This paper evaluates Indo-Pacific warm pool (IPWP) sea surface temperature (SST) warming biases of Coupled Model Intercomparison Project Phase 5 (CMIP5) and CMIP6. The IPWP warming trend in the CMIP5 multi-model ensemble (MME) is closer to observation than in CMIP6 MME, but the IPWP expanding trend is the opposite. There is no qualitative improvement in the simulation of IPWP warming from CMIP5 to CMIP6. In addition, four metrics were used to investigate the performance of Indo-Pacific region warming trends in all models. CMIP6 models perform better than CMIP5 with smaller root mean square error and bias in MME and higher skill scores in MME and top models, which is tightly linked to their better performance in simulating associated physical processes in CMIP6 models. IPWP warming biases are mainly attributed to the combined effects of positive atmospheric process biases and negative ocean dynamics term biases. The positive atmospheric process biases are primarily related to the shortwave radiation and latent heat flux from atmospheric forcing, the latter of which can be attributed to the biases in surface wind fields. Compared with CMIP5 models, the IPWP warming simulated by CMIP6 models is weaker, related to the less robust atmospheric processes and the shallower thermocline anomalies simulated by CMIP6.
{"title":"The simulation of the Indo-Pacific warm pool SST warming trend in CMIP5 and CMIP6","authors":"Wenrong Bai, Hailong Liu, Pengfei Lin, Hongyan Shen","doi":"10.1186/s40562-024-00346-6","DOIUrl":"https://doi.org/10.1186/s40562-024-00346-6","url":null,"abstract":"This paper evaluates Indo-Pacific warm pool (IPWP) sea surface temperature (SST) warming biases of Coupled Model Intercomparison Project Phase 5 (CMIP5) and CMIP6. The IPWP warming trend in the CMIP5 multi-model ensemble (MME) is closer to observation than in CMIP6 MME, but the IPWP expanding trend is the opposite. There is no qualitative improvement in the simulation of IPWP warming from CMIP5 to CMIP6. In addition, four metrics were used to investigate the performance of Indo-Pacific region warming trends in all models. CMIP6 models perform better than CMIP5 with smaller root mean square error and bias in MME and higher skill scores in MME and top models, which is tightly linked to their better performance in simulating associated physical processes in CMIP6 models. IPWP warming biases are mainly attributed to the combined effects of positive atmospheric process biases and negative ocean dynamics term biases. The positive atmospheric process biases are primarily related to the shortwave radiation and latent heat flux from atmospheric forcing, the latter of which can be attributed to the biases in surface wind fields. Compared with CMIP5 models, the IPWP warming simulated by CMIP6 models is weaker, related to the less robust atmospheric processes and the shallower thermocline anomalies simulated by CMIP6.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"39 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141502476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-06-21DOI: 10.1186/s40562-024-00345-7
Zhongqiang Sun, Shuangyue Lin, Guangqun Wang, Longlong Liu, Mengqi Wang
Identifying and characterizing sedimentary evolution patterns are crucial for assessing the distributions of source and reservoir rocks, which are fundamental to hydrocarbon exploration. This study analyzed the stratigraphic sequence, lithological characteristics, sedimentary lithofacies, individual well sedimentary sequences, and seismic reflection properties. The analysis revealed six fourth-order sequences, including progradational and regressive sequences, indicative of water level changes. The sediment sources for the second and third sub-members of the Eocene Shahejie Formation's third member (Es32+3) in the Nanpu Sag were identified as the Baigezhuang and Xinanzhuang Uplifts. Predominantly, the sandstones are lithic arkose and feldspathic litharenite, both of which exhibit low compositional and structural maturity. Notably, 22 lithofacies and 8 lithofacies associations suggest fan delta processes. This study identified three fundamental seismic reflection package reflection types. These lithofacies associations, sedimentary sequences, and seismic reflections serve as critical indicators for determining sedimentary environments. The results from the sedimentary facies analysis indicate that the Es32+3 Formation developed fan delta deposits, controlled by the sequence of the sedimentary evolution pattern. The potential of these fan delta sediments to form oil and gas reservoirs is significant. Therefore, precise characterization of the sedimentary evolution pattern is essential for a comprehensive understanding of basin dynamics and hydrocarbon potential.
{"title":"Sedimentary evolution pattern influenced by sequence stratigraphy: a case study of the Nanpu Sag, Bohai Bay Basin, China","authors":"Zhongqiang Sun, Shuangyue Lin, Guangqun Wang, Longlong Liu, Mengqi Wang","doi":"10.1186/s40562-024-00345-7","DOIUrl":"https://doi.org/10.1186/s40562-024-00345-7","url":null,"abstract":"Identifying and characterizing sedimentary evolution patterns are crucial for assessing the distributions of source and reservoir rocks, which are fundamental to hydrocarbon exploration. This study analyzed the stratigraphic sequence, lithological characteristics, sedimentary lithofacies, individual well sedimentary sequences, and seismic reflection properties. The analysis revealed six fourth-order sequences, including progradational and regressive sequences, indicative of water level changes. The sediment sources for the second and third sub-members of the Eocene Shahejie Formation's third member (Es32+3) in the Nanpu Sag were identified as the Baigezhuang and Xinanzhuang Uplifts. Predominantly, the sandstones are lithic arkose and feldspathic litharenite, both of which exhibit low compositional and structural maturity. Notably, 22 lithofacies and 8 lithofacies associations suggest fan delta processes. This study identified three fundamental seismic reflection package reflection types. These lithofacies associations, sedimentary sequences, and seismic reflections serve as critical indicators for determining sedimentary environments. The results from the sedimentary facies analysis indicate that the Es32+3 Formation developed fan delta deposits, controlled by the sequence of the sedimentary evolution pattern. The potential of these fan delta sediments to form oil and gas reservoirs is significant. Therefore, precise characterization of the sedimentary evolution pattern is essential for a comprehensive understanding of basin dynamics and hydrocarbon potential.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"19 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141502477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study explored closed-circuit television (CCTV) networks in northeastern Toyama Prefecture, Japan, as a new data source for tsunami detection following the 2024 Noto Peninsula earthquake. We analyzed CCTV footage and extracted time-series water level fluctuations at Yokoyama, Shimoiino, and Ekko. Spectral analysis of these waveforms revealed several long-period peaks (more than 100 s) in power spectral density (PSD), suggesting the presence of tsunami components. Notably, relatively large PSD peaks at approximately 5–10 min were observed at all CCTV locations in this study and at offshore wave observation points (Tanaka and Toyama). At Yokoyama, a maximum run-up of approximately 3 m was confirmed around 16:28. Although water level fluctuations at Shimoiino and Ekko were detected, identifying tsunami components proved challenging due to their small magnitude compared to other wave components. Despite these challenges, this study demonstrates the potential of CCTV networks for tsunami detection, and further research is needed to achieve real-time detection.
{"title":"Potential for tsunami detection via CCTV cameras in northeastern Toyama Prefecture, Japan, following the 2024 Noto Peninsula earthquake","authors":"Tomoki Shirai, Yota Enomoto, Keisuke Haga, Tatsuhiko Tokuta, Taro Arikawa, Nobuhito Mori, Fumihiko Imamura","doi":"10.1186/s40562-024-00343-9","DOIUrl":"https://doi.org/10.1186/s40562-024-00343-9","url":null,"abstract":"This study explored closed-circuit television (CCTV) networks in northeastern Toyama Prefecture, Japan, as a new data source for tsunami detection following the 2024 Noto Peninsula earthquake. We analyzed CCTV footage and extracted time-series water level fluctuations at Yokoyama, Shimoiino, and Ekko. Spectral analysis of these waveforms revealed several long-period peaks (more than 100 s) in power spectral density (PSD), suggesting the presence of tsunami components. Notably, relatively large PSD peaks at approximately 5–10 min were observed at all CCTV locations in this study and at offshore wave observation points (Tanaka and Toyama). At Yokoyama, a maximum run-up of approximately 3 m was confirmed around 16:28. Although water level fluctuations at Shimoiino and Ekko were detected, identifying tsunami components proved challenging due to their small magnitude compared to other wave components. Despite these challenges, this study demonstrates the potential of CCTV networks for tsunami detection, and further research is needed to achieve real-time detection.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"38 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141259603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-05-27DOI: 10.1186/s40562-024-00341-x
Ke-Sheng Cheng, Gwo‑Hsing Yu, Yuan-Li Tai, Kuo-Chan Huang, Sheng‑Fu Tsai, Dong‑Hong Wu, Yun-Ching Lin, Ching-Teng Lee, Tzu-Ting Lo
A hypothesis testing approach, based on the theorem of probability integral transformation and the Kolmogorov–Smirnov one-sample test, for performance evaluation of probabilistic seasonal rainfall forecasts is proposed in this study. By considering the probability distribution of monthly rainfalls, the approach transforms the tercile forecast probabilities into a forecast distribution and tests whether the observed data truly come from the forecast distribution. The proposed approach provides not only a quantitative measure for performance evaluation but also a cumulative probability plot for insightful interpretations of forecast characteristics such as overconfident, underconfident, mean-overestimated, and mean-underestimated. The approach has been applied for the performance evaluation of probabilistic season rainfall forecasts in northern Taiwan, and it was found that the forecast performance is seasonal dependent. Probabilistic seasonal rainfall forecasts of the Meiyu season are likely to be overconfident and mean-underestimated, while forecasts of the winter-to-spring season are overconfident. A relatively good forecast performance is observed for the summer season.
{"title":"Hypothesis testing for performance evaluation of probabilistic seasonal rainfall forecasts","authors":"Ke-Sheng Cheng, Gwo‑Hsing Yu, Yuan-Li Tai, Kuo-Chan Huang, Sheng‑Fu Tsai, Dong‑Hong Wu, Yun-Ching Lin, Ching-Teng Lee, Tzu-Ting Lo","doi":"10.1186/s40562-024-00341-x","DOIUrl":"https://doi.org/10.1186/s40562-024-00341-x","url":null,"abstract":"A hypothesis testing approach, based on the theorem of probability integral transformation and the Kolmogorov–Smirnov one-sample test, for performance evaluation of probabilistic seasonal rainfall forecasts is proposed in this study. By considering the probability distribution of monthly rainfalls, the approach transforms the tercile forecast probabilities into a forecast distribution and tests whether the observed data truly come from the forecast distribution. The proposed approach provides not only a quantitative measure for performance evaluation but also a cumulative probability plot for insightful interpretations of forecast characteristics such as overconfident, underconfident, mean-overestimated, and mean-underestimated. The approach has been applied for the performance evaluation of probabilistic season rainfall forecasts in northern Taiwan, and it was found that the forecast performance is seasonal dependent. Probabilistic seasonal rainfall forecasts of the Meiyu season are likely to be overconfident and mean-underestimated, while forecasts of the winter-to-spring season are overconfident. A relatively good forecast performance is observed for the summer season.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"57 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141172275","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In response to the growing demand for high-resolution rainfall data to support disaster prevention in Taiwan, this study presents an innovative approach for downscaling precipitation data. We employed a hierarchical architecture of Multi-Scale Residual Networks (MSRN) to downscale rainfall from a coarse 0.25-degree resolution to a fine 0.0125-degree resolution, representing a substantial challenge due to a resolution increase of over 20 times. Our results demonstrate that the hierarchical MSRN outperforms both the one-step MSRN and linear interpolation methods when reconstructing high-resolution daily rainfall. It surpasses the linear interpolation method by 15.1 and 9.1% in terms of mean absolute error and root mean square error, respectively. Furthermore, the hierarchical MSRN excels in accurately reproducing high-resolution rainfall for various rainfall thresholds, displaying minimal biases. The threat score (TS) highlights the hierarchical MSRN's capability to replicate extreme rainfall events, achieving TS scores exceeding 0.54 and 0.46 at rainfall thresholds of 350 and 500 mm per day, outperforming alternative methods. This method is also applied to an operational global model, the ECMWF’s daily rainfall forecasts over Taiwan. The evaluation results indicate that our approach is effective at improving rainfall forecasts for thresholds greater than 100 mm per day, with more significant improvement for the 1- to 3-day lead forecast. This approach also offers a realistic visual representation of fine-grained rainfall distribution, showing promise for making significant contributions to disaster preparedness and weather forecasting in Taiwan.
{"title":"Downscaling Taiwan precipitation with a residual deep learning approach","authors":"Li-Huan Hsu, Chou-Chun Chiang, Kuan-Ling Lin, Hsin-Hung Lin, Jung-Lien Chu, Yi-Chiang Yu, Chin-Shyurng Fahn","doi":"10.1186/s40562-024-00340-y","DOIUrl":"https://doi.org/10.1186/s40562-024-00340-y","url":null,"abstract":"In response to the growing demand for high-resolution rainfall data to support disaster prevention in Taiwan, this study presents an innovative approach for downscaling precipitation data. We employed a hierarchical architecture of Multi-Scale Residual Networks (MSRN) to downscale rainfall from a coarse 0.25-degree resolution to a fine 0.0125-degree resolution, representing a substantial challenge due to a resolution increase of over 20 times. Our results demonstrate that the hierarchical MSRN outperforms both the one-step MSRN and linear interpolation methods when reconstructing high-resolution daily rainfall. It surpasses the linear interpolation method by 15.1 and 9.1% in terms of mean absolute error and root mean square error, respectively. Furthermore, the hierarchical MSRN excels in accurately reproducing high-resolution rainfall for various rainfall thresholds, displaying minimal biases. The threat score (TS) highlights the hierarchical MSRN's capability to replicate extreme rainfall events, achieving TS scores exceeding 0.54 and 0.46 at rainfall thresholds of 350 and 500 mm per day, outperforming alternative methods. This method is also applied to an operational global model, the ECMWF’s daily rainfall forecasts over Taiwan. The evaluation results indicate that our approach is effective at improving rainfall forecasts for thresholds greater than 100 mm per day, with more significant improvement for the 1- to 3-day lead forecast. This approach also offers a realistic visual representation of fine-grained rainfall distribution, showing promise for making significant contributions to disaster preparedness and weather forecasting in Taiwan.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"160 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140935867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-04-29DOI: 10.1186/s40562-024-00337-7
V. Manu, N. Balan, Y. Ebihara, Qing-He Zhang, Zan-Yang Xing
We notice that the important early decreasing part of the main phase (MP) from the positive main phase onset (MPO) to 0-level of Dst and SymH indices is missed in the treatment of the main phase (MP) of geomagnetic storms. We correct this inconsistency in 848 storms having positive MPO (out of 1164 storms) in SymH during 1981–2019 by raising the 0-level of SymH to the MPO-level. The correction considers the full range of the main phase, increases the corrected (revised) storm intensity (SymHMin*) and impulsive strength (IpsSymH*) by up to − 149 nT and − 134 nT, respectively, and seems important for all aspects of global space weather. For example, the corrected SymHMin* changes the conventional storm identification and classification and corrected IpsSymH* clearly identifies all 3 severe space weather (SvSW) events from over 1100 normal space weather (NSW) events with a separation of 52 nT; it also identifies all 8 minor-system-damage space weather (MSW) events from the NSW events. Large fluctuations occur in the global geomagnetic field during space weather events. The fluctuations at low latitudes are referred as geomagnetic storms. The Dst and SymH indices have been used for studying the storms and other aspects of global space weather. However, we notice that the Dst and SymH values during the main phase and recovery phase of the storms having positive main phase onset (MPO > 0 nT) are significantly less than their actual values. We correct this inconsistency in 848 such storms (out of 1164 storms) in SymH during 1981-2019 by raising the 0-level of SymH to the MPO-level. The corrected/revised storm intensity (SymHMin*) and impulsive strength (IpsSymH*) increase by up to − 149 and − 134 nT. The correction seems important for studying all aspects global space weather. For example, the correction identifies the storms corresponding to severe space weather causing power outage and/or telecommunication failure from those corresponding to normal space weather.
{"title":"A fresh look at the intensity and impulsive strength of geomagnetic storms","authors":"V. Manu, N. Balan, Y. Ebihara, Qing-He Zhang, Zan-Yang Xing","doi":"10.1186/s40562-024-00337-7","DOIUrl":"https://doi.org/10.1186/s40562-024-00337-7","url":null,"abstract":"We notice that the important early decreasing part of the main phase (MP) from the positive main phase onset (MPO) to 0-level of Dst and SymH indices is missed in the treatment of the main phase (MP) of geomagnetic storms. We correct this inconsistency in 848 storms having positive MPO (out of 1164 storms) in SymH during 1981–2019 by raising the 0-level of SymH to the MPO-level. The correction considers the full range of the main phase, increases the corrected (revised) storm intensity (SymHMin*) and impulsive strength (IpsSymH*) by up to − 149 nT and − 134 nT, respectively, and seems important for all aspects of global space weather. For example, the corrected SymHMin* changes the conventional storm identification and classification and corrected IpsSymH* clearly identifies all 3 severe space weather (SvSW) events from over 1100 normal space weather (NSW) events with a separation of 52 nT; it also identifies all 8 minor-system-damage space weather (MSW) events from the NSW events. Large fluctuations occur in the global geomagnetic field during space weather events. The fluctuations at low latitudes are referred as geomagnetic storms. The Dst and SymH indices have been used for studying the storms and other aspects of global space weather. However, we notice that the Dst and SymH values during the main phase and recovery phase of the storms having positive main phase onset (MPO > 0 nT) are significantly less than their actual values. We correct this inconsistency in 848 such storms (out of 1164 storms) in SymH during 1981-2019 by raising the 0-level of SymH to the MPO-level. The corrected/revised storm intensity (SymHMin*) and impulsive strength (IpsSymH*) increase by up to − 149 and − 134 nT. The correction seems important for studying all aspects global space weather. For example, the correction identifies the storms corresponding to severe space weather causing power outage and/or telecommunication failure from those corresponding to normal space weather. ","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"43 6 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140834641","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-04-18DOI: 10.1186/s40562-024-00334-w
Rong Cui, Xuhua Cheng, Wei Duan, Long Jiang, Yifei Zhou
In response to abundant freshwater input from rainfall and river discharge, the northern Bay of Bengal (BoB) is featured by low sea surface salinity (SSS) and strong intraseasonal variability (ISV). This study investigates the characteristic and dynamic mechanisms of SSS ISV in the northern BoB based on satellite observations and the output of Simple Ocean Data Assimilation (SODA). The strong SSS ISV is mainly concentrated near the mouth of the Ganges–Brahmaputra River and along the east coast of India, where the horizontal salinity gradient varies greatly. SSS ISV in the northern BoB is notably in phase with freshwater transport, which peaks from July to November. The contribution of riverine freshwater is significant both geographically and temporally. The SSS budget analysis indicates that the horizontal advection plays a dominant role in SSS ISV. Once currents cross the salinity field, large horizontal advection anomalies become important and favor SSS ISV. Altered SSS patterns can impact water density, potentially influencing the strength and direction of currents. This, in turn, may have cascading effects on local and regional climate patterns.
{"title":"Features and mechanisms of sea surface salinity intraseasonal variability in the Northern Bay of Bengal","authors":"Rong Cui, Xuhua Cheng, Wei Duan, Long Jiang, Yifei Zhou","doi":"10.1186/s40562-024-00334-w","DOIUrl":"https://doi.org/10.1186/s40562-024-00334-w","url":null,"abstract":"In response to abundant freshwater input from rainfall and river discharge, the northern Bay of Bengal (BoB) is featured by low sea surface salinity (SSS) and strong intraseasonal variability (ISV). This study investigates the characteristic and dynamic mechanisms of SSS ISV in the northern BoB based on satellite observations and the output of Simple Ocean Data Assimilation (SODA). The strong SSS ISV is mainly concentrated near the mouth of the Ganges–Brahmaputra River and along the east coast of India, where the horizontal salinity gradient varies greatly. SSS ISV in the northern BoB is notably in phase with freshwater transport, which peaks from July to November. The contribution of riverine freshwater is significant both geographically and temporally. The SSS budget analysis indicates that the horizontal advection plays a dominant role in SSS ISV. Once currents cross the salinity field, large horizontal advection anomalies become important and favor SSS ISV. Altered SSS patterns can impact water density, potentially influencing the strength and direction of currents. This, in turn, may have cascading effects on local and regional climate patterns.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"38 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140625794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-04-13DOI: 10.1186/s40562-024-00336-8
Soonyoung Roh, Park Sa Kim, Hwan-Jin Song
This study aimed to identify the optimal configuration for neural network (NN) emulators in numerical weather prediction, minimizing trial and error by comparing emulator performance across multiple hidden layers (1–5 layers), as automatically defined by the Sherpa library. Our findings revealed that Sherpa-applied emulators consistently demonstrated good results and stable performance with low errors in numerical simulations. The optimal configurations were observed with one and two hidden layers, improving results when two hidden layers were employed. The Sherpa-defined average neurons per hidden layer ranged between 153 and 440, resulting in a speedup relative to the CNT of 7–12 times. These results provide valuable insights for developing radiative physical NN emulators. Utilizing automatically determined hyperparameters can effectively reduce trial-and-error processes while maintaining stable outcomes. However, further experimentation is needed to establish the most suitable hyperparameter values that balance both speed and accuracy, as this study did not identify optimized values for all hyperparameters.
{"title":"Streamlining hyperparameter optimization for radiation emulator training with automated Sherpa","authors":"Soonyoung Roh, Park Sa Kim, Hwan-Jin Song","doi":"10.1186/s40562-024-00336-8","DOIUrl":"https://doi.org/10.1186/s40562-024-00336-8","url":null,"abstract":"This study aimed to identify the optimal configuration for neural network (NN) emulators in numerical weather prediction, minimizing trial and error by comparing emulator performance across multiple hidden layers (1–5 layers), as automatically defined by the Sherpa library. Our findings revealed that Sherpa-applied emulators consistently demonstrated good results and stable performance with low errors in numerical simulations. The optimal configurations were observed with one and two hidden layers, improving results when two hidden layers were employed. The Sherpa-defined average neurons per hidden layer ranged between 153 and 440, resulting in a speedup relative to the CNT of 7–12 times. These results provide valuable insights for developing radiative physical NN emulators. Utilizing automatically determined hyperparameters can effectively reduce trial-and-error processes while maintaining stable outcomes. However, further experimentation is needed to establish the most suitable hyperparameter values that balance both speed and accuracy, as this study did not identify optimized values for all hyperparameters. ","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"61 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140559968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-03-25DOI: 10.1186/s40562-024-00330-0
Satrio Muhammad Alif, Kuo-En Ching, Takeshi Sagiya, Widya Nabila Wahyuni
To provide a precise Euler pole parameter of Sundaland plate for earthquake potential evaluation in Sumatra, Indonesia after the 2004 M9.2 Aceh earthquake, we adopted 37 new Global Navigation Satellite System (GNSS) observations in Sumatra and 30 transformed published velocities in Indochina and Malaysia under the International Terrestrial Reference Frame 2014 (ITRF2014). The 37 GNSS data were processed using the software Bernese v.5.2. The GNSS velocities were calculated by the coordinate time series analysis with the least squares method. The grid search algorithm was used in Euler pole parameter estimation, which was validated using the bootstrap resampling. The optimized Euler pole parameters are the latitude of 45.63 ± 0.45°, the longitude of − 88.71 ± 0.38° and the angular velocity of 0.337 ± 0.002°/Myr in counterclockwise direction. Besides, the distinguishable and systematic pattern in space is shown in the residual velocities, which may imply the possibility of minor postseismic deformation, Tibetan crustal flows, or the hypothesis that the Sundaland Plate is composed of several microplates.
{"title":"Determination of Euler pole parameters for Sundaland plate based on updated GNSS observations in Sumatra, Indonesia","authors":"Satrio Muhammad Alif, Kuo-En Ching, Takeshi Sagiya, Widya Nabila Wahyuni","doi":"10.1186/s40562-024-00330-0","DOIUrl":"https://doi.org/10.1186/s40562-024-00330-0","url":null,"abstract":"To provide a precise Euler pole parameter of Sundaland plate for earthquake potential evaluation in Sumatra, Indonesia after the 2004 M9.2 Aceh earthquake, we adopted 37 new Global Navigation Satellite System (GNSS) observations in Sumatra and 30 transformed published velocities in Indochina and Malaysia under the International Terrestrial Reference Frame 2014 (ITRF2014). The 37 GNSS data were processed using the software Bernese v.5.2. The GNSS velocities were calculated by the coordinate time series analysis with the least squares method. The grid search algorithm was used in Euler pole parameter estimation, which was validated using the bootstrap resampling. The optimized Euler pole parameters are the latitude of 45.63 ± 0.45°, the longitude of − 88.71 ± 0.38° and the angular velocity of 0.337 ± 0.002°/Myr in counterclockwise direction. Besides, the distinguishable and systematic pattern in space is shown in the residual velocities, which may imply the possibility of minor postseismic deformation, Tibetan crustal flows, or the hypothesis that the Sundaland Plate is composed of several microplates.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"273 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140298757","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2024-03-22DOI: 10.1186/s40562-024-00323-z
Ryan Paulik, Alec Wild, Conrad Zorn, Liam Wotherspoon, Shaun Williams
Reliable flood damage models are informed by detailed damage assessments. Damage models are critical in flood risk assessments, representing an elements vulnerability to damage. This study evaluated residential building damage for the July 2021 flood in Westport, New Zealand. We report on flood hazard, exposure and damage features observed for 247 residential buildings. Damage samples were applied to evaluate univariable and multivariable model performance using different variable sample sizes and regression-based supervised learning algorithms. Feature analysis for damage prediction showed high importance of water depth variables and low importance for commonly observed building variables such as structural frame and storeys. Overfitting occurred for most models evaluated when more than 150 samples were used. This resulted from limited damage heterogeneity observed, and variables of low importance affecting model learning. The Random Forest algorithm, which considered multiple important variables (water depth above floor level, area and floor height) improved predictive precision by 17% relative to other models when over 150 damage samples were considered. Our findings suggest the evaluated model performance could be improved by incorporating heterogeneous damage samples from similar flood contexts, in turn increasing capacity for reliable spatial transfer.
{"title":"Evaluation of residential building damage for the July 2021 flood in Westport, New Zealand","authors":"Ryan Paulik, Alec Wild, Conrad Zorn, Liam Wotherspoon, Shaun Williams","doi":"10.1186/s40562-024-00323-z","DOIUrl":"https://doi.org/10.1186/s40562-024-00323-z","url":null,"abstract":"Reliable flood damage models are informed by detailed damage assessments. Damage models are critical in flood risk assessments, representing an elements vulnerability to damage. This study evaluated residential building damage for the July 2021 flood in Westport, New Zealand. We report on flood hazard, exposure and damage features observed for 247 residential buildings. Damage samples were applied to evaluate univariable and multivariable model performance using different variable sample sizes and regression-based supervised learning algorithms. Feature analysis for damage prediction showed high importance of water depth variables and low importance for commonly observed building variables such as structural frame and storeys. Overfitting occurred for most models evaluated when more than 150 samples were used. This resulted from limited damage heterogeneity observed, and variables of low importance affecting model learning. The Random Forest algorithm, which considered multiple important variables (water depth above floor level, area and floor height) improved predictive precision by 17% relative to other models when over 150 damage samples were considered. Our findings suggest the evaluated model performance could be improved by incorporating heterogeneous damage samples from similar flood contexts, in turn increasing capacity for reliable spatial transfer.","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"29 1","pages":""},"PeriodicalIF":4.0,"publicationDate":"2024-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140197584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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：