Browse
Abstract
This study analyzes horizontal and vertical wind-driven circulation responses in small semienclosed bays, the associated offshore dynamic conditions, and the relative importance of each term in the momentum balance equations using a multiplatform observational system. The observational platform consists of three ADCPs and a land-based radar monitoring the velocity field within the bay and in the contiguous offshore area. The wind-driven patterns in the bay can switch from a barotropic cyclonic or anticyclonic circulation to a two-layer baroclinic mode response as a function of the wind regime (its direction and magnitude). For the baroclinic mode, the vertical location of the inflection point in the velocity profile can vary according to the proximity of the boundary current to the entrance of the bay. The influence of offshore combined meteorological and marine conditions on the inner-bay dynamics is evidenced under moderate to strong wind conditions and is almost nonexistent under negligible wind. The momentum balance analysis as well as the nondimensional numbers evidence the impact of wind stress, coastline shape, stratification, and the nonlinear advective terms. Advection can be at the same order of magnitude as pressure gradient, Coriolis, or wind stress terms and can be greater than the bottom stress terms. The nonlinear terms in the momentum equations are frequently neglected when analyzing wind-driven circulation by means of in situ data or analytical models.
Abstract
This study analyzes horizontal and vertical wind-driven circulation responses in small semienclosed bays, the associated offshore dynamic conditions, and the relative importance of each term in the momentum balance equations using a multiplatform observational system. The observational platform consists of three ADCPs and a land-based radar monitoring the velocity field within the bay and in the contiguous offshore area. The wind-driven patterns in the bay can switch from a barotropic cyclonic or anticyclonic circulation to a two-layer baroclinic mode response as a function of the wind regime (its direction and magnitude). For the baroclinic mode, the vertical location of the inflection point in the velocity profile can vary according to the proximity of the boundary current to the entrance of the bay. The influence of offshore combined meteorological and marine conditions on the inner-bay dynamics is evidenced under moderate to strong wind conditions and is almost nonexistent under negligible wind. The momentum balance analysis as well as the nondimensional numbers evidence the impact of wind stress, coastline shape, stratification, and the nonlinear advective terms. Advection can be at the same order of magnitude as pressure gradient, Coriolis, or wind stress terms and can be greater than the bottom stress terms. The nonlinear terms in the momentum equations are frequently neglected when analyzing wind-driven circulation by means of in situ data or analytical models.
Abstract
This work evaluates the fidelity of various upper-ocean turbulence parameterizations subject to realistic monsoon forcing and presents a finite-time ensemble vector (EV) method to better manage the design and numerical principles of these parameterizations. The EV method emphasizes the dynamics of a turbulence closure multimodel ensemble and is applied to evaluate 10 different ocean surface boundary layer (OSBL) parameterizations within a single-column (SC) model against two boundary layer large-eddy simulations (LES). Both LES include realistic surface forcing, but one includes wind-driven shear turbulence only, while the other includes additional Stokes forcing through the wave-average equations that generate Langmuir turbulence. The finite-time EV framework focuses on what constitutes the local behavior of the mixed layer dynamical system and isolates the forcing and ocean state conditions where turbulence parameterizations most disagree. Identifying disagreement provides the potential to evaluate SC models comparatively against the LES. Observations collected during the 2018 monsoon onset in the Bay of Bengal provide a case study to evaluate models under realistic and variable forcing conditions. The case study results highlight two regimes where models disagree 1) during wind-driven deepening of the mixed layer and 2) under strong diurnal forcing.
Abstract
This work evaluates the fidelity of various upper-ocean turbulence parameterizations subject to realistic monsoon forcing and presents a finite-time ensemble vector (EV) method to better manage the design and numerical principles of these parameterizations. The EV method emphasizes the dynamics of a turbulence closure multimodel ensemble and is applied to evaluate 10 different ocean surface boundary layer (OSBL) parameterizations within a single-column (SC) model against two boundary layer large-eddy simulations (LES). Both LES include realistic surface forcing, but one includes wind-driven shear turbulence only, while the other includes additional Stokes forcing through the wave-average equations that generate Langmuir turbulence. The finite-time EV framework focuses on what constitutes the local behavior of the mixed layer dynamical system and isolates the forcing and ocean state conditions where turbulence parameterizations most disagree. Identifying disagreement provides the potential to evaluate SC models comparatively against the LES. Observations collected during the 2018 monsoon onset in the Bay of Bengal provide a case study to evaluate models under realistic and variable forcing conditions. The case study results highlight two regimes where models disagree 1) during wind-driven deepening of the mixed layer and 2) under strong diurnal forcing.
Abstract
The continuous, moored observation revealed significant variability in the strength of the Atlantic meridional overturning circulation (AMOC). The cause of such AMOC variability is an extensively studied subject. This study focuses on the short-term variability, which ranges up to seasonal and interannual time scales. A mechanism is proposed from the perspective of ocean water redistribution by layers. By offering explanations for four phenomena of AMOC variability in the subtropical and tropical oceans (seasonality, meridional coherence, layered-transport compensation as observed at 26.5°N, and the 2009/10 downturn that occurred at 26.5°N), this mechanism suggests that the short-term AMOC variabilities in the entire subtropical and tropical regions are governed by a basinwide adiabatic water redistribution process, or the so-called sloshing dynamics, rather than diapycnal processes.
Significance Statement
The Atlantic meridional overturning circulation (AMOC) is a key component in the global climate system due to its immense power in redistributing heat meridionally, which contributes to the hospitable climate of the United Kingdom and western Europe. Therefore, any changes in AMOC can have significant impacts on both global and local climate variability. Here I propose a mechanism to explain the short-term (up to interannual) AMOC variability in the subtropical and tropical regions from the perspective of ocean water redistribution. This mechanism suggests that the short-term variability of AMOC strength is dominated by an adiabatic process, and thus, its large-amplitude variation is mostly a reversible process. In other words, AMOC may be more resilient to short-term variability than previously believed, and it could recover autonomously from the abrupt changes.
Abstract
The continuous, moored observation revealed significant variability in the strength of the Atlantic meridional overturning circulation (AMOC). The cause of such AMOC variability is an extensively studied subject. This study focuses on the short-term variability, which ranges up to seasonal and interannual time scales. A mechanism is proposed from the perspective of ocean water redistribution by layers. By offering explanations for four phenomena of AMOC variability in the subtropical and tropical oceans (seasonality, meridional coherence, layered-transport compensation as observed at 26.5°N, and the 2009/10 downturn that occurred at 26.5°N), this mechanism suggests that the short-term AMOC variabilities in the entire subtropical and tropical regions are governed by a basinwide adiabatic water redistribution process, or the so-called sloshing dynamics, rather than diapycnal processes.
Significance Statement
The Atlantic meridional overturning circulation (AMOC) is a key component in the global climate system due to its immense power in redistributing heat meridionally, which contributes to the hospitable climate of the United Kingdom and western Europe. Therefore, any changes in AMOC can have significant impacts on both global and local climate variability. Here I propose a mechanism to explain the short-term (up to interannual) AMOC variability in the subtropical and tropical regions from the perspective of ocean water redistribution. This mechanism suggests that the short-term variability of AMOC strength is dominated by an adiabatic process, and thus, its large-amplitude variation is mostly a reversible process. In other words, AMOC may be more resilient to short-term variability than previously believed, and it could recover autonomously from the abrupt changes.
Abstract
The large-scale ocean circulation in an ocean basin was previously thought to be impacted cumulatively by all the overlying tropical cyclones (TCs). Based on idealized numerical experiments and altimetry observation, this study reveals that, unnecessarily by cumulative impacts, a single TC actually has the ability to plow the large-scale sea surface height (SSH) field due to the TC-induced geostrophic response. This ability is dictated by the along-track length scale of the geostrophic response, i.e., the total track length. Some of the observed along-track signals, including the SSH trough, jet, and SSH rise, can confirm the TC-induced large-scale impacts. Shortly after the TC passage, the observable large-scale signals are generally the SSH trough. However, the jet and the SSH rise easily emerge from the evolved SSH trough due to Rossby wave dispersion. By identifying and tracking the observable signals, this study demonstrates that the subtropical gyre primarily over 4°–20°N, 122°E–180° is plowed by nine typhoons (2015) into several large blocks of SSH troughs and SSH rises. These long-lived SSH troughs and SSH rises dominate the upper-layer circulation from April to December in 2015. If the large-scale signals cannot be observed, the estimated TC-induced mean SSH decreases suggest that the large-scale impacts may still exist but merely cannot be seen intuitively. This study provides compelling observational evidence for the TC-induced large-scale impacts, further highlighting that TCs may play a nonnegligible role in the upper-ocean dynamics in the subtropical gyre.
Significance Statement
This study aims to demonstrate the ability of a typhoon to affect the large-scale ocean dynamics. The ability manifests as some along-track signals in altimetry observations, including sea surface height trough, jet, and sea surface height rise, which can be frequently observed after some typhoons in 2015. The sea surface height field in the western North Pacific is continuously plowed by these typhoons into several large blocks of sea surface height troughs and rises. These long-lived sea surface height troughs and rises dominate the upper-layer circulation from April to December in 2015. This study indicates that typhoons play a vital role in the upper-ocean dynamics in the western North Pacific.
Abstract
The large-scale ocean circulation in an ocean basin was previously thought to be impacted cumulatively by all the overlying tropical cyclones (TCs). Based on idealized numerical experiments and altimetry observation, this study reveals that, unnecessarily by cumulative impacts, a single TC actually has the ability to plow the large-scale sea surface height (SSH) field due to the TC-induced geostrophic response. This ability is dictated by the along-track length scale of the geostrophic response, i.e., the total track length. Some of the observed along-track signals, including the SSH trough, jet, and SSH rise, can confirm the TC-induced large-scale impacts. Shortly after the TC passage, the observable large-scale signals are generally the SSH trough. However, the jet and the SSH rise easily emerge from the evolved SSH trough due to Rossby wave dispersion. By identifying and tracking the observable signals, this study demonstrates that the subtropical gyre primarily over 4°–20°N, 122°E–180° is plowed by nine typhoons (2015) into several large blocks of SSH troughs and SSH rises. These long-lived SSH troughs and SSH rises dominate the upper-layer circulation from April to December in 2015. If the large-scale signals cannot be observed, the estimated TC-induced mean SSH decreases suggest that the large-scale impacts may still exist but merely cannot be seen intuitively. This study provides compelling observational evidence for the TC-induced large-scale impacts, further highlighting that TCs may play a nonnegligible role in the upper-ocean dynamics in the subtropical gyre.
Significance Statement
This study aims to demonstrate the ability of a typhoon to affect the large-scale ocean dynamics. The ability manifests as some along-track signals in altimetry observations, including sea surface height trough, jet, and sea surface height rise, which can be frequently observed after some typhoons in 2015. The sea surface height field in the western North Pacific is continuously plowed by these typhoons into several large blocks of sea surface height troughs and rises. These long-lived sea surface height troughs and rises dominate the upper-layer circulation from April to December in 2015. This study indicates that typhoons play a vital role in the upper-ocean dynamics in the western North Pacific.
Abstract
The East Mediterranean Sea (EMS) circulation has previously been characterized as dominated by gyres, mesoscale eddies, and disjoint boundary currents. We develop nested high-resolution numerical simulations in the EMS to examine the circulation variability with an emphasis on the yet unexplored regional submesoscale currents. Rather than several disjoint currents, a continuous cyclonic boundary current (BC) encircling the Levantine basin is identified in both model solution and altimetry data. This EMS BC advects eddy chains downstream and is identified as a principal source of regional mesoscale and submesoscale current variability. During the seasonal fall to winter mixed layer deepening, energetic submesoscale [O(10) km] eddies, fronts, and filaments emerge throughout the basin, characterized by O(1) Rossby numbers. A submesoscale time scale range of ≈1–5 days is identified using spatiotemporal analysis of the numerical solutions and confirmed through mooring data. The submesoscale kinetic energy (KE) wavenumber (k) spectral slope is found to be k −2, shallower than the quasigeostrophic-like ∼k −3 slope diagnosed in summer. The shallowness of the winter spectral slope is shown to be due to divergent subinertial motions, consistent with the Boyd theoretical model, rather than with the surface quasigeostrophic model. Using a coarse-graining approach, we diagnose a seasonal inverse (forward) KE cascade above (below) 30-km scales due to rotational (divergent) motions and show that these commence after completion of the fall submesoscale energization. We also show that at scales larger than several hundred kilometers, the spectral density becomes near constant and a weak forward cascade occurs, from gyre scales to mesoscales.
Abstract
The East Mediterranean Sea (EMS) circulation has previously been characterized as dominated by gyres, mesoscale eddies, and disjoint boundary currents. We develop nested high-resolution numerical simulations in the EMS to examine the circulation variability with an emphasis on the yet unexplored regional submesoscale currents. Rather than several disjoint currents, a continuous cyclonic boundary current (BC) encircling the Levantine basin is identified in both model solution and altimetry data. This EMS BC advects eddy chains downstream and is identified as a principal source of regional mesoscale and submesoscale current variability. During the seasonal fall to winter mixed layer deepening, energetic submesoscale [O(10) km] eddies, fronts, and filaments emerge throughout the basin, characterized by O(1) Rossby numbers. A submesoscale time scale range of ≈1–5 days is identified using spatiotemporal analysis of the numerical solutions and confirmed through mooring data. The submesoscale kinetic energy (KE) wavenumber (k) spectral slope is found to be k −2, shallower than the quasigeostrophic-like ∼k −3 slope diagnosed in summer. The shallowness of the winter spectral slope is shown to be due to divergent subinertial motions, consistent with the Boyd theoretical model, rather than with the surface quasigeostrophic model. Using a coarse-graining approach, we diagnose a seasonal inverse (forward) KE cascade above (below) 30-km scales due to rotational (divergent) motions and show that these commence after completion of the fall submesoscale energization. We also show that at scales larger than several hundred kilometers, the spectral density becomes near constant and a weak forward cascade occurs, from gyre scales to mesoscales.
Abstract
Indonesian Throughflow (ITF) waters move along multiple pathways within the Indian Ocean. The western route is within the thermocline of the South Equatorial Current (SEC), and the southern route is via injection into the Leeuwin Current (LC) along western Australia. We use gridded Argo data to examine heat content anomaly (HCa) within three boxes in the eastern Indian Ocean, one adjacent to the ITF outflow from the Indonesian Seas (ITF box), the second in the eastern portion of the SEC (SEC box), and the third in the LC (LC box). Although interannual HCa variability in the SEC and ITF boxes is well correlated, a large increase in HCa within the ITF box does not appear in the SEC box in 2011 but is evident in the LC box. The 2011 change in the SEC–LC partitioning is investigated using GODAS reanalysis by examining the strength of the SEC and LC during a 2009 HCa increase within the ITF box and the subsequent increase in 2011. During 2009, a strong SEC and weakened LC spread the increased ITF HCa into the central Indian Ocean, whereas a weak SEC and strengthened LC during 2011 transmit the HCa signal to the south. Near-surface winds and mean sea level pressure from NCEP–NCAR reanalysis reveal that Ningaloo Niño events led to shifts in ocean circulation during 2000 and 2011. LC and SEC exports show a high negative correlation at interannual time scales, indicating that a reduction of outflow from one pathway is partially compensated by an increase from the other.
Abstract
Indonesian Throughflow (ITF) waters move along multiple pathways within the Indian Ocean. The western route is within the thermocline of the South Equatorial Current (SEC), and the southern route is via injection into the Leeuwin Current (LC) along western Australia. We use gridded Argo data to examine heat content anomaly (HCa) within three boxes in the eastern Indian Ocean, one adjacent to the ITF outflow from the Indonesian Seas (ITF box), the second in the eastern portion of the SEC (SEC box), and the third in the LC (LC box). Although interannual HCa variability in the SEC and ITF boxes is well correlated, a large increase in HCa within the ITF box does not appear in the SEC box in 2011 but is evident in the LC box. The 2011 change in the SEC–LC partitioning is investigated using GODAS reanalysis by examining the strength of the SEC and LC during a 2009 HCa increase within the ITF box and the subsequent increase in 2011. During 2009, a strong SEC and weakened LC spread the increased ITF HCa into the central Indian Ocean, whereas a weak SEC and strengthened LC during 2011 transmit the HCa signal to the south. Near-surface winds and mean sea level pressure from NCEP–NCAR reanalysis reveal that Ningaloo Niño events led to shifts in ocean circulation during 2000 and 2011. LC and SEC exports show a high negative correlation at interannual time scales, indicating that a reduction of outflow from one pathway is partially compensated by an increase from the other.
Abstract
Turbulence in the ocean surface layer is forced by a mixture of buoyancy, wind, and wave processes that evolves over time scales from the diurnal scale of buoyancy forcing, through storm time scales, to the annual cycle. This study seeks a predictor for root-mean-square w (rmsw), a time and surface layer average of turbulent vertical velocity w measured by bottom-mounted vertical-beam acoustic Doppler current profilers, in terms of concurrently measured surface forcing fields. Data used are from two coastal sites, one shallow (LEO, 15-m depth) and one deeper (R2, 26-m depth). The analysis demonstrates that it is possible to predict observed rmsw with a simple linear combination of two scale velocities, one the convective scale velocity
Abstract
Turbulence in the ocean surface layer is forced by a mixture of buoyancy, wind, and wave processes that evolves over time scales from the diurnal scale of buoyancy forcing, through storm time scales, to the annual cycle. This study seeks a predictor for root-mean-square w (rmsw), a time and surface layer average of turbulent vertical velocity w measured by bottom-mounted vertical-beam acoustic Doppler current profilers, in terms of concurrently measured surface forcing fields. Data used are from two coastal sites, one shallow (LEO, 15-m depth) and one deeper (R2, 26-m depth). The analysis demonstrates that it is possible to predict observed rmsw with a simple linear combination of two scale velocities, one the convective scale velocity
Abstract
Sea level variabilities in the southwest Pacific contribute to the variations of equatorial current bifurcation and the Indonesian Throughflow transport. These processes are closely related to the recharge/discharge of equatorial heat content and dynamic distribution of anthropogenic ocean heating over the Indo-Pacific basin, thus being of profound significance for climate variability and change. Here we identify the major features of seasonal and interannual sea level variabilities in this region, confirming the dominance of the first baroclinic mode in the tropics (contributing 60%–80% of the variances) and higher baroclinic modes in the extratropics (40%–60% of the seasonal variance). Seasonally, except in the western Coral Sea where the Ekman pumping is significant, the wind-driven first-mode baroclinic Rossby waves originating to the east of the date line control the sea level variations over most tropical Pacific regions. In the domain where the 1.5-layer reduced gravity model becomes deficient, the surface heat fluxes dominate, explaining ∼40%–80% of sea level variance. For interannual variability, ∼40%–60% of the variance are El Niño–Southern Oscillation (ENSO) related. The wind-driven Rossby and Kelvin waves east of the date line explain ∼40%–78% of the interannual variance in the tropical Pacific. Outside the tropics, small-scale diffusive processes are presumed critical for interannual variability according to a thermodynamic analysis using an eddy-permitting ocean model simulation. Further process and predictive understandings can be achieved with the coupled climate models properly parameterizing the subgrid-scale processes.
Abstract
Sea level variabilities in the southwest Pacific contribute to the variations of equatorial current bifurcation and the Indonesian Throughflow transport. These processes are closely related to the recharge/discharge of equatorial heat content and dynamic distribution of anthropogenic ocean heating over the Indo-Pacific basin, thus being of profound significance for climate variability and change. Here we identify the major features of seasonal and interannual sea level variabilities in this region, confirming the dominance of the first baroclinic mode in the tropics (contributing 60%–80% of the variances) and higher baroclinic modes in the extratropics (40%–60% of the seasonal variance). Seasonally, except in the western Coral Sea where the Ekman pumping is significant, the wind-driven first-mode baroclinic Rossby waves originating to the east of the date line control the sea level variations over most tropical Pacific regions. In the domain where the 1.5-layer reduced gravity model becomes deficient, the surface heat fluxes dominate, explaining ∼40%–80% of sea level variance. For interannual variability, ∼40%–60% of the variance are El Niño–Southern Oscillation (ENSO) related. The wind-driven Rossby and Kelvin waves east of the date line explain ∼40%–78% of the interannual variance in the tropical Pacific. Outside the tropics, small-scale diffusive processes are presumed critical for interannual variability according to a thermodynamic analysis using an eddy-permitting ocean model simulation. Further process and predictive understandings can be achieved with the coupled climate models properly parameterizing the subgrid-scale processes.
Abstract
Mesoscale eddies can alter the propagation of wind-generated near-inertial waves (NIWs). Different from previous studies, the subsurface mooring observed NIWs are generated outside an anticyclonic eddy (ACE) and then interact with the arriving ACE. It is found that with the arrival of the ACE, the NIWs accelerate to propagate downward and the maximum vertical wavelength and group velocity of NIWs reach ∼500 m and ∼35 m day−1, respectively. When entering the core of the ACE, the near-inertial energy is trapped and finally stalls at a critical depth, which basically corresponds to the base of the ACE located at around 750-m depth. Through a ray-tracing model and dynamic analyses, this critical depth is much deeper than that of NIWs generated directly inside an ACE. By using depth–time varying stratification and relative vorticity, ray-tracing experiments further demonstrate that NIWs generated outside and passed over by an ACE can propagate to deep depths. Furthermore, energy budget analyses indicate that the net energy transfer from the ACE to NIWs plays an important role in the enhancement of downward-propagating near-inertial energy and its long-term persistence (∼45 days) in the critical layer. Within the critical layer, the enhancement of shear instability and nonlinear interactions among internal waves account for the loss of the trapped near-inertial energy and provide energy for furnishing deep ocean mixing.
Significance Statement
The interactions between near-inertial waves and a westward-moving anticyclonic eddy are investigated in this study. Knowledge about the propagation of near-inertial waves continues to be a topic of interest because near-inertial waves transfer energy from the mixed layer to the interior ocean, which is an important source of turbulent mixing. While much is known about how near-inertial energy propagates inside an anticyclonic eddy, few studies have examined how near-inertial energy propagates when it is generated outside an anticyclonic eddy and then enters the arriving anticyclonic eddy. In this study, the deep propagation and trapping of near-inertial energy by a westward-moving anticyclonic eddy is observed, which contributes greatly to the energy budget and the deep-ocean mixing.
Abstract
Mesoscale eddies can alter the propagation of wind-generated near-inertial waves (NIWs). Different from previous studies, the subsurface mooring observed NIWs are generated outside an anticyclonic eddy (ACE) and then interact with the arriving ACE. It is found that with the arrival of the ACE, the NIWs accelerate to propagate downward and the maximum vertical wavelength and group velocity of NIWs reach ∼500 m and ∼35 m day−1, respectively. When entering the core of the ACE, the near-inertial energy is trapped and finally stalls at a critical depth, which basically corresponds to the base of the ACE located at around 750-m depth. Through a ray-tracing model and dynamic analyses, this critical depth is much deeper than that of NIWs generated directly inside an ACE. By using depth–time varying stratification and relative vorticity, ray-tracing experiments further demonstrate that NIWs generated outside and passed over by an ACE can propagate to deep depths. Furthermore, energy budget analyses indicate that the net energy transfer from the ACE to NIWs plays an important role in the enhancement of downward-propagating near-inertial energy and its long-term persistence (∼45 days) in the critical layer. Within the critical layer, the enhancement of shear instability and nonlinear interactions among internal waves account for the loss of the trapped near-inertial energy and provide energy for furnishing deep ocean mixing.
Significance Statement
The interactions between near-inertial waves and a westward-moving anticyclonic eddy are investigated in this study. Knowledge about the propagation of near-inertial waves continues to be a topic of interest because near-inertial waves transfer energy from the mixed layer to the interior ocean, which is an important source of turbulent mixing. While much is known about how near-inertial energy propagates inside an anticyclonic eddy, few studies have examined how near-inertial energy propagates when it is generated outside an anticyclonic eddy and then enters the arriving anticyclonic eddy. In this study, the deep propagation and trapping of near-inertial energy by a westward-moving anticyclonic eddy is observed, which contributes greatly to the energy budget and the deep-ocean mixing.
Abstract
Directional wave spectra are of importance for numerous practical applications such as seafaring and ocean engineering. The wave spectral densities at a certain point in the open ocean are significantly correlated to the local wind field and historical remote wind field. This feature can be used to predict the wave spectrum at that point using the wind field. In this study, a convolutional neural network (CNN) model was established to estimate wave spectra at a target point using the wind field from the ERA5 dataset. A geospatial range where the wind could impact the target point was selected, and then the historical wind field data within the range were analyzed to extract the nonlinear quantitative relationships between wind fields and wave spectra. For the spectral densities at a given direction, the wind data along the direction where waves come from were used as the input of the CNN. The model was trained to minimize the mean square error between the CNN-predicted and ERA5 reanalysis spectral density. The data structure of the wind input is reorganized into a polar grid centered on the target point to make the model applicable to different open-ocean locations worldwide. The results show that the model can predict well the wave spectrum shapes and integral wave parameters. The model allows for the prediction of single-point wave spectra in the open ocean with low computational cost and can be helpful for the study of spectral wave climate.
Significance Statement
The directional wave spectra (DWS) describe the distribution of wave energy among different frequencies and directions. They are useful for many marine practical applications. Usually, DWS are modeled using numerical wave models (NWMs) based on wave action balance differential equations. Although contemporary NWMs perform well after years of development, their computational costs are relatively high. The fast-developed artificial intelligence (AI) might provide an alternative solution to this task. In this study, convolutional neural networks are used to model the DWS at some selected points in the open ocean. By “learning” from NWM data, AI can effectively simulate single-point DWS in open oceans with low computational cost, which can serve as a faster data-driven surrogate model in related applications.
Abstract
Directional wave spectra are of importance for numerous practical applications such as seafaring and ocean engineering. The wave spectral densities at a certain point in the open ocean are significantly correlated to the local wind field and historical remote wind field. This feature can be used to predict the wave spectrum at that point using the wind field. In this study, a convolutional neural network (CNN) model was established to estimate wave spectra at a target point using the wind field from the ERA5 dataset. A geospatial range where the wind could impact the target point was selected, and then the historical wind field data within the range were analyzed to extract the nonlinear quantitative relationships between wind fields and wave spectra. For the spectral densities at a given direction, the wind data along the direction where waves come from were used as the input of the CNN. The model was trained to minimize the mean square error between the CNN-predicted and ERA5 reanalysis spectral density. The data structure of the wind input is reorganized into a polar grid centered on the target point to make the model applicable to different open-ocean locations worldwide. The results show that the model can predict well the wave spectrum shapes and integral wave parameters. The model allows for the prediction of single-point wave spectra in the open ocean with low computational cost and can be helpful for the study of spectral wave climate.
Significance Statement
The directional wave spectra (DWS) describe the distribution of wave energy among different frequencies and directions. They are useful for many marine practical applications. Usually, DWS are modeled using numerical wave models (NWMs) based on wave action balance differential equations. Although contemporary NWMs perform well after years of development, their computational costs are relatively high. The fast-developed artificial intelligence (AI) might provide an alternative solution to this task. In this study, convolutional neural networks are used to model the DWS at some selected points in the open ocean. By “learning” from NWM data, AI can effectively simulate single-point DWS in open oceans with low computational cost, which can serve as a faster data-driven surrogate model in related applications.
