The accurate representation of mixed-phase monsoon clouds and their phase distribution is of great importance for numerical models used to predict monsoon rainfall. Therefore, it is essential for these models to correctly capture the phase fraction of clouds, which includes the proportions of liquid and ice. Ice particle formation in clouds occurs through primary ice production and secondary ice production (SIP). Most weather and climate models tend to overlook secondary SIP mechanisms, often only including rime-splintering. This oversight can introduce biases in the phase partitioning of mixed-phase clouds and monsoon rainfall predictions.
{"title":"Importance of secondary ice production in mixed-phase monsoon clouds over the Indian subcontinent","authors":"Sachin Patade, Gayatri Kulkarni, Sonali Patade, Deepak Waman, Georgia Sotiropoulou, Soumya Samanta, Neelam Malap, Thara Prabhakaran","doi":"10.1016/j.atmosres.2024.107890","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107890","url":null,"abstract":"The accurate representation of mixed-phase monsoon clouds and their phase distribution is of great importance for numerical models used to predict monsoon rainfall. Therefore, it is essential for these models to correctly capture the phase fraction of clouds, which includes the proportions of liquid and ice. Ice particle formation in clouds occurs through primary ice production and secondary ice production (SIP). Most weather and climate models tend to overlook secondary SIP mechanisms, often only including rime-splintering. This oversight can introduce biases in the phase partitioning of mixed-phase clouds and monsoon rainfall predictions.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"88 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142918009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Short-term forecasting of extreme weather is crucial for disaster warning and prevention. Many extreme weather events are often accompanied by significant water vapor changes, therefore, assimilating high-precision, high-resolution water vapor observations into numerical models is essential. This study explores the impact of GNSS ZTD assimilation on short-term forecasting of extreme weather using the WRF model on the case of “21.7” Henan extreme heavy rain. The impacts of GNSS ZTD assimilation on model fields and forecast results are analyzed, compared with scenarios where no data or only conventional observational data are assimilated. The results indicate that GNSS products outperform radiosonde data in temporal and spatial resolution, significantly affecting humidity fields in assimilation and providing more detailed water vapor distribution. In terms of precipitation forecasting, the analysis of POD, FAR, and ETS scores shows that GNSS data assimilation primarily impacts moderate to heavy rainfall for this case. During most simulation periods, the scores are higher when GNSS products are assimilated, with the most notable improvements observed at the threshold of 30 mm for 3-h accumulated precipitation, where ETS scores increase by an average of 21 %. However, despite the general improvement in precipitation forecast accuracy, limitations remain in forecasting peak rainfall periods.
{"title":"Improving forecast of “21.7” Henan extreme heavy rain by assimilating high spatial resolution GNSS ZTDs","authors":"Mengjie Liu, Yidong Lou, Weixing Zhang, Rong Wan, Zhenyi Zhang, Zhikang Fu, Xiaohong Zhang","doi":"10.1016/j.atmosres.2024.107880","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107880","url":null,"abstract":"Short-term forecasting of extreme weather is crucial for disaster warning and prevention. Many extreme weather events are often accompanied by significant water vapor changes, therefore, assimilating high-precision, high-resolution water vapor observations into numerical models is essential. This study explores the impact of GNSS ZTD assimilation on short-term forecasting of extreme weather using the WRF model on the case of “21.7” Henan extreme heavy rain. The impacts of GNSS ZTD assimilation on model fields and forecast results are analyzed, compared with scenarios where no data or only conventional observational data are assimilated. The results indicate that GNSS products outperform radiosonde data in temporal and spatial resolution, significantly affecting humidity fields in assimilation and providing more detailed water vapor distribution. In terms of precipitation forecasting, the analysis of POD, FAR, and ETS scores shows that GNSS data assimilation primarily impacts moderate to heavy rainfall for this case. During most simulation periods, the scores are higher when GNSS products are assimilated, with the most notable improvements observed at the threshold of 30 mm for 3-h accumulated precipitation, where ETS scores increase by an average of 21 %. However, despite the general improvement in precipitation forecast accuracy, limitations remain in forecasting peak rainfall periods.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"2 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888660","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-25DOI: 10.1016/j.atmosres.2024.107883
Lijie Zhang, Shanshan Wang, Yang Wang, Miao Lei, Yang Zhao, Jianjun He
Precipitation and its diurnal cycle are closely connected with the thermodynamic and dynamic processes of the Earth system, significantly influencing the climate. This study evaluates the performance of reanalysis and CMIP6 datasets in reproducing hourly precipitation events during warm seasons across China from 1980 to 2014. Statistical results indicate that while these datasets better reproduce the distribution of mean intensity than mean duration, neither fully captures the trends in mean duration or intensity. The reanalysis and CMIP6 datasets can reproduce the late afternoon precipitation peak, but it is difficult to capture the early morning peak of long-duration events except for ERA5 and HadGEM3-GC31-MM. Short-duration events show advanced peak time trends over most stations, while long-duration events exhibit delayed trends, a pattern not comprehensively replicated by other datasets. Additionally, the impact of aerosols on precipitation peak times varies across three regions: North China Plain (NCP), Yangtze River Delta (YRD) and the Pearl River Delta (PRD). In the NCP, early morning and midnight peaks advance for precipitation that lasts 4–6 h, potentially linked to aerosol radiative effects. In contrast, in the YRD and PRD, both early morning and late afternoon peaks are delayed 1–2 h, associated with the radiative and microphysical effects of aerosols. This study also highlights that aerosol impacts on precipitation peak times are dependent on meteorological conditions. In the NCP, the radiative effect of absorbing aerosols is enhanced under low-CAPE conditions. In the YRD, the aerosol invigoration effect is inhibited under high-WS conditions, whereas in the PRD, a low-WS environment enhances the microphysical effect of aerosols.
{"title":"Long-term variations in diurnal precipitation pattern and their attribution to aerosols across China","authors":"Lijie Zhang, Shanshan Wang, Yang Wang, Miao Lei, Yang Zhao, Jianjun He","doi":"10.1016/j.atmosres.2024.107883","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107883","url":null,"abstract":"Precipitation and its diurnal cycle are closely connected with the thermodynamic and dynamic processes of the Earth system, significantly influencing the climate. This study evaluates the performance of reanalysis and CMIP6 datasets in reproducing hourly precipitation events during warm seasons across China from 1980 to 2014. Statistical results indicate that while these datasets better reproduce the distribution of mean intensity than mean duration, neither fully captures the trends in mean duration or intensity. The reanalysis and CMIP6 datasets can reproduce the late afternoon precipitation peak, but it is difficult to capture the early morning peak of long-duration events except for ERA5 and HadGEM3-GC31-MM. Short-duration events show advanced peak time trends over most stations, while long-duration events exhibit delayed trends, a pattern not comprehensively replicated by other datasets. Additionally, the impact of aerosols on precipitation peak times varies across three regions: North China Plain (NCP), Yangtze River Delta (YRD) and the Pearl River Delta (PRD). In the NCP, early morning and midnight peaks advance for precipitation that lasts 4–6 h, potentially linked to aerosol radiative effects. In contrast, in the YRD and PRD, both early morning and late afternoon peaks are delayed 1–2 h, associated with the radiative and microphysical effects of aerosols. This study also highlights that aerosol impacts on precipitation peak times are dependent on meteorological conditions. In the NCP, the radiative effect of absorbing aerosols is enhanced under low-CAPE conditions. In the YRD, the aerosol invigoration effect is inhibited under high-WS conditions, whereas in the PRD, a low-WS environment enhances the microphysical effect of aerosols.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"28 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142917986","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.atmosres.2024.107886
Inmaculada Foyo-Moreno, Ismael L. Lozano, Inmaculada Alados, Juan Luis Guerrero-Rascado
Net surface radiation is a crucial parameter across various fields, as it represents the available energy for the energy exchange between the surface and the atmosphere. This work presents a new model for estimating instantaneous daytime net surface radiation (Rn) under all sky conditions, using solar position via cos ϴz and the clearness index (kt) as predictors. Global solar radiation (G↓) is the primary factor influencing Rn and is extensively measured at numerous radiometric stations. Consequently, this model takes advantage of using a single input (G↓). The model was validated against other empirical models at various sites with diverse climatological characteristics. Two types of models were evaluated, one including reflected global solar irradiance (G↑) as an additional input variable alongside G↓. The best results were obtained when incorporating G↑. However, this poses a challenge as G↑ is not measured at most radiometric stations. Nevertheless, in both types, the simplest model consistently outperformed the others, revealing no significant improvements with the addition of extra variables. Overall, the proposed model demonstrated good fit with the experimental data, although with some overestimation. The coefficient of determination (R2) is over 0,94, except at sites with extreme surface albedo conditions (α > 0,55). Mean bias error values ranged from 4 Wm−2 to 44 Wm−2, while root mean square error values varied from 25 Wm−2 to 62 Wm−2. Additional assessments across different seasons and sky conditions revealed improved performance during colder seasons and under cloudy conditions. Finally, the statistical analysis of the proposed model falls within the range of other more sophisticated models that involve additional input variables.
{"title":"A new model to estimate daytime net surface radiation under all sky conditions","authors":"Inmaculada Foyo-Moreno, Ismael L. Lozano, Inmaculada Alados, Juan Luis Guerrero-Rascado","doi":"10.1016/j.atmosres.2024.107886","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107886","url":null,"abstract":"Net surface radiation is a crucial parameter across various fields, as it represents the available energy for the energy exchange between the surface and the atmosphere. This work presents a new model for estimating instantaneous daytime net surface radiation (R<ce:inf loc=\"post\">n</ce:inf>) under all sky conditions, using solar position via cos ϴ<ce:inf loc=\"post\">z</ce:inf> and the clearness index (k<ce:inf loc=\"post\">t</ce:inf>) as predictors. Global solar radiation (G<ce:inf loc=\"post\">↓</ce:inf>) is the primary factor influencing R<ce:inf loc=\"post\">n</ce:inf> and is extensively measured at numerous radiometric stations. Consequently, this model takes advantage of using a single input (G<ce:inf loc=\"post\">↓</ce:inf>). The model was validated against other empirical models at various sites with diverse climatological characteristics. Two types of models were evaluated, one including reflected global solar irradiance (G<ce:inf loc=\"post\">↑</ce:inf>) as an additional input variable alongside G<ce:inf loc=\"post\">↓</ce:inf>. The best results were obtained when incorporating G<ce:inf loc=\"post\">↑</ce:inf>. However, this poses a challenge as G<ce:inf loc=\"post\">↑</ce:inf> is not measured at most radiometric stations. Nevertheless, in both types, the simplest model consistently outperformed the others, revealing no significant improvements with the addition of extra variables. Overall, the proposed model demonstrated good fit with the experimental data, although with some overestimation. The coefficient of determination (R<ce:sup loc=\"post\">2</ce:sup>) is over 0,94, except at sites with extreme surface albedo conditions (α > 0,55). Mean bias error values ranged from 4 Wm<ce:sup loc=\"post\">−2</ce:sup> to 44 Wm<ce:sup loc=\"post\">−2</ce:sup>, while root mean square error values varied from 25 Wm<ce:sup loc=\"post\">−2</ce:sup> to 62 Wm<ce:sup loc=\"post\">−2</ce:sup>. Additional assessments across different seasons and sky conditions revealed improved performance during colder seasons and under cloudy conditions. Finally, the statistical analysis of the proposed model falls within the range of other more sophisticated models that involve additional input variables.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"15 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142917990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.atmosres.2024.107888
Kalliopi-Mikaela Papa, Aristeidis G. Koutroulis
The spatiotemporal precipitation patterns in Greece are influenced by several factors, including the complex topography and the multifaceted climatic regimes of the country. Rain gauges, albeit a reliable tool for the accurate quantification of precipitation, are scarce, sporadic, and not properly maintained. In these instances, gridded datasets may provide a solution by administering spatially and temporally continuous precipitation data. The products, however, reveal limitations in the realistic simulation of precipitation, primarily caused by the intrinsic flaws of the underlying methods used. The assessment of eight of the most spatially and temporally detailed precipitation datasets, namely ERA5-Land (ERA5L), AgERA5, CHELSA-W5E5 v1.1 (CHELSA), MSWEP V2.8, CHIRPS05, IMERG V06 (Final), and E-OBS, compared against field observations acquired from 304 gauging stations across Greece has not been previously attempted. The evaluation is conducted on a daily and a monthly timescale, over a 32-year period (1984–2016), assessing the performance of the gridded products by considering both the country as a whole and its individual regions. The ability of the datasets to correctly portray the occurrence of extreme events and precipitation patterns is examined by statistical metrics and further insights are provided by the application and statistical analysis of climate indices on ground observations. CHELSA, CERRAL and AgERA5 consistently yield acceptable results across statistical metrics, outperforming the other datasets, which exhibit inferior performance in both temporal scales. The statistical analysis reveals distinct patterns of heavier precipitation in northern and western regions, with strong seasonal variability in the West and South and a possible average decennial increase of over 110 mm in mean annual and over 30 mm in extreme precipitation, along the assessment period. Overall, the datasets fail to accurately depict precipitable extremes, but CHELSA and CERRAL stand out as more reliable options for describing the precipitation dynamics in Greece.
{"title":"Evaluation of precipitation datasets over Greece. Insights from comparing multiple gridded products with observations","authors":"Kalliopi-Mikaela Papa, Aristeidis G. Koutroulis","doi":"10.1016/j.atmosres.2024.107888","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107888","url":null,"abstract":"The spatiotemporal precipitation patterns in Greece are influenced by several factors, including the complex topography and the multifaceted climatic regimes of the country. Rain gauges, albeit a reliable tool for the accurate quantification of precipitation, are scarce, sporadic, and not properly maintained. In these instances, gridded datasets may provide a solution by administering spatially and temporally continuous precipitation data. The products, however, reveal limitations in the realistic simulation of precipitation, primarily caused by the intrinsic flaws of the underlying methods used. The assessment of eight of the most spatially and temporally detailed precipitation datasets, namely ERA5-Land (ERA5L), AgERA5, CHELSA-W5E5 v1.1 (CHELSA), MSWEP V2.8, CHIRPS05, IMERG V06 (Final), and <ce:italic>E</ce:italic>-OBS, compared against field observations acquired from 304 gauging stations across Greece has not been previously attempted. The evaluation is conducted on a daily and a monthly timescale, over a 32-year period (1984–2016), assessing the performance of the gridded products by considering both the country as a whole and its individual regions. The ability of the datasets to correctly portray the occurrence of extreme events and precipitation patterns is examined by statistical metrics and further insights are provided by the application and statistical analysis of climate indices on ground observations. CHELSA, CERRAL and AgERA5 consistently yield acceptable results across statistical metrics, outperforming the other datasets, which exhibit inferior performance in both temporal scales. The statistical analysis reveals distinct patterns of heavier precipitation in northern and western regions, with strong seasonal variability in the West and South and a possible average decennial increase of over 110 mm in mean annual and over 30 mm in extreme precipitation, along the assessment period. Overall, the datasets fail to accurately depict precipitable extremes, but CHELSA and CERRAL stand out as more reliable options for describing the precipitation dynamics in Greece.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"123 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888665","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As a serious compound disaster, the drought-flood abrupt alternation (DFAA) has become consensus and is urgent to be analyzed. Previous studies have mainly focused on the monthly time scale, making it difficult to reveal the detailed variation characteristics of DFAA. To address this limitation, this study proposed a daily-scale method that combines narrow and generalized DFAA intensity to identify, screen, and evaluate DFAA phenomena. Taking the Pearl River Basin (PRB) as an example, the results obtained through the new method demonstrate good agreement with the observed reality. The DFAA events mainly occurred in the west, east and southeast of PRB, and their frequency, coverage, intensity and average duration exhibit a significant upward trend, while the maximum duration shows an insignificant downward trend. Generally, the proposed method offers a clear representation of the evolution of DFAA events, and enables comprehensive analysis and comparison of DFAA events within the same watershed, thereby promoting a deeper understanding of DFAA on a broader temporal scale.
{"title":"A novel daily-scale index for detecting drought-flood abrupt alternation events: Proof from Pearl River Basin, China","authors":"Chengguang Lai, Yuxing Wang, Yuxiang Zhao, Zhaoli Wang, Xushu Wu, Xiaoyan Bai","doi":"10.1016/j.atmosres.2024.107892","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107892","url":null,"abstract":"As a serious compound disaster, the drought-flood abrupt alternation (DFAA) has become consensus and is urgent to be analyzed. Previous studies have mainly focused on the monthly time scale, making it difficult to reveal the detailed variation characteristics of DFAA. To address this limitation, this study proposed a daily-scale method that combines narrow and generalized DFAA intensity to identify, screen, and evaluate DFAA phenomena. Taking the Pearl River Basin (PRB) as an example, the results obtained through the new method demonstrate good agreement with the observed reality. The DFAA events mainly occurred in the west, east and southeast of PRB, and their frequency, coverage, intensity and average duration exhibit a significant upward trend, while the maximum duration shows an insignificant downward trend. Generally, the proposed method offers a clear representation of the evolution of DFAA events, and enables comprehensive analysis and comparison of DFAA events within the same watershed, thereby promoting a deeper understanding of DFAA on a broader temporal scale.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"26 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142888664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.atmosres.2024.107893
Kazutoshi Sato, Kazu Takahashi, Jun Inoue
Cloud particle phase is an important controlling factor for the Earth's surface heat budget, through the radiative balance. Thus, it exerts a strong influence on climate change in the Arctic. Aerosols transported from lower latitudes modify Arctic cloud properties, including cloud phase. In this study, we investigated ice cloud formation and high aerosol concentrations over the Arctic Ocean using a combination of observations obtained by an Arctic voyage, reanalysis data, and backward trajectory analyses. On 12 September 2023, in an atmospheric river over the Arctic Ocean, ice clouds at temperatures warmer than −15 °C were observed in the middle troposphere by a Cloud Particle Sensor sonde. In the lower troposphere, a particle counter onboard a drone detected particle counts two orders of magnitude higher than the voyage average. Backward trajectories indicated that a lower tropospheric air mass with a high concentration of organic carbon (OC) aerosols over northern and coastal western Canada, where wildfire-induced OC emissions were evident, reached the mid-troposphere over the Arctic Ocean. These results suggest that OC aerosols from severe Canadian wildfires in the summer of 2023 acted as ice-nucleating particles for ice cloud formation under high-temperature conditions exceeding −15 °C over the Arctic Ocean.
{"title":"Impact of Canadian wildfires on aerosol and ice clouds in the early-autumn Arctic","authors":"Kazutoshi Sato, Kazu Takahashi, Jun Inoue","doi":"10.1016/j.atmosres.2024.107893","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107893","url":null,"abstract":"Cloud particle phase is an important controlling factor for the Earth's surface heat budget, through the radiative balance. Thus, it exerts a strong influence on climate change in the Arctic. Aerosols transported from lower latitudes modify Arctic cloud properties, including cloud phase. In this study, we investigated ice cloud formation and high aerosol concentrations over the Arctic Ocean using a combination of observations obtained by an Arctic voyage, reanalysis data, and backward trajectory analyses. On 12 September 2023, in an atmospheric river over the Arctic Ocean, ice clouds at temperatures warmer than −15 °C were observed in the middle troposphere by a Cloud Particle Sensor sonde. In the lower troposphere, a particle counter onboard a drone detected particle counts two orders of magnitude higher than the voyage average. Backward trajectories indicated that a lower tropospheric air mass with a high concentration of organic carbon (OC) aerosols over northern and coastal western Canada, where wildfire-induced OC emissions were evident, reached the mid-troposphere over the Arctic Ocean. These results suggest that OC aerosols from severe Canadian wildfires in the summer of 2023 acted as ice-nucleating particles for ice cloud formation under high-temperature conditions exceeding −15 °C over the Arctic Ocean.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"72 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142917987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.atmosres.2024.107887
Jiamu Ding, Renlong Hang, Rui Zhang, Luhui Yue, Qingshan Liu
Deep learning has attracted more and more attention in the field of tropical cyclone (TC) intensity estimation (TCIE). It is able to achieve promising results when the testing data follows the same distribution as the training data. However, due to the difference of geographical locations, TC intensity distributions, and imaging sensors, TC in different basins often show diverse distributions, making deep learning models trained on one basin can hardly be generalized to other basins. In this paper, we propose a cross-basin incremental learning model (CBIL-TCIE) to estimate the intensity of TC in multiple basins. CBIL-TCIE consists of domain-shared and domain-specific layers within the framework of multi-task learning. The domain-shared layers learn the common knowledge of all basins, and the domain-specific layers learn the specific knowledge of the current basin. Additionally, most of the existing studies have primarily focused on utilizing either maximum sustained wind (MSW) or minimum sea level pressure (MSLP) to represent TC intensity. Differently, in order to better characterize the intensity of TCs, our model can output MSW and MSLP concurrently as the TC intensity in different basins. To test the performance of our proposed model, we conduct experiments on a widely used dataset named GridSat, which consists of TC data across multiple basins. The performance of the CBIL-TCIE in multiple basins can improve by 19.2 % compared to the widely used fine-tuning method. Furthermore, the experiment demonstrates that concurrently outputting MSW and MSLP can effectively facilitate the ability of TC intensity estimation.
{"title":"Cross-basin incremental learning for tropical cyclone intensity estimation","authors":"Jiamu Ding, Renlong Hang, Rui Zhang, Luhui Yue, Qingshan Liu","doi":"10.1016/j.atmosres.2024.107887","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107887","url":null,"abstract":"Deep learning has attracted more and more attention in the field of tropical cyclone (TC) intensity estimation (TCIE). It is able to achieve promising results when the testing data follows the same distribution as the training data. However, due to the difference of geographical locations, TC intensity distributions, and imaging sensors, TC in different basins often show diverse distributions, making deep learning models trained on one basin can hardly be generalized to other basins. In this paper, we propose a cross-basin incremental learning model (CBIL-TCIE) to estimate the intensity of TC in multiple basins. CBIL-TCIE consists of domain-shared and domain-specific layers within the framework of multi-task learning. The domain-shared layers learn the common knowledge of all basins, and the domain-specific layers learn the specific knowledge of the current basin. Additionally, most of the existing studies have primarily focused on utilizing either maximum sustained wind (MSW) or minimum sea level pressure (MSLP) to represent TC intensity. Differently, in order to better characterize the intensity of TCs, our model can output MSW and MSLP concurrently as the TC intensity in different basins. To test the performance of our proposed model, we conduct experiments on a widely used dataset named GridSat, which consists of TC data across multiple basins. The performance of the CBIL-TCIE in multiple basins can improve by 19.2 % compared to the widely used fine-tuning method. Furthermore, the experiment demonstrates that concurrently outputting MSW and MSLP can effectively facilitate the ability of TC intensity estimation.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"83 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142917988","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.atmosres.2024.107891
Zongbao Bai, Zhong Zhong, Yuan Sun, Yunying Li
In this study, the thermodynamic feedback effect of excess cloud ice possibly originating from tropical cyclones (TCs) on the western Pacific subtropical high (WPSH) and, in turn, the TC track was investigated by use of sensitivity experiments with the Weather Research and Forecast (WRF) model. The results indicated that excess cloud ice in the area north of the TC track can alter the temperature and the geopotential height there via latent heat release (cooling due to melting and evaporation) of hydrometeors in the upper (lower) troposphere, accompanying by the enhanced local downdrafts. Moreover, under the influence of the changed steering flow, the TC turns northward ahead of time relative to its original track. The sensitivity experiments confirm that the thermodynamic feedback effect of hydrometeors aloft can ultimately affect the TC track, which is also supported by observations of some selected northward-turning TCs.
{"title":"Effect of the thermodynamic feedback of tropical cyclones on the subtropical high","authors":"Zongbao Bai, Zhong Zhong, Yuan Sun, Yunying Li","doi":"10.1016/j.atmosres.2024.107891","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107891","url":null,"abstract":"In this study, the thermodynamic feedback effect of excess cloud ice possibly originating from tropical cyclones (TCs) on the western Pacific subtropical high (WPSH) and, in turn, the TC track was investigated by use of sensitivity experiments with the Weather Research and Forecast (WRF) model. The results indicated that excess cloud ice in the area north of the TC track can alter the temperature and the geopotential height there via latent heat release (cooling due to melting and evaporation) of hydrometeors in the upper (lower) troposphere, accompanying by the enhanced local downdrafts. Moreover, under the influence of the changed steering flow, the TC turns northward ahead of time relative to its original track. The sensitivity experiments confirm that the thermodynamic feedback effect of hydrometeors aloft can ultimately affect the TC track, which is also supported by observations of some selected northward-turning TCs.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"82 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142917993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.atmosres.2024.107882
Jian Rao, Xiaoqi Zhang, Qian Lu, Siming Liu
The stratospheric disturbances and their impact on predictability of near surface extreme events are one of crucial issues in the subseasonal to seasonal (S2S) prediction project. This study examines the 2023/24 winter when frequent stratospheric disturbances occurred, including minor and major sudden stratospheric warmings (SSWs). The stratospheric circulation was disturbed multiple times, with rapid circumpolar westerly wind deceleration and even zonal wind reversal. Corresponding wave pulses were observed in the troposphere and lower stratosphere, with large eddy heat flux pulses appearing before every stratospheric perturbation. The stratospheric perturbation was examined from the spatiotemporal evolution of the annular mode index, revealing two instances of evident downward propagation. Nevertheless, the near surface did not respond in a typical negative NAM pattern associated with the stratospheric signal. The study further analyzed the predictability of the near surface and its relation with the stratospheric disturbance using common initializations in January and February 2024 from S2S models. The results indicate that the near surface predictability was not enhanced in the 2023/24 winter albeit with frequent stratospheric disturbances, and the contribution of the stratospheric disturbance to the surface predictability was limited. Although the multimodel ensemble means forecast warm spots over broad regions of lands and dry spots in part of China and US, the stratospheric circulation error nearly did not explain the near surface forecasting error among S2S models most of the time. The subseasonal predictability of the near surface conditions over the course of the 2023/24 winter seldom originated from the stratospheric disturbances, and other predictability sources such as the warm tropical Pacific Ocean and increased Arctic sea ice should be considered.
{"title":"Prediction of near-surface conditions following the 2023/24 sudden stratospheric warming by the S2S project models","authors":"Jian Rao, Xiaoqi Zhang, Qian Lu, Siming Liu","doi":"10.1016/j.atmosres.2024.107882","DOIUrl":"https://doi.org/10.1016/j.atmosres.2024.107882","url":null,"abstract":"The stratospheric disturbances and their impact on predictability of near surface extreme events are one of crucial issues in the subseasonal to seasonal (S2S) prediction project. This study examines the 2023/24 winter when frequent stratospheric disturbances occurred, including minor and major sudden stratospheric warmings (SSWs). The stratospheric circulation was disturbed multiple times, with rapid circumpolar westerly wind deceleration and even zonal wind reversal. Corresponding wave pulses were observed in the troposphere and lower stratosphere, with large eddy heat flux pulses appearing before every stratospheric perturbation. The stratospheric perturbation was examined from the spatiotemporal evolution of the annular mode index, revealing two instances of evident downward propagation. Nevertheless, the near surface did not respond in a typical negative NAM pattern associated with the stratospheric signal. The study further analyzed the predictability of the near surface and its relation with the stratospheric disturbance using common initializations in January and February 2024 from S2S models. The results indicate that the near surface predictability was not enhanced in the 2023/24 winter albeit with frequent stratospheric disturbances, and the contribution of the stratospheric disturbance to the surface predictability was limited. Although the multimodel ensemble means forecast warm spots over broad regions of lands and dry spots in part of China and US, the stratospheric circulation error nearly did not explain the near surface forecasting error among S2S models most of the time. The subseasonal predictability of the near surface conditions over the course of the 2023/24 winter seldom originated from the stratospheric disturbances, and other predictability sources such as the warm tropical Pacific Ocean and increased Arctic sea ice should be considered.","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"19 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142917991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：