共查询到20条相似文献，搜索用时 406 毫秒
1.
Both finite-element and finite-difference numerical models are applied to simulate storm surges and associated currents generated
by tropical cyclones that struck the coast of Andhra Pradesh, located on the east coast of India. During a cyclone, the total
water level at any location on the coast is made up of the storm surge, surge–wind wave interaction and the tide. The advanced
circulation two-dimensional depth-integrated (ADCIRC-2DDI) model based on finite-element formulation and the two-dimensional
finite-difference model of storm surges developed at IIT Delhi, hereafter referred as IITD storm surge model, are used. These
models are driven by astronomical tides at the open ocean boundary and cyclonic asymmetric winds over the surface of the computational
domain. Comparison of model simulated sea-surface elevations with coarse and finer spatial resolutions suggests that the grid
resolution near the coast is very crucial for accurate determination of the surges in addition to the local bathymetry. The
model underpredicts surges, and the peak surge location shifts more to the right of the landfall as the spatial resolution
of the model becomes coarser. The numerical experiments also demonstrate that the ADCIRC model is robust over the IITD storm
surge model for surge computations as the coastline is better represented in the former. 相似文献
2.
Cyclone-generated surface waves are simulated using state-of-art SWAN (Simulating WAves Nearshore) model coupled with hydrodynamic
model inputs. A severe cyclonic storm passed over the Arabian Sea during 4–9th November 1982 is selected from UNISYS track
records. The cyclone lasted for nearly 6 days and subsided with a land fall at Gujarat coast, west coast of India. In this
study, cyclonic wind fields are generated using a well-established relationship suggested by Jelesnianski and Taylor (1973). The associated water level variations due to storm surge and surge generated currents are simulated using POM (Princeton
Ocean Model). The outputs are one-way coupled with the wave model SWAN for simulating wave parameters off Gujarat, north-east
basin of Arabian Sea. An extensive literature review is carried out on the progress and methodology adopted for storm wave
modelling and analysis. The results presented in this paper reveal the severity of the storm event and would be highly useful
for assessing the extreme wave event/climate especially for the south coast of Gujarat. 相似文献
3.
Judith Wolf 《Natural Hazards》2009,49(2):241-260
Wind waves and elevated water levels together can cause flooding in low-lying coastal areas, where the water level may be
a combination of mean sea level, tides and surges generated by storm events. In areas with a wide continental shelf a travelling
external surge may combine with the locally generated surge and waves and there can be significant interaction between the
propagation of the tide and surge. Wave height at the coast is controlled largely by water depth. So the effect of tides and
surges on waves must also be considered, while waves contribute to the total water level by means of wave setup through radiation
stress. These processes are well understood and accurately predicted by models, assuming good bathymetry and wind forcing
is available. Other interactions between surges and waves include the processes of surface wind-stress and bottom friction
as well as depth and current refraction of waves by surge water levels and currents, and some of the details of these processes
are still not well understood. The recent coastal flooding in Myanmar (May 2008) in the Irrawaddy River Delta is an example
of the severity of such events, with a surge of over 3 m exacerbated by heavy precipitation. Here, we review the existing
capability for combined modelling of tides, surges and waves, their interactions and the development of coupled models. 相似文献
4.
The influence of seasonal and cyclonic winds is studied on the characteristics of internal waves (IWs) over the western Bay of Bengal (BoB) by using MITgcm model. As the BoB experiences reversal of seasonal winds and also tropical cyclones during pre-monsoon and post-monsoon months, its effect is seen through the computation of spectral estimates of the IWs. It is seen that the peak estimate is associated with the semidiurnal frequency at all the depths and is found higher in May compared to November. This is attributed to the presence of shallow mixed layer depth and deep thermocline due to the upwelling favorable winds. The computation of isopycnal displacement infers that the internal tides are present from 40 to 120 m depth in case of upwelling favorable winds of May, whereas, the presence of internal tides is restricted between 90 and 120 m for the downwelling favorable winds of November. During May, the available potential energy is also seen in a narrow coastal stretch, whilst it is absent in November. During the Hudhud cyclone period of October 7–14, 2014, it is noticed from the spectral estimates that the IWs of tidal frequency are replaced by inertial frequency with a periodicity of about 2 days as a consequence of strong cyclonic winds. The progressive vector diagram shows the mean current is initially westward up to October 17, 2014 and then northeastward with well-defined clockwise circulation. The maximum radius of inertial oscillation of 15 km is observed. After the cyclone ceases, the estimate associated with inertial frequency slowly diminishes and enhances the estimates related to internal tides. The simulations also suggest that the internal tides are absent for about 6 weeks as a response of the cyclonic winds.
相似文献5.
The northeastern Gulf of Mexico in the USA is extremely susceptible to the impacts of tropical cyclones because of its unique geometric and topographic features. Focusing on Hurricanes Ivan (2004) and Katrina (2005), this paper has addressed four scientific questions on this area’s response to hurricanes: (1) How does the shallow, abandoned Mississippi delta contribute to the storm surge? (2) What was the controlling factor that caused the record-high storm surge of Hurricane Katrina? (3) Why are the responses of an estuary to Hurricanes Ivan and Katrina so different from the corresponding surges on the open coast? (4) How would the storm surge differ if Hurricane Katrina had taken a different course? Guided by field observations of winds, waves, water levels, and currents, two state-of-the-art numerical models for storm surges and wind waves have been coupled to hindcast the relevant hydrodynamic conditions, including storm surges, surface waves, and depth-averaged currents. Fairly good agreement between the modeled and measured surge hydrographs was found. The quantitative numerical simulations and simple qualitative analysis have revealed that the record-high storm surge of Hurricane Katrina was caused by the interaction of the surge with the extremely shallow, ancient deltaic lobe of Mississippi River. A hypothetical scenario formed by shifting the path of Hurricane Katrina to the observed path of Hurricane Frederic (1979) resulted in a much smaller surge than that observed in coastal Mississippi and Louisiana. However, this scenario did still result in a high surge near the head of Mobile Bay. One of the important lessons learned from Hurricane Katrina is that the Saffir–Simpson scale should be systematically revised to reflect the topographic and geometric features of a complex, heterogeneous coast, including the possible surge amplification in an estuary or a submerged river delta. 相似文献
6.
A three-dimensional numerical model is described to study theresponse of a coastal ocean excited by a tropical cyclone in the Bay of Bengal. The numericalexperiments have been carried out using the model to understand the dynamics and thermodynamics ofthe ocean due to different cyclonic systems approaching in different directions. In the firstexperiment, the model is used to simulate the vertical thermal structure of the ocean as a response ofpassage of the less intensified 1997 cyclone, which was skirting the east coast of India before crossingthe Bangladesh coast. The simulations are compared with the buoy data available during the storm period.In the next experiment, it is considered an idealized cyclone with hurricane winds movingnormal to the east coast of India crossing between Visakhapatnam and Kakinada to evolve thermalstructure and currents of the ocean. A net decrease of the SST of 6–7 °C is simulated whenthe severe cyclonic storm moved over the coastal ocean. 相似文献
7.
P N Ananth Rao V S N Tatavarti J Swain C V K Prasada Rao 《Journal of Earth System Science》1993,102(2):351-366
Directional wave data collected during an experiment at a location on the continental shelf of the south west coast of India
using a WAVEC buoy, have been analysed based on the technique of Kuiket al (1988). The observed wave spectra indicate that the wave field is composed of sea waves (with peaks around 0·18 and 0·23
Hz) travelling nearly in the wind direction (WNW-N), and lower frequency (0·09 Hz) swell waves from the South. The parameterization
of the wave directional spread shows that both local wind conditions and nonlinear wave-wave interactions control the shape
of the directional distribution. The directional distribution is generally bimodal in the transition region between sea and
swell and at higher frequencies when rapid changes in wind speed and direction occur. 相似文献
8.
Storm surge induced by hurricane is a major threat to the Gulf Coasts of the United States. A numerical modeling study was conducted to simulate the storm surge during Hurricane Michael, a category 5 hurricane that landed on the Florida Panhandle in 2018. A high-resolution model mesh was used in the ADCIRC hydrodynamic model to simulate storm surge and tides during the hurricane. Two parametric wind models, Holland 1980 model and Holland 2010 model, have been evaluated for their effects on the accuracy of storm surge modeling by comparing simulated and observed maximum water levels along the coast. The wind model parameters are determined by observed hurricane wind and pressure data. Results indicate that both Holland 1980 and Holland 2010 wind models produce reasonable accuracy in predicting maximum water level in Mexico Beach, with errors between 1 and 3.7%. Comparing to the observed peak water level of 4.74 m in Mexico Beach, Holland 1980 wind model with radius of 64-knot wind speed for parameter estimation results in the lowest error of 1%. For a given wind model, the wind profiles are also affected by the wind data used for parameter estimation. Away from hurricane eye wall, using radius of 64-knot wind speed for parameter estimation generally produces weaker wind than those using radius of 34-knot wind speed for parameter estimation. Comparing model simulated storm tides with 17 water marks observed along the coast, Holland 2010 wind model using radius of 34-knot wind speed for parameter estimation leads to the minimum mean absolute error. The results will provide a good reference for researchers to improve storm surge modeling. The validated model can be used to support coastal hazard mitigation planning.
相似文献9.
A. D. Rao P. L. N. Murty Indu Jain R. S. Kankara S. K. Dube T. S. Murty 《Natural Hazards》2013,66(3):1431-1441
The devastation due to storm surge flooding caused by extreme wind waves generated by the cyclones is a severe apprehension along the coastal regions of India. In order to coexist with nature’s destructive forces in any vulnerable coastal areas, numerical ocean models are considered today as an essential tool to predict the sea level rise and associated inland extent of flooding that could be generated by a cyclonic storm crossing any coastal stretch. For this purpose, the advanced 2D depth-integrated (ADCIRC-2DDI) circulation model based on finite-element formulation is configured for the simulation of surges and water levels along the east coast of India. The model is integrated using wind stress forcing, representative of 1989, 1996, and 2000 cyclones, which crossed different parts of the east coast of India. Using the long-term inventory of cyclone database, synthesized tracks are deduced for vulnerable coastal districts of Tamil Nadu. Return periods are also computed for the intensity and frequency of cyclones for each coastal district. Considering the importance of Kalpakkam region, extreme water levels are computed based on a 50-year return period data, for the generation of storm surges, induced water levels, and extent of inland inundation. Based on experimental evidence, it is advocated that this region could be inundated/affected by a storm with a threshold pressure drop of 66 hpa. Also it is noticed that the horizontal extent of inland inundation ranges between 1 and 1.5 km associated with the peak surge. Another severe cyclonic storm in Tamil Nadu (November 2000 cyclone), which made landfall approximately 20 km south of Cuddalore, has been chosen to simulate surges and water levels. Two severe cyclonic storms that hit Andhra coast during 1989 and 1996, which made landfall near Kavali and Kakinada, respectively, are also considered and computed run-up heights and associated water levels. The simulations exhibit a good agreement with available observations from the different sources on storm surges and associated inundation caused by these respective storms. It is believed that this study would help the coastal authorities to develop a short- and long-term disaster management, mitigation plan, and emergency response in the event of storm surge flooding. 相似文献
10.
Auto-correlation analysis of ocean surface wind vectors 总被引：1，自引：0，他引：1
Abhijit Sarkar Sujit Basu A. K. Varma Jignesh Kshatriya 《Journal of Earth System Science》2002,111(3):297-303
The nature of the inherent temporal variability of surface winds is analyzed by comparison of winds obtained through different
measurement methods. In this work, an auto-correlation analysis of a time series data of surface winds measuredin situ by a deep water buoy in the Indian Ocean has been carried out. Hourly time series data available for 240 hours in the month
of May, 1999 were subjected to an auto-correlation analysis. The analysis indicates an exponential fall of the autocorrelation
in the first few hours with a decorrelation time scale of about 6 hours. For a meaningful comparison between satellite derived
products andin situ data, satellite data acquired at different time intervals should be used with appropriate ‘weights’, rather than treating
the data as concurrent in time. This paper presents a scheme for temporal weighting using the auto-correlation analysis. These
temporal ‘weights’ can potentially improve the root mean square (rms) deviation between satellite andin situ measurements. A case study using the TRMM Microwave Imager (TMI) and Indian Ocean buoy wind speed data resulted in an improvement
of about 10%. 相似文献
11.
Ki-Young Heo Jeong-Wook Lee Kyung-Ja Ha Ki-Cheon Jun Kwang-Soon Park Jae-Il Kwon 《Natural Hazards》2009,51(1):151-162
High-quality informations on sea level pressure and sea surface wind stress are required to accurately predict storm surges
over the Korean Peninsula. The storm surge on 31 March 2007 at Yeonggwang, on the western coast, was an abrupt response to
mesocyclone development. In the present study, we attempted to obtain reliable surface winds and sea level pressures. Using
an optimal physical parameterization for wind conditions, MM5, WRF and COAMPS were used to simulate the atmospheric states
that accompanied the storm surge. The use of MM5, WRF and COAMPS simulations indicated the development of high winds in the
strong pressure gradient due to an anticyclone and a mesocyclone in the southern part of the western coast. The response to
this situation to the storm surge was sensitive. A low-level warm advection was examined as a possible causal mechanism for
the development of a mesocyclone in the generating storm surge. The low-level warm temperature advection was simulated using
the three models, but MM5 and WRF tended to underestimate the warm tongue and overestimate the wind speed. The WRF simulation
was closer to the observed data than the other simulations in terms of wind speed and the intensity of the mesocyclone. It
can be concluded that the magnitude of the storm surge at Yeonggwang was dependent, not only on the development of a mesocyclone
but on ocean effects as well. 相似文献
12.
Susanne M. Moskalski Christopher K. Sommerfield Kuo-Chuin Wong 《Estuaries and Coasts》2011,34(4):800-813
Influences of tides, freshwater discharge, and winds on water properties in the St. Jones River estuary (USA), a Delaware
National Estuarine Research Reserve, were investigated using multiyear records of sea level, salinity, and turbidity, supplemented
by a current profiler time series in 2007. Results demonstrate that instantaneous properties fluctuate with semidiurnal tides
and resonant overtides, whereas tidal mean variations are forced by seasonal freshwater inflow and offshore winds. Mean sea
level and salinity are highest in summer and vary with seasonal water temperature and rainfall, whereas sea level variability
and turbidity are highest in winter on account of storm effects. Salinity and discharge modeling suggest that much (43–65%)
of the freshwater resident in the estuary is derived from non-point sources below the head of tide. This diffuse freshwater
inflow produces a seaward surface slope and weak mean current, which temporarily reverses under the influence of storm–wind
setup within Delaware Bay. 相似文献
13.
The Environmental Fluid Dynamic Code, an estuarine and coastal ocean circulation model, is used to simulate the distribution
of the salinity plume in the vicinity of the mouth of the Cape Fear River Estuary, North Carolina. The individual and coupled
effects of the astronomical tides, river discharge, and atmospheric winds on the spatial and temporal distributions of coastal
water levels and the salinity plume were investigated. These modeled effects were compared with water level observations made
by the National Oceanic and Atmospheric Administration and salinity surveys conducted by the Coastal Ocean Research and Monitoring
Program. Model results and observations of salinity distributions and coastal water level showed good agreement. The simulations
indicate that strong winds tend to reduce the surface plume size and distort the bulge shape near the estuary mouth due to
enhanced wind-induced surface mixing. Under normal discharge conditions, tides, and light winds, the southward outwelling
plume veers west. Relatively moderate winds can mechanically reverse the flow direction of the plume. Under conditions of
weak to moderate winds the water column does not mix vertically to the bottom, while in strong wind cases the plume becomes
vertically well mixed. Under conditions of high river discharge the plume increases in size and reaches the bottom. Vertical
mixing induced by strong spring tides can also enable the plume to reach the bottom. 相似文献
14.
P. R. Shanas V. M. Aboobacker Khalid M. Zubier Alaa M. A. Albarakati 《Arabian Journal of Geosciences》2018,11(5):90
The Jeddah coast lies in the central eastern Red Sea, which is characterized by the predominant northwest winds and the associated wind waves throughout the year. A detailed investigation on the spectral wave characteristics in the nearshore regions of the Jeddah coast has not been carried out yet, primarily due to the lack of data. In the present study, we have analyzed the available wave spectra measured at two nearshore locations along the Jeddah coast using wave gauges. The wave spectra were separated into wind sea and swell components using a frequency-based algorithm, and the integral wave parameters corresponding to each component were derived. Although the measurements were limited to the summer season, notable features such as the diurnal variability and the superimposition of wind seas and swells have been identified from the spectra. The superimposition is mainly due to the interaction of the young swells propagated from the northern Red Sea and the local breezes prevailing along the coast at certain periods in a diurnal cycle. Based on the wave speed calculations and the estimated time shifts between the wind and waves, the potential swell regions have been backtracked. 相似文献
15.
Amy F. Waterhouse Bilge Tutak Arnoldo Valle-Levinson Y. Peter Sheng 《Estuaries and Coasts》2013,36(5):1037-1053
The mechanisms responsible for the modulation of laterally sheared non-tidal (residual) exchange flow in a subtropical inlet, with special emphasis on tropical storm influence, are studied using a combination of current velocity profiles and hydrographic and meteorological data. The mouth of the inlet, St. Augustine Inlet in northeast Florida, is characterized by a 15-m-deep channel flanked by shoals (<6 m deep). Residual flows across the inlet mouth were laterally sheared with inflow in the channel and outflow over the shoals. This pattern persisted during four separate semi-diurnal tidal cycle surveys effected over 3 years. During spring tides, residual exchange flows intensified relative to neap tides. Residual inflow in the channel only reversed immediately after tropical storms because of their extreme winds and major temporal changes in water level. After the residual flow reversed in the channel, along-channel baroclinicity drove gravitational circulation that persisted for 4.5 days and was enhanced by offshore winds. A depth-averaged along-basin momentum budget highlighted the importance of bottom friction to help balance the barotropic pressure gradient. The rest of the momentum budget was likely provided by advective terms. During and after tropical storms, accelerations from wind stress and baroclinic pressure gradients also became influential in the along-basin momentum budget. 相似文献
16.
Antony Joseph R. G. Prabhudesai Prakash Mehra V. Sanil Kumar K. V. Radhakrishnan Vijay Kumar K. Ashok Kumar Yogesh Agarwadekar U. G. Bhat Ryan Luis Pradhan Rivankar Blossom Viegas 《Natural Hazards》2011,57(2):293-312
Response of the coastal regions of eastern Arabian Sea (AS) and Kavaratti Island lagoon in the AS to the tropical cyclonic storm `Phyan??, which developed in winter in the south-eastern AS and swept northward along the eastern AS during 9?C12 November 2009 until its landfall at the northwest coast of India, is examined based on in situ and satellite-derived measurements. Wind was predominantly south/south-westerly and the maximum wind speed (U10) of ~16 m/s occurred at Kavaratti Island region followed by ~8 m/s at Dwarka (Gujarat) and ~7 m/s at Diu (located south of Dwarka) as well as two southwest Indian coastal locations (Mangalore and Malpe). All other west Indian coastal sites recorded maximum wind speed of ~5?C6 m/s. Gust factor (i.e., gust-to-speed ratio) during peak storm event was highly variable with respect to topography, with steep hilly stations (Karwar and Ratnagiri) and proximate thick and tall vegetation-rich site (Kochi) exhibiting large values (~6), whereas Island station (Kavaratti) exhibiting ~1 (indicating consistently steady wind). Rainfall in association with Phyan was temporally scattered, with the highest 24-h accumulated precipitation (~60 mm) at Karwar and ~45 mm at several other west Indian coastal sites. Impact of Phyan on the west Indian coastal regions was manifested in terms of intensified significant waves (~2.2 m at Karwar and Panaji), sea surface cooling (~5°C at Calicut), and moderate surge (~50 cm at Verem, Goa). The surface waves were south-westerly and the peak wave period (T p) shortened from ~10?C17 s to ~5?C10 s during Phyan, indicating their transition from the long-period `swell?? to the short-period `sea??. Reduction in the spread of the mean wave period (T z) from ~5?C10 s to a steady period of ~6 s was another manifestation of the influence of the cyclone on the surface wave field. Several factors such as (1) water piling-up at the coast supported by south/south-westerly wind and seaward flow of the excess water in the rivers due to heavy rains, (2) reduction of piling-up at the coast, supported by the upstream penetration of seawater into the rivers, and (3) possible interaction of upstream flow with river run-off, together resulted in the observed moderate surge at the west Indian coast. Despite the intense wind forcing, Kavaratti Island lagoon experienced insignificantly weak surge (~7 cm) because of lack of river influx and absence of a sufficiently large land boundary required for the generation and sustenance of wave/wind-driven water mass piling-up at the land?Csea interface. 相似文献
17.
Nicolas Bruneau Juergen Grieser Thomas Loridan Enrica Bellone Shree Khare 《Natural Hazards》2017,85(2):649-667
Catastrophe risk models are used to assess and manage the economic and societal impacts of natural perils such as tropical cyclones. Large ensembles of event simulations are required to generate useful model output. For example, to estimate the risk due to wind-driven storm surge and waves in tropical cyclone risk models, computationally efficient parametric representations of the wind forcing are required to enable the generation of large ensembles. This paper presents new results on the impact of including explicit representations of extra-tropical transitioning in parametric wind models used to force storm surge and wave simulations in a catastrophe risk modelling context. Extra-tropical transitioning is particularly important in modelling risk on the Japanese coastline, as roughly 40 % of typhoons hitting the Japanese mainland are transitioning before landfall. Using both a historical and idealized track set, we compare maximum storm surge and wave footprints along the Japanese coastline for models that include, and do not include, explicit representations of extra-tropical transitioning. We find that the inclusion of extra-tropical transitioning leads to lower storm surge (10–20 %) and waves (5–15 %) on the southern Japanese coast, with significantly higher storm surge and waves along the northern coast (25–50 %). The results of this paper demonstrate that useful risk assessment of coastal flood risk in Japan must consider the extra-tropical transitioning process. 相似文献
18.
Jean-Michel Lefèvre 《Natural Hazards》2009,49(2):361-370
Long and high swells are dangerous for many islands located in the Tropics because they can generate large breakers and long
run up associated with large set up when reaching the coast. Most of the time those islands do not experience large waves
especially in usually protected areas, for instance, by coral reefs or wind protected. Long waves have the ability to reach
such areas, thanks to wave set up, shoaling and bottom refraction. This article describes an example of such high swell events
and its impact on the islands. The buoy network used by the French National Weather Service and all available satellite observations
related to waves are presented together with numerical sea-state models used to issue early warnings. 相似文献
19.
During 23–30 September 1997, a rare cyclonic storm has developed close to the Andhra coast, and it has later travelled parallel to coastline northward and finally crossed the land at Chittagong (22°N, 91°E) on 27 September. While translating along the east coast of India, it has produced heavy to very heavy rainfall on the coastal stations causing devastating floods. In this study, we made an attempt to understand the salient causes of this unique cyclone movement. We have analyzed daily fields of wind and relative humidity for 850, 700, 500 hPa and mean daily OLR data to understand the plausible reasons for its movement. The buoy data deployed by National Institute of Ocean Technology, Chennai, Viz. DS5 (15°N, 81°E), DS4 (19°N, 88°E) and SW7 (20°N, 86°E) were analyzed to understand the ocean–atmosphere interaction processes in the west Bay of Bengal during formation of the system. Analysis of OLR over the cyclonic storm region has revealed that the heavy rainfall areas coincide with low OLR (120–180 W m?2). The persistent southward movement of 500 hPa ridge on the eastern wedge of the system along with the steering current at 200 hPa has helped in maintaining the movement of the system parallel to the east coast of India during its life cycle. 相似文献
20.
A bottom-mounted Recording Doppler Current Profiler was placed at an offshore location (depth of 34 m) in the southeast Chukchi Sea, Alaska, from July through December 2007 (UTC) with the objective of linking observed wave activity—wind-sea and swells—to their synoptic drivers. A total of 47 intervals of elevated wave state were recorded: 29 exceeding 1 m significant wave height (SWH), 16 exceeding 2 m SWH, and 3 m exceeded on two occasions; during one of those, a SWH of 4 m was observed. Detailed analysis of the two large events, including comparison with high-resolution reanalysis wind data (North America Regional Reanalysis), showed wave direction from the east, varied about 15° to the north (counterclockwise) from the wind direction, and current flow in the opposite direction (from the west). This is thought to be the influence of a strong “wind-sea” presence. Regarding classic wave limitations, although the SE Chukchi Sea is a large embayment bordered by land to the east, fetch limitations from the northeast and southeast did not appear to be a constraint for the wind speeds indicated by reanalysis. These two events appeared to be driven by winds associated with cyclonic systems that moved into the eastern Bering Sea and stalled. Examination of smaller waves associated with these events suggested that waves of 1.5 m SWH or less are likely part of another regime and can either be swell or wind-sea, moving in from the open Chukchi Sea to the northwest or through the Bering Strait to the south. 相似文献