Search Results
You are looking at 1 - 10 of 20 items for :
- Author or Editor: Wei Li x
- Journal of Atmospheric and Oceanic Technology x
- Refine by Access: All Content x
Atmospheric Cold Front–Induced Exchange Flows through a Microtidal Multi-Inlet Bay: Analysis Using Multiple Horizontal ADCPs and FVCOM SimulationsAbstract
Atmospheric cold fronts provide recurring forcing for circulations and long-term transport in estuaries with microtides. Multiple horizontal ADCPs were used to obtain time series data from three inlets in Barataria Bay. The data cover a period of 51 atmospheric cold fronts between 2013 and 2015. The weather and subtidal ocean response are highly correlated in the “weather band” (3–7 days). The cold front–associated winds produce alternating flows into, out of, and then back into the bay, forming an asymmetric “M” for low-pass filtered flows. Results show that cold front–induced flows are the most important component in this region, and the flows can be predicted based on wind vector time series. Numerical simulations using a validated Finite-Volume Coastal Ocean Model (FVCOM) demonstrate that the wind-driven oscillations within the bay are consistent with the quasi-steady state with little influence of the Coriolis effect for cold front–related wind-driven flows. The four major inlets (from the southwest to the northeast) consistently carry 10%, 57%, 21%, and 12% of the tidal exchange of the bay, respectively. The subtidal exchange rates through them however fluctuate greatly with averages of 18% ± 13%, 35% ± 18%, 31% ± 16%, and 16% ± 9%, respectively. Several modes of exchange flows through the multiple inlets are found, consisting of the all-in and all-out mode (45% occurrence) under strong winds perpendicular to the coastline; the shallow-downwind, deep-upwind mode (41%), particularly during wind-relaxation periods; and the upwind-in and downwind-out mode (13%) under northerly or southerly winds. These modes are discussed with the low-pass filtered model results and verified by a forcing–response joint EOF analysis.
Abstract
Atmospheric cold fronts provide recurring forcing for circulations and long-term transport in estuaries with microtides. Multiple horizontal ADCPs were used to obtain time series data from three inlets in Barataria Bay. The data cover a period of 51 atmospheric cold fronts between 2013 and 2015. The weather and subtidal ocean response are highly correlated in the “weather band” (3–7 days). The cold front–associated winds produce alternating flows into, out of, and then back into the bay, forming an asymmetric “M” for low-pass filtered flows. Results show that cold front–induced flows are the most important component in this region, and the flows can be predicted based on wind vector time series. Numerical simulations using a validated Finite-Volume Coastal Ocean Model (FVCOM) demonstrate that the wind-driven oscillations within the bay are consistent with the quasi-steady state with little influence of the Coriolis effect for cold front–related wind-driven flows. The four major inlets (from the southwest to the northeast) consistently carry 10%, 57%, 21%, and 12% of the tidal exchange of the bay, respectively. The subtidal exchange rates through them however fluctuate greatly with averages of 18% ± 13%, 35% ± 18%, 31% ± 16%, and 16% ± 9%, respectively. Several modes of exchange flows through the multiple inlets are found, consisting of the all-in and all-out mode (45% occurrence) under strong winds perpendicular to the coastline; the shallow-downwind, deep-upwind mode (41%), particularly during wind-relaxation periods; and the upwind-in and downwind-out mode (13%) under northerly or southerly winds. These modes are discussed with the low-pass filtered model results and verified by a forcing–response joint EOF analysis.
A Three-Step Method for Estimating Vortex Center Locations in Four-Dimensional Space from Radar-Observed Tornadic MesocyclonesAbstract
The recently developed two-dimensional variational methods for analyzing vortex winds from radar-observed mesocyclones can be extended to analyze three-dimensional vortex winds, but the first task for this extension is to estimate the vortex center location and its continuous variations in four-dimensional space so that the horizontal location of the vortex center can be expressed as a continuous function of height and time. To accomplish this task, a three-step method is developed in this paper. The method is applied to the Moore, Oklahoma, tornadic mesocyclone observed by the operational KTLX radar (Oklahoma City, Oklahoma) and the NSSL phased-array radar on 20 May 2013. The estimated vortex center trajectory at the ground level is verified with the tornado damage survey data. The estimated vortex center trajectories above the ground (up to 4-km height) reveal that the vortex core was initially tilted northeastward along the direction of the environmental flow and its vertical shear but became nearly vertical about 16 min later and 4 min before the vortex started to cause EF5 damages.
Abstract
The recently developed two-dimensional variational methods for analyzing vortex winds from radar-observed mesocyclones can be extended to analyze three-dimensional vortex winds, but the first task for this extension is to estimate the vortex center location and its continuous variations in four-dimensional space so that the horizontal location of the vortex center can be expressed as a continuous function of height and time. To accomplish this task, a three-step method is developed in this paper. The method is applied to the Moore, Oklahoma, tornadic mesocyclone observed by the operational KTLX radar (Oklahoma City, Oklahoma) and the NSSL phased-array radar on 20 May 2013. The estimated vortex center trajectory at the ground level is verified with the tornado damage survey data. The estimated vortex center trajectories above the ground (up to 4-km height) reveal that the vortex core was initially tilted northeastward along the direction of the environmental flow and its vertical shear but became nearly vertical about 16 min later and 4 min before the vortex started to cause EF5 damages.
Evaluation of a Support Vector Machine–Based Single-Doppler Wind Retrieval AlgorithmAbstract
Wind retrieval algorithms are required for Doppler weather radars. In this article, a new wind retrieval algorithm of single-Doppler radar with a support vector machine (SVM) is analyzed and compared with the original algorithm with the least squares technique. Through an analysis of coefficient matrices of equations corresponding to the optimization problems for the two algorithms, the new algorithm, which contains a proper penalization parameter, is found to effectively reduce the condition numbers of the matrices and thus has the ability to acquire accurate results, and the smaller the analysis volume is, the smaller the condition number of the matrix. This characteristic makes the new algorithm suitable to retrieve mesoscale and small-scale and high-resolution wind fields. Afterward, the two algorithms are applied to retrieval experiments to implement a comparison and a discussion. The results show that the penalization parameter cannot be too small, otherwise it may cause a large condition number; it cannot be too large either, otherwise it may change the properties of equations, leading to retrieved wind direction along the radial direction. Compared with the original algorithm, the new algorithm has definite superiority with the appropriate penalization parameters for small analysis volumes. When the suggested small analysis volume dimensions and penalization parameter values are adopted, the retrieval accuracy can be improved by 10 times more than the traditional method. As a result, the new algorithm has the capability to analyze the dynamical structures of severe weather, which needs high-resolution retrieval, and the potential for quantitative applications such as the assimilation in numerical models, but the retrieval accuracy needs to be further improved in the future.
Abstract
Wind retrieval algorithms are required for Doppler weather radars. In this article, a new wind retrieval algorithm of single-Doppler radar with a support vector machine (SVM) is analyzed and compared with the original algorithm with the least squares technique. Through an analysis of coefficient matrices of equations corresponding to the optimization problems for the two algorithms, the new algorithm, which contains a proper penalization parameter, is found to effectively reduce the condition numbers of the matrices and thus has the ability to acquire accurate results, and the smaller the analysis volume is, the smaller the condition number of the matrix. This characteristic makes the new algorithm suitable to retrieve mesoscale and small-scale and high-resolution wind fields. Afterward, the two algorithms are applied to retrieval experiments to implement a comparison and a discussion. The results show that the penalization parameter cannot be too small, otherwise it may cause a large condition number; it cannot be too large either, otherwise it may change the properties of equations, leading to retrieved wind direction along the radial direction. Compared with the original algorithm, the new algorithm has definite superiority with the appropriate penalization parameters for small analysis volumes. When the suggested small analysis volume dimensions and penalization parameter values are adopted, the retrieval accuracy can be improved by 10 times more than the traditional method. As a result, the new algorithm has the capability to analyze the dynamical structures of severe weather, which needs high-resolution retrieval, and the potential for quantitative applications such as the assimilation in numerical models, but the retrieval accuracy needs to be further improved in the future.
Deep Learning–Based Prediction of Kuroshio Path South of JapanAbstract
Variation of the Kuroshio path south of Japan has an important impact on weather, climate, and ecosystems due to its distinct features. Motivated by the ever-popular deep learning methods using neural network architectures in areas where more accurate reference data for oceanographic observations and reanalysis are available, we build four deep learning models based on the long short-term memory (LSTM) neural network, combined with the empirical orthogonal function (EOF) and complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), namely, the LSTM, EOF–LSTM, CEEMDAN–LSTM, and EOF–CEEMDAN–LSTM. Using these models, we conduct long-range predictions (120 days) of the Kuroshio path south of Japan based on 50-yr ocean reanalysis and nearly 15 years of satellite altimeter data. We show that the EOF–CEEMDAN–LSTM performs the best among the four models, by attaining approximately 0.739 anomaly correlation coefficient and 0.399° root-mean-square error for the 120-day prediction of the Kuroshio path south of Japan. The hindcasts of the EOF–CEEMDAN–LSTM are successful in reproducing the observed formation and decay of the Kuroshio large meander during 2004/05, and the formation of the latest large meander in 2017. Finally, we present predictions of the Kuroshio path south of Japan at 120-day lead time, which suggest that the Kuroshio will remain in the state of the large meander until November 2022.
Abstract
Variation of the Kuroshio path south of Japan has an important impact on weather, climate, and ecosystems due to its distinct features. Motivated by the ever-popular deep learning methods using neural network architectures in areas where more accurate reference data for oceanographic observations and reanalysis are available, we build four deep learning models based on the long short-term memory (LSTM) neural network, combined with the empirical orthogonal function (EOF) and complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), namely, the LSTM, EOF–LSTM, CEEMDAN–LSTM, and EOF–CEEMDAN–LSTM. Using these models, we conduct long-range predictions (120 days) of the Kuroshio path south of Japan based on 50-yr ocean reanalysis and nearly 15 years of satellite altimeter data. We show that the EOF–CEEMDAN–LSTM performs the best among the four models, by attaining approximately 0.739 anomaly correlation coefficient and 0.399° root-mean-square error for the 120-day prediction of the Kuroshio path south of Japan. The hindcasts of the EOF–CEEMDAN–LSTM are successful in reproducing the observed formation and decay of the Kuroshio large meander during 2004/05, and the formation of the latest large meander in 2017. Finally, we present predictions of the Kuroshio path south of Japan at 120-day lead time, which suggest that the Kuroshio will remain in the state of the large meander until November 2022.
Abstract
In recent years, the Earth System Research Laboratory (ESRL) of the National Oceanic and Atmospheric Administration (NOAA) has developed a space and time mesoscale analysis system (STMAS), which is currently a sequential three-dimensional variational data assimilation (3DVAR) system and is developing into a sequential 4DVAR in the near future. It is implemented by using a multigrid method based on a variational approach to generate grid analyses. This study is to test how STMAS deals with 2D Doppler radar radial velocity and to what degree the 2D Doppler radar radial velocity can improve the conventional (in situ) observation analysis. Two idealized experiments and one experiment with real Doppler radar radial velocity data, handled by STMAS, demonstrated significant improvement of the conventional observation analysis. Because the radar radial wind data can provide additional wind information (even it is incomplete: e.g., missing tangential wind vector), the analyses by assimilating both radial wind data and conventional data showed better results than those by assimilating only conventional data. Especially in the case of sparse conventional data, radar radial wind data can provide significant information and improve the analyses considerably.
Abstract
In recent years, the Earth System Research Laboratory (ESRL) of the National Oceanic and Atmospheric Administration (NOAA) has developed a space and time mesoscale analysis system (STMAS), which is currently a sequential three-dimensional variational data assimilation (3DVAR) system and is developing into a sequential 4DVAR in the near future. It is implemented by using a multigrid method based on a variational approach to generate grid analyses. This study is to test how STMAS deals with 2D Doppler radar radial velocity and to what degree the 2D Doppler radar radial velocity can improve the conventional (in situ) observation analysis. Two idealized experiments and one experiment with real Doppler radar radial velocity data, handled by STMAS, demonstrated significant improvement of the conventional observation analysis. Because the radar radial wind data can provide additional wind information (even it is incomplete: e.g., missing tangential wind vector), the analyses by assimilating both radial wind data and conventional data showed better results than those by assimilating only conventional data. Especially in the case of sparse conventional data, radar radial wind data can provide significant information and improve the analyses considerably.
Weather-Induced Transport through a Tidal Channel Calibrated by an Unmanned BoatAbstract
An unmanned surface vehicle (USV) was designed and constructed to operate continuously for covering both flood and ebb and preferably a complete tidal cycle (e.g., ~24 h) to measure the vertical profiles of horizontal flow velocity. It was applied in a tidal channel at Port Fourchon, Louisiana. A bottom-mounted ADCP was deployed for 515 days. The first EOF mode of the velocity profiles showed a barotropic type of flow that explained more than 98.2% of the variability. The second mode showed a typical estuarine flow with two layers, which explained 0.47% of the variability. Using a linear regression of the total transport from the USV with the vertically averaged velocity from the bottom-mounted ADCP, with an R-squared value of 98%, the total along-channel transport throughout the deployment was calculated. A low-pass filtering of the transport allowed for examining the impact of 76 events with cold, warm, or combined cold–warm fronts passing the area. The top seven most severe events were discussed, as their associated transports obviously stood out in the time series, indicating the importance of weather. It is shown that large-scale weather systems with frontal lines of ~1500–3000-km horizontal length scale control the subtidal transport in the area. Cold (warm) fronts tend to generate outward (inward) transports, followed by a rebound. The maximum coherence between the atmospheric forcing and the ocean response reached ~71%–84%, which occurred at about a frequency f of ~0.29 cycle per day or T of ~3.4 days in the period, consistent with the atmospheric frontal return periods (~3–7 days).
Abstract
An unmanned surface vehicle (USV) was designed and constructed to operate continuously for covering both flood and ebb and preferably a complete tidal cycle (e.g., ~24 h) to measure the vertical profiles of horizontal flow velocity. It was applied in a tidal channel at Port Fourchon, Louisiana. A bottom-mounted ADCP was deployed for 515 days. The first EOF mode of the velocity profiles showed a barotropic type of flow that explained more than 98.2% of the variability. The second mode showed a typical estuarine flow with two layers, which explained 0.47% of the variability. Using a linear regression of the total transport from the USV with the vertically averaged velocity from the bottom-mounted ADCP, with an R-squared value of 98%, the total along-channel transport throughout the deployment was calculated. A low-pass filtering of the transport allowed for examining the impact of 76 events with cold, warm, or combined cold–warm fronts passing the area. The top seven most severe events were discussed, as their associated transports obviously stood out in the time series, indicating the importance of weather. It is shown that large-scale weather systems with frontal lines of ~1500–3000-km horizontal length scale control the subtidal transport in the area. Cold (warm) fronts tend to generate outward (inward) transports, followed by a rebound. The maximum coherence between the atmospheric forcing and the ocean response reached ~71%–84%, which occurred at about a frequency f of ~0.29 cycle per day or T of ~3.4 days in the period, consistent with the atmospheric frontal return periods (~3–7 days).
Abstract
Correlation scales have been used in the traditional scheme of three-dimensional variational data assimilation (3DVAR) to estimate the background (or first guess) error covariance matrix (the 𝗕 matrix in brief) for the numerical forecast and reanalysis of ocean for decades. However, it is challenging to implement this scheme. On the one hand, determining the correlation scales accurately can be difficult. On the other hand, the positive definite of the 𝗕 matrix cannot be guaranteed unless the correlation scales are sufficiently small. Xie et al. indicated that a traditional 3DVAR only corrects certain wavelength errors, and its accuracy depends on the accuracy of the 𝗕 matrix. Generally speaking, the shortwave error cannot be sufficiently corrected until the longwave error is corrected. An inaccurate 𝗕 matrix may mistake longwave errors as shortwave ones, resulting in erroneous analyses.
A new 3DVAR data assimilation scheme, called a multigrid data assimilation scheme, is proposed in this paper for quickly minimizing longwave and shortwave errors successively. By assimilating the sea surface temperature and temperature profile observations into a numerical model of the China Seas, this scheme is applied to a retroactive real-time forecast experiment and favorable results are obtained. Compared to the traditional scheme of 3DVAR, this new scheme has higher forecast accuracy and lower root-mean-square errors. Note that the new scheme demonstrates greatly improved numerical efficiency in the analysis procedure.
Abstract
Correlation scales have been used in the traditional scheme of three-dimensional variational data assimilation (3DVAR) to estimate the background (or first guess) error covariance matrix (the 𝗕 matrix in brief) for the numerical forecast and reanalysis of ocean for decades. However, it is challenging to implement this scheme. On the one hand, determining the correlation scales accurately can be difficult. On the other hand, the positive definite of the 𝗕 matrix cannot be guaranteed unless the correlation scales are sufficiently small. Xie et al. indicated that a traditional 3DVAR only corrects certain wavelength errors, and its accuracy depends on the accuracy of the 𝗕 matrix. Generally speaking, the shortwave error cannot be sufficiently corrected until the longwave error is corrected. An inaccurate 𝗕 matrix may mistake longwave errors as shortwave ones, resulting in erroneous analyses.
A new 3DVAR data assimilation scheme, called a multigrid data assimilation scheme, is proposed in this paper for quickly minimizing longwave and shortwave errors successively. By assimilating the sea surface temperature and temperature profile observations into a numerical model of the China Seas, this scheme is applied to a retroactive real-time forecast experiment and favorable results are obtained. Compared to the traditional scheme of 3DVAR, this new scheme has higher forecast accuracy and lower root-mean-square errors. Note that the new scheme demonstrates greatly improved numerical efficiency in the analysis procedure.
Abstract
A recursive filter or parameterized curve fitting technique is usually used in a three-dimensional variational data assimilation (3DVAR) scheme to approximate the background error covariance, which can only represent the errors of an ocean field over a predetermined scale. Without an accurate flow-dependent error covariance that is also local and time dependent, a 3DVAR system may not provide good analyses because it is optimal only under the assumption of an accurate covariance. In this study, a sequential 3DVAR (S3DVAR) is formulated in model grid space to examine if there is useful information that can be extracted from the observation. This formulation is composed of a series of 3DVARs, each of which uses recursive filters with different length scales. It can provide an inhomogeneous and anisotropic analysis for the wavelengths that can be resolved by the observation network, just as with the conventional Barnes analysis or successive corrections. Being a variational formulation, S3DVAR can deal with data globally with an explicit specification of the observation errors; explicit physical balances or constraints; and advanced datasets, such as satellite and radar. Even though the S3DVAR analysis can be viewed as a set of isotropic functions superpositioned together, this superposition is not prespecified as in a single 3DVAR approach but is determined by the information that can be resolved by observation. The S3DVAR is adopted in a global sea surface temperature (SST) data assimilation system, into which the shipboard SSTs and the 4-km Advanced Very High Resolution Radiometer (AVHRR) Pathfinder daily SSTs are assimilated, respectively. The results demonstrate that the proposed S3DVAR works better in practice than a single 3DVAR.
Abstract
A recursive filter or parameterized curve fitting technique is usually used in a three-dimensional variational data assimilation (3DVAR) scheme to approximate the background error covariance, which can only represent the errors of an ocean field over a predetermined scale. Without an accurate flow-dependent error covariance that is also local and time dependent, a 3DVAR system may not provide good analyses because it is optimal only under the assumption of an accurate covariance. In this study, a sequential 3DVAR (S3DVAR) is formulated in model grid space to examine if there is useful information that can be extracted from the observation. This formulation is composed of a series of 3DVARs, each of which uses recursive filters with different length scales. It can provide an inhomogeneous and anisotropic analysis for the wavelengths that can be resolved by the observation network, just as with the conventional Barnes analysis or successive corrections. Being a variational formulation, S3DVAR can deal with data globally with an explicit specification of the observation errors; explicit physical balances or constraints; and advanced datasets, such as satellite and radar. Even though the S3DVAR analysis can be viewed as a set of isotropic functions superpositioned together, this superposition is not prespecified as in a single 3DVAR approach but is determined by the information that can be resolved by observation. The S3DVAR is adopted in a global sea surface temperature (SST) data assimilation system, into which the shipboard SSTs and the 4-km Advanced Very High Resolution Radiometer (AVHRR) Pathfinder daily SSTs are assimilated, respectively. The results demonstrate that the proposed S3DVAR works better in practice than a single 3DVAR.
Abstract
A new, fully conserved minimal adjustment scheme with temperature and salinity (T, S) coherency is presented for eliminating false static instability generated from analyzing and assimilating stable ocean (T, S) profiles data, that is, from generalized averaging over purely observed data (data analysis) or over modeled/observed data (data assimilation). This approach consists of a variational method with (a) fully (heat, salt, and potential energy) conserved conditions, (b) minimal adjustment, and (c) (T, S) coherency. Comparison with three existing schemes (minimal adjustment, conserved minimal adjustment, and convective adjustment) using observational profiles and a simple one-dimensional ocean mixed layer model shows the superiority of this new scheme.
Abstract
A new, fully conserved minimal adjustment scheme with temperature and salinity (T, S) coherency is presented for eliminating false static instability generated from analyzing and assimilating stable ocean (T, S) profiles data, that is, from generalized averaging over purely observed data (data analysis) or over modeled/observed data (data assimilation). This approach consists of a variational method with (a) fully (heat, salt, and potential energy) conserved conditions, (b) minimal adjustment, and (c) (T, S) coherency. Comparison with three existing schemes (minimal adjustment, conserved minimal adjustment, and convective adjustment) using observational profiles and a simple one-dimensional ocean mixed layer model shows the superiority of this new scheme.
Abstract
A variational method is used to estimate wave-affected parameters in a two-equation turbulence model with assimilation of temperature data into an ocean boundary layer model. Enhancement of turbulent kinetic energy dissipation due to breaking waves is considered. The Mellor–Yamada level 2.5 turbulence closure scheme (MY2.5) with the two uncertain wave-affected parameters (wave energy factor α and Charnock coefficient β) is selected as the two-equation turbulence model for this study. Two types of experiments are conducted. First, within an identical synthetic experiment framework, the upper-layer temperature “observations” in summer generated by a “truth” model are assimilated into a biased simulation model to investigate if (α, β) can be successfully estimated using the variational method. Second, real temperature profiles from Ocean Weather Station Papa are assimilated into the biased simulation model to obtain the optimal wave-affected parameters. With the optimally estimated parameters, the upper-layer temperature can be well predicted. Furthermore, the horizontal distribution of the wave-affected parameters employed in a high-order turbulence closure scheme can be estimated optimally by using the four-dimensional variational method that assimilates the upper-layer available temperature data into an ocean general circulation model.
Abstract
A variational method is used to estimate wave-affected parameters in a two-equation turbulence model with assimilation of temperature data into an ocean boundary layer model. Enhancement of turbulent kinetic energy dissipation due to breaking waves is considered. The Mellor–Yamada level 2.5 turbulence closure scheme (MY2.5) with the two uncertain wave-affected parameters (wave energy factor α and Charnock coefficient β) is selected as the two-equation turbulence model for this study. Two types of experiments are conducted. First, within an identical synthetic experiment framework, the upper-layer temperature “observations” in summer generated by a “truth” model are assimilated into a biased simulation model to investigate if (α, β) can be successfully estimated using the variational method. Second, real temperature profiles from Ocean Weather Station Papa are assimilated into the biased simulation model to obtain the optimal wave-affected parameters. With the optimally estimated parameters, the upper-layer temperature can be well predicted. Furthermore, the horizontal distribution of the wave-affected parameters employed in a high-order turbulence closure scheme can be estimated optimally by using the four-dimensional variational method that assimilates the upper-layer available temperature data into an ocean general circulation model.
