the spatial resolution of predicted cloud images is 4 km and the temporal resolution is 1 hour. The results show that the evaluation values of SSIM (structural similarity), PSNR (peak signal to noise ratio) and RMSE (root mean square error) predicted by the proposed GAN-based cloud images extrapolation model are 0.75, 20.92 and 10.00 K, respectively. In addition, the MAE (mean absolute error), MSE (mean squared error), and SSIM are chosen as loss function and analyzed, aiming to verify the rationality of the loss function in the generator. Comparative experiments of different loss functions show that it is reasonable and effective to choose SSIM combined with MAE as the loss function. Furthermore, to verify the effectiveness of the proposed GAN-based model, the prediction results are compared with those of the optical flow method and the TrajGRU model with the GAN-based model. The experimental results show that the cloud image extrapolation model based on GAN has the superior prediction performance, with the highest SSIM and PSNR, and the lowest RMSE within 1-3 h of cloud images nowcasting. The observational examples show that the cloud images predicted by the proposed model can well describe the movement, development and dissipation trend of clouds. Meanwhile, the experiments obtain accurate prediction performance on the intensity, location and shape of clouds in the study region. It indicates that the cloud image extrapolation model based on GAN is rational and feasible, which can be effectively applied to the meteorological business to monitor the occurrence and movement of clouds and warn the occurrence of severe weather in advance, and can provide an important reference for weather forecasting.

Abstract:
keep_len="250">Hail refers to the solid precipitation with a diameter greater than or equal to 5 mm caused by convection. Hail is also one main disastrous weather phenomenon in eastern Hubei, while Doppler weather radar is the most favorable tool for hail identification. At present, there are two hail identification methods used in the actual operation in eastern Hubei, one is artificial conceptual model, the other is the self-contained identification technology in short-time and proximity prediction system. The conceptual model needs to be judged by human, which is too subjective and threshold values of radar echo characteristic are not clear, while the false alarm rate(FAR) of existing automatic technologies in prediction system are too high. To overcome the shortcoming of the above methods, feasibilities of machine learning algorithms for hail identification are explored and a decision tree algorithm is established. Based on hail disaster data of Wuhan, Huanggang, Huangshi, Ezhou, Xianning and Xiaogan, Doppler weather radar data and convention high altitude sounding data of Wuhan from 2015 to 2021, the height of wet bulb 0℃(HWB0) and the height of wet bulb -20℃(HWB-20) are introduced into the hail identification factors, and artificial intelligence technology is applied in hail recognition. The performance is evaluated according to probability of detection(POD), FAR and critical success index(CSI). The result shows that both the decision tree algorithm with radar echo intensity (intensity decision tree) and the decision tree algorithm with radar echo intensity and wet bulb temperature height (intensity-height decision tree) can identify hail effectively. The POD results of the two decision tree algorithms are higher than 0.88, while the FAR are lower than 0.12, and the CSI are higher than 0.8, but the intensity-height decision tree performs better, with the POD and CSI increased by 5.68% and 7.5% than intensity decision tree respectively, while the FAR decrease 41.67%. The key factor of hail recognition by intensity decision tree is the combined reflectivity factor, and the bottom layer is the reflectivity factor of 0.5° and 1.5° elevation. The key factor of intensity-height decision tree is the reflectivity factor of 0.5° elevation and the judgment modules of radar echo extension height with the height of wet bulb temperature, especially with HWB0 included in the middle, and the bottom layer is the strength attributes of storm (vertically integrated liquid water and combined reflectivity). The analysis results of three cases with different occurrence time, location and hail size show that, due to the introduction of height of wet bulb temperature, the intensity-height decision tree reduces the number of empty alarm when the height of HWB0 and HWB-20 are high, especially when the HWB0 is high, thus reduces its FAR and improves its CSI, which indicate its potential wide prospect for operational application. Hail refers to the solid precipitation with a diameter greater than or equal to 5 mm caused by convection. Hail is also one main disastrous weather phenomenon in eastern Hubei, while Doppler weather radar is the most favorable tool for hail identification. At present, there are two hail identification methods used in the actual operation in eastern Hubei, one is artificial conceptual model, the other is the self-contained identification technology in short-time and proximity prediction system. The conceptual model needs to be judged by human, which is too subjective and threshold values of radar echo characteristic are not clear, while the false alarm rate(FAR) of existing automatic technologies in prediction system are too high. To overcome the shortcoming of the above methods, feasibilities of machine learning algorithms for hail identification are explored and a decision tree algorithm is established. Based on hail disaster data of Wuhan, Huanggang, Huangshi, Ezhou, Xianning and Xiaogan, Doppler weather radar data and convention high altitude sounding data of Wuhan from 2015 to 2021, the height of wet bulb 0℃(HWB0) and the height of wet bulb -20℃(HWB-20) are introduced into the hail identification factors, and artificial intelligence technology is applied in hail recognition. The performance is evaluated according to probability of detection(POD), FAR and critical success index(CSI). The result shows that both the decision tree algorithm with radar echo intensity (intensity decision tree) and the decision tree algorithm with radar echo intensity and wet bulb temperature height (intensity-height decision tree) can identify hail effectively. The POD results of the two decision tree algorithms are higher than 0.88, while the FAR are lower than 0.12, and the CSI are higher than 0.8, but the intensity-height decision tree performs better, with the POD and CSI increased by 5.68% and 7.5% than intensity decision tree respectively, while the FAR decrease 41.67%. The key factor of hail recognition by intensity decision tree is the combined reflectivity factor, and the bottom layer is the reflectivity factor of 0.5° and 1.5° elevation. The key factor of intensity-height decision tree is the reflectivity factor of 0.5° elevation and the judgment modules of radar echo extension height with the height of wet bulb temperature, especially with HWB0 included in the middle, and the bottom layer is the strength attributes of storm (vertically integrated liquid water and combined reflectivity). The analysis results of three cases with different occurrence time, location and hail size show that, due to the introduction of height of wet bulb temperature, the intensity-height decision tree reduces the number of empty alarm when the height of HWB0 and HWB-20 are high, especially when the HWB0 is high, thus reduces its FAR and improves its CSI, which indicate its potential wide prospect for operational application.

Abstract:
keep_len="250">Interannual variation in net ecosystem carbon production (NEP) plays an important role in the carbon cycle processes. An agricultural ecosystem may fluctuate between carbon net source and carbon sink, or it may remain neutral. Thus, the long-term trends in NEP and the relevant meteorological, soil and biotic control of interannual variation in NEP remain unclear in farmland agroecosystems. To effectively assess the carbon sequestration potential of the farmland ecosystem, the eddy covariance dataset of rain-fed spring maize in Northeast China from 2005 to 2020 are used to investigate the interannual variations in NEP and the relevant meteorological, soil and biotic influences. NEP is partitioned into gross ecosystem productivity (GEP) and ecosystem respiration (RE) to explain the interannual variations of NEP and its influencing factors. The average annual NEP, GEP and RE are 272±109, 1086±177, 820±130 g·m^-2·a^-1, respectively, with no significant changes. The day-to-day dynamics of NEP, GEP and RE show single peak curves. NEP and GEP reach the maximums at the very time of maize tasseling, and the maximum value of RE occurs 13 days after NEP and GEP. Compared with RE, NEP variations are mainly caused by GEP. The redundancy analysis shows the interannual variations in NEP are mainly affected by precipitation as the meteorological factor and water use efficiency as the biotic factor, and the influence weights of the meteorological and biotic factors are 28.4% and 31.4%. Meanwhile, the influence weights of the meteorological factors (photosynthetically active radiation, carbon dioxide and precipitation), soil (soil volumetric water content and soil organic carbon) and biotic factors (leaf area index and water use efficiency) are 61.0%, 43.8% and 62.8% for the interannual variations in GEP. The interannual variations in RE are mainly affected by the soil (soil volumetric water content and soil organic carbon) and the biotic factors (leaf area index), and the influence weight of the soil factors (39.3%) is larger than that of the biotic factor (29.2%). The results indicate that, under the background of climate warming, interannual variations in NEP in rain-fed spring maize agroecosystems are likely to be more sensitive to changes in moisture, while radiation and temperature will contribute to interannual NEP variations by affecting vapor pressure difference and soil water content. Interannual variation in net ecosystem carbon production (NEP) plays an important role in the carbon cycle processes. An agricultural ecosystem may fluctuate between carbon net source and carbon sink, or it may remain neutral. Thus, the long-term trends in NEP and the relevant meteorological, soil and biotic control of interannual variation in NEP remain unclear in farmland agroecosystems. To effectively assess the carbon sequestration potential of the farmland ecosystem, the eddy covariance dataset of rain-fed spring maize in Northeast China from 2005 to 2020 are used to investigate the interannual variations in NEP and the relevant meteorological, soil and biotic influences. NEP is partitioned into gross ecosystem productivity (GEP) and ecosystem respiration (RE) to explain the interannual variations of NEP and its influencing factors. The average annual NEP, GEP and RE are 272±109, 1086±177, 820±130 g·m^-2·a^-1, respectively, with no significant changes. The day-to-day dynamics of NEP, GEP and RE show single peak curves. NEP and GEP reach the maximums at the very time of maize tasseling, and the maximum value of RE occurs 13 days after NEP and GEP. Compared with RE, NEP variations are mainly caused by GEP. The redundancy analysis shows the interannual variations in NEP are mainly affected by precipitation as the meteorological factor and water use efficiency as the biotic factor, and the influence weights of the meteorological and biotic factors are 28.4% and 31.4%. Meanwhile, the influence weights of the meteorological factors (photosynthetically active radiation, carbon dioxide and precipitation), soil (soil volumetric water content and soil organic carbon) and biotic factors (leaf area index and water use efficiency) are 61.0%, 43.8% and 62.8% for the interannual variations in GEP. The interannual variations in RE are mainly affected by the soil (soil volumetric water content and soil organic carbon) and the biotic factors (leaf area index), and the influence weight of the soil factors (39.3%) is larger than that of the biotic factor (29.2%). The results indicate that, under the background of climate warming, interannual variations in NEP in rain-fed spring maize agroecosystems are likely to be more sensitive to changes in moisture, while radiation and temperature will contribute to interannual NEP variations by affecting vapor pressure difference and soil water content.