Abstract

Characteristics of ocean surface winds around an isolated island are examined in relation to atmospheric stability using a synthetic aperture radar (SAR) and rawinsonde sounding observations. The SAR-derived winds on 22 May 2009 indicate a low-level jet extending over 30 km behind the island. Around the time of SAR acquisition, winds intensified on the leeward side in association with the stabilization of stratified flows, which suggests the connection of the SAR-derived jet with downslope winds. A number of SAR-derived winds elucidate typical wind patterns and their transitions depending on the nondimensional mountain height . For cases of large (>2), a wake is formed in the lee of the island and low-level jets produce strong wind shear on both sides of the wake. For cases of relatively small (<1.75), although a weak wind region is formed behind the mountain, no wind jets develop. As a transition of the above two cases , a low-level jet develops in the lee of the island, as in the case on 22 May 2009. These wake configurations and their -dependent transitions seem to correspond to major regimes for hydrostatic flow over topography with constant upstream speed and buoyancy frequency: small-amplitude waves , wave breaking , and flow splitting . It is noted that the ocean surface winds behind the island are very sensitive to around the transition point, changing up and down. The occurrence of each wind pattern shows clear seasonal features dependent on atmospheric stability.

1. Introduction

Atmospheric flows become intricately deformed by topographic forcing and their behavior directly affects human activities. Many theoretical (e.g., Drazin 1961; Lilly and Klemp 1979; Smith 1980, 1985), experimental (e.g., Brighton 1978; Hunt and Snyder 1980), and numerical (e.g., Smolarkiewicz and Rotunno 1989; Schär and Durran 1997; Epifanio and Rotunno 2005) studies on stratified flow over mountains have been conducted, mainly focusing on its effect on aviation and the spreading of polluted air. The behavior of stratified flow passing over an isolated mountain has been of great interest. For the isolated mountain characterized by an along-stream length scale a, a cross-stream length scale b, and a maximum height h, the hydrostatic flow with uniform upstream wind speed U and buoyancy frequency (or Brunt–Väisälä frequency) N is defined by the two following nondimensional parameters: 1) the nondimensional mountain height , which is the mountain height normalized by a scale for the wavelength of a linear two-dimensional mountain wave, and 2) the horizontal aspect ratio β = b/a (e.g., Epifanio 2003; Reinecke and Durran 2008). The parameter is also recognized as the inverse Froude number (Fr). For an axisymmetric mountain, the controlling parameter is reduced to only because of a constant β (= 1). Previous studies suggested that the flow showed nonlinear nature with the increase of and had the following four regimes: 1) small-amplitude waves, 2) wave breaking, 3) upstream stagnation and flow splitting, and 4) lee vortices (e.g., Bauer et al. 2000; Epifanio 2003).

For , a mountain-induced flow can be described as a small-amplitude mountain wave. Smith (1980) developed a linear theory of three-dimensional mountain waves for the flow over an isolated bell-shaped mountain. The theory gives an asymmetric surface pressure distribution situating high (low) pressure anomalies over the windward (leeward) slope of the hill. Continuous horizontal divergence in the lee of the hill, which is induced by pressure distribution, is partially attributed to the sinking of warm air aloft. The increase of surface wind speeds associated with the sinking air reaches its maximum in the leeward slope and is proportional to h and N. With the increase of , streamlines over the leeward slope gradually grow steeper due to nonlinearity.

For , a streamline overturns over the leeward slope resulting in wave breaking when exceeds a critical value (generally 0.7–1.2) dependent on the shapes of the topography. A nonlinear lee wave theory has been developed that incorporates wave breaking to explain the mechanism of leeward severe winds (e.g., Lilly and Klemp 1979; Smith 1985). Smith (1985) proposed a two-dimensional nonlinear theory of the flow over the mountain based on Long’s (1953) equation and pointed out the importance of a wave-induced well-mixed stagnant layer to the onset of downslope winds. The wave breaking forms a well-mixed region over the leeward slope, which results in an acceleration of the low-level lee-slope flow from the flow aloft. This is similar to the supercritical flow in shallow-water theory.

For , the flow splits and goes around the mountain, being confined to horizontal planes, or remains stagnate on the windward side of the mountain. The upstream splitting reduces the wave breaking. As increases, a counterrotating vortex pair with reversed flow appears behind the mountain. The theory for large flow was developed by Drazin (1961), where the flows confined to each horizontal plane are given by two-dimensional potential solutions on their respective planes. Brighton (1978) conducted laboratory experiments of very stable stratified flows and confirmed the validity of Drazin’s (1961) theory except at the layer around the top of the obstacle and at the wake where the horizontal flow regime breaks down. Many numerical studies have been performed to investigate the configurations of the large flow, where mechanisms contributing to lee vortex formation and shedding were the focus (e.g., Smolarkiewicz and Rotunno 1989; Schär and Durran 1997; Epifanio and Rotunno 2005). They pointed out that the lee vortex formation is attributed to either wave breaking or upstream blocking depending on . For the large flow, instability of the symmetric wake associated with the intensive reversed flow causes the transition into a vortex-shedding regime. Recently, the effects of other controlling parameters, such as the horizontal aspect ratio β (e.g., Bauer et al. 2000), the vertical wind shear (e.g., Miranda and Valente 1997), the earth’s rotation (e.g., Thorsteinsson 1988), and the vertical gradient of atmospheric stability (Reinecke and Durran 2008), have also been examined.

Some observational studies on the flow over an isolated mountain using intensive actual atmospheric measurements have been conducted for relatively large mountains [O(100 km)] (e.g., Walter and Overland 1982; Mass and Ferber 1990; Ferber and Mass 1990; Smith and Grubišić 1993). Walter and Overland (1982) investigated the relationship between wind measurements in the lee of the Olympic Mountains in the state of Washington and the atmospheric conditions. They demonstrated that flows were principally horizontally confined for large (low Fr) cases, which is consistent with Drazin’s (1961) theory and the laboratory experiments of Brighton (1978) and Hunt and Snyder (1980), and an asymmetric pressure distribution for the low (= 0.22) (large Fr) case, which is approximately consistent with Smith’s (1980) theory. Smith and Grubišić (1993) elucidated from aircraft measurements that a quasi-steady counterrotating eddy pair was formed behind the island of Hawaii under the large (=3) trade winds. Xie et al. (2001) demonstrated through the use of scatterometer wind data that a stationary alternate wind pattern with strong and weak winds is produced in the lee and on the flanks of the major Hawaiian Islands under the trade winds and that it develops a steady ocean current. Satellite images also contributed to our understanding of an island–atmosphere interaction (e.g., Etling 1990; Smith et al. 1997). A Kármán-type vortex shedding from an isolated island, which had been confirmed in satellite cloud images, was investigated in relation to the Rossby and Froude numbers (Etling 1990). The sunglint images evinced long, straight 20-km-wide wakes extending over 300 km (Smith et al. 1997). They suggested that the wakes were closely related to the generation of potential vorticity and were induced by wave breaking (stratification-induced dissipation), as pointed out by Schär and Smith (1993a).

The synthetic aperture radar (SAR), which is a microwave imaging sensor, can provide high spatial resolution images (~10 m) of ocean surface roughness and the related wind fields even in the coastal seas where scatterometers cannot observe. The SAR in some cases has shed light on previously unknown phenomena, especially about coastal wind fields. There are numerous works of research dealing with the orographically forced coastal wind fields such as katabatic winds, bora winds, gap winds, and barrier jets (e.g., Alpers et al. 1998, 2009; Pan and Smith 1999; Sandvik and Furevik 2002; Loescher et al. 2006; Colle et al. 2006; Winstead et al. 2006). Alpers et al. (1998, 2009) investigated katabatic wind fields over Mediterranean coastal waters and bora events over the Adriatic and Black Seas using SAR-derived wind patterns. Sandvik and Furevik (2002) simulated a mesoscale coastal jet confirmed through in situ ship measurements and SAR images using a high-resolution numerical model. They found that the coastal jets were the result of stratified flows around mountains. Loescher et al. (2006), Colle et al. (2006), and Winstead et al. (2006) advanced the research of SAR-derived winds from case studies to climatological ones. They investigated the temporal and spatial climatologies of barrier jets in the Gulf of Alaska using a number of SAR-derived wind maps and clarified the frequency, spatial distribution, dynamics, and morphology of these phenomena.

There are some studies dealing with the effects of SAR-revealed gap winds on wind-wave development (Shimada and Kawamura 2004, 2006; Isoguchi and Kawamura 2007; Shimada et al. 2008). Through the combined use of a scatterometer, altimeter, and SAR-derived winds, Shimada and Kawamura (2004, 2006) demonstrated that the SAR winds could fill coastal wind fields that connect those on land and those on the open ocean, and these orographically modified winds affected wind-wave development in the downstream region. Isoguchi and Kawamura (2007) and Shimada et al. (2008) clarified the picture and a formative mechanism of coastal wind jets and their effects on wind-wave development using satellite observations and numerical simulations.

In the present study, characteristics of ocean surface wind fields around Rishiri Island are examined in relation to atmospheric conditions using SAR images and in situ aerological and ground measurements. Rishiri Island, located west of Hokkaido, Japan, is an almost cone-shaped isolated island with a diameter of about 16 km and a height of 1719 m (Fig. 1). Thus, a major flow-determining parameter is only under the assumption of the hydrostatic flow with constant U and N. A series of SAR images elucidates characteristic island wakes with weak and strong surface flows and the effect of on their configurations is analyzed. The remainder of this paper is constructed as follows. In section 2, the satellite and in situ data used and the derivation of wind fields by the SAR are described. In section 3, a case study on a low-level jet observed in the lee of Rishiri Island is presented. In section 4, with a number of SAR images, the spatial and temporal climatologies of the wind fields are investigated. The wind maps are morphologically classified into four types and their dependence on is examined. The occurrence of each type indicates clear seasonal characteristics corresponding to atmospheric conditions. A discussion is provided in section 5. Section 6 gives a summary and our conclusions.

Fig. 1.

Maps and topography of (a) northern Japan and (b) the area around Rishiri Island. The area in (b) is indicated by the dashed rectangular in (a). (c) Elevation profiles of Rishiri Island along meridional (solid) and zonal (dashed) lines that cross the peak. The x axis shows distance from the peak.

Fig. 1.

Maps and topography of (a) northern Japan and (b) the area around Rishiri Island. The area in (b) is indicated by the dashed rectangular in (a). (c) Elevation profiles of Rishiri Island along meridional (solid) and zonal (dashed) lines that cross the peak. The x axis shows distance from the peak.

2. Data

The SAR images used in this study are collected from the Phased-Array L-band Synthetic Aperture Radar (PALSAR) aboard the Advanced Land Observing Satellite (ALOS) and the Active Microwave Instruments (AMI) aboard the Earth Resources Satellite-½ (ERS-½). The AMI operates in C band (5.3 GHz) with vertical polarization and a constant incidence angle of 23°. It obtains strips of high-resolution (about 30 m) images of a 100-km width. The PALSAR, operated in L band (1.27 GHz), has variable observational modes, where the off-nadir angles and polarization are changeable. The image we used was observed over a width of 70 km with horizontal polarization and an off-nadir angle of 34.3°. Acquisition times and sensors of the SAR images used are listed in the  appendix (Table A1). A SIGMA-SAR processor (Shimada 1999) performs imaging processes from raw signal data and each digital number in the SAR images is converted to a normalized radar cross section (NRCS) by applying conversion parameters for AMI (Shimada 2002) and PALSAR (Shimada et al. 2009).

To obtain the wind speed from an NRCS value, the C-band model function (CMOD4; Stoffelen and Anderson 1997) and the L-band horizontal transmit–horizontal receive (HH) geophysical model function (GMF; Isoguchi and Shimada 2009) are applied to the AMI and PALSAR images, respectively. Unlike the well-established C-band vertical transmit–vertical receive (VV) GMF (CMOD4), which was originally developed for scatterometers, the L-band HH GMF we used was empirically developed for PALSAR by relating the PALSAR NRCSs to the collocated wind vectors obtained from scatterometers. This type of wind conversion from an SAR had been proposed and applied to previous SARs: the RADARSAT-1 SAR operating at C-band HH polarization (Horstmann et al. 2000; Monaldo et al. 2004) and the Japanese Earth Resources Satellite-1 (JERS-1) SAR operating at L-band HH polarization (Shimada et al. 2003).

To estimate wind speed, an incidence angle, a beam view angle, a wind direction, and an NRCS value are required. The first two parameters are taken from satellite orbital information. For PALSAR data, the wind direction is estimated by interpolating the model surface winds, which is a well-established method for SAR wind speed mapping (i.e., Monaldo et al. 2001; Monaldo et al. 2004). We used the Grid-Point Value-Mesoscale Model (GPV-MSM) data produced by the Japan Meteorological Agency (JMA), which are calculated at 3-h intervals onto a 0.05° × 0.0625° grid around Japan. For the AMI data, because the GPV-MSM does not cover the entire AMI observation period from 1991 to 2003, the wind direction was estimated from the image itself as the direction of an island wake and was defined as a constant over the whole image (100 km × 85 km). As mentioned above, because a reversed flow is expected to appear in a wake for a large flow, the assumption of the constant wind direction might give rise to errors in the wind speed estimation. We estimate the following errors for typical values: a 180° (90°) inaccurate direction causes a 0.5 m s−1 (5 m s−1) error for the AMI (C band) data in the case of the incidence angle of 23° and the NRCS of −4 dB in the wake, and a 1.0 m s−1 (1.5 m s−1) error for the PALSAR (L band) data in the case of an incidence angle of 35° and an NRCS of −17 dB. Thus, whereas relatively large errors are expected for wind components perpendicular to the defined constant wind direction, those for the reversed flow are not as significant. In addition, since qualitative wind patterns are mainly discussed for the data with the weak wind wake, the errors do not affect the main results in the present study.

To capture wind fields over the land and ocean at and around the SAR observation times, Automated Meteorological Data Acquisition System (AMeDAS) data are used. AMeDAS, a land-based regional meteorological observation system operated by JMA, automatically obtains precipitation amounts, wind direction, wind speed, temperature, and sunshine duration every 10 min. Aerological observations of daily rawinsondes and rawin ascents provided by JMA are used to investigate atmospheric stability conditions. The rawinsonde soundings record the atmospheric pressure, air temperature, humidity, wind directions, and wind speeds at altitudes up to about 30 km. We mainly use the measurements at Wakkanai, Japan, located 35 km east-northeast of Rishiri Island (see Fig. 1), assuming that they are representative of atmospheric conditions around Rishiri Island. In addition, the measurements at Sapporo, Japan (see Fig. 1), are added for some cases to capture atmospheric conditions more widely. From the data, N and are calculated for the respective SAR observations (Table A1) as

 
formula

where g is the acceleration of gravity, h0 is the mountain height defined as 1700 m, θ is the potential temperature, and U is the wind speed defined as an average for the 0–1700-m layer.

3. A case study on a low-level jet formed in the lee of Rishiri Island

The PALSAR image acquired at 0113 UTC 22 May 2009 represents a characteristic wind pattern in the lee of Rishiri Island. Figure 2 shows the PALSAR-derived wind speeds, on which the GPV-MSM 10-m wind vectors at 0000 UTC are overlaid with black arrows, and the AMeDAS surface wind speeds at 0110 UTC with white arrows. For south-southwesterly (SSW) ambient winds of about 7 m s−1, a low-level jet developed in the lee of Rishiri Island, extending north-northeastward over 30 km. The AMeDAS observations have been conducted since 2003 at two stations, Kutsugata and Motodomari, located west and north of the island (see Fig. 1), which enable us to investigate the wind field affected by topography. A relatively weak southerly wind of 4 m s−1 is observed at the western tip of the island (Kutsugata), while a stronger SSW one of 11.6 m s−1 is observed in the north of the island (Motodomari). The in situ wind distribution is thus qualitatively consistent with that derived from the SAR observations.

Fig. 2.

The SAR-derived wind speed map around Rishiri Island at 0113 UTC 22 May 2009 (shaded image), along with wind vectors over the land and ocean, which were obtained from AMeDAS at 0110 UTC (white arrows) and GPV-MSM at 0000 UTC (black arrows).

Fig. 2.

The SAR-derived wind speed map around Rishiri Island at 0113 UTC 22 May 2009 (shaded image), along with wind vectors over the land and ocean, which were obtained from AMeDAS at 0110 UTC (white arrows) and GPV-MSM at 0000 UTC (black arrows).

In general, atmospheric stability and wind speeds are of importance when considering downslope winds. Vertical profiles of temperature, potential temperature, and wind observed by the rawinsonde sounding at Sapporo and Wakkanai (see Fig. 1) at 0000 UTC 22 May 2009 are depicted in Fig. 3. Also shown in Fig. 4 are the rawinsonde-derived time–height sections of the potential temperature and wind at (a) Sapporo and (b) Wakkanai, (c) the AMeDAS hourly wind speed and direction, and (d) the temperatures at the Kutsugata and Motodomari stations (see Fig. 1) from 19 to 24 May. The PALSAR acquisition time is shown with solid triangles in Fig. 4. Both the temperature profiles T at Wakkanai and Sapporo show temperature inversions (Fig. 3), suggesting stably stratified flows in the lower layer. In particular, the profile at Wakkanai features the strong temperature inversion at a height around 200–300 m with warmer air of about 291 K (18°C) at the 300–700-m layer. The temperature at the Motodomari station was 18.5°C at 0110 UTC, implying that the warmer air above the inversion descends to Motodomari. The time progression of this phenomenon is illustrated in Fig. 4. Relatively warmer air appearing from 20 May made the potential temperature gradient strong at the lower layer. Along with it, the wind speed at Motodomari increased rapidly by 12 m s−1 from 2 m s−1 at 1800 UTC 20 May and accelerated up to 15 m s−1 at 0600 UTC 21 May. On the other hand, the wind speed at Kutsugata did not undergo a large change, remaining around 5 m s−1. This condition remained unchanged for about 30 h while the southwesterly wind blew, until 0300 UTC 22 May, 2 h after the PALSAR observation. The average wind speeds during the period were 12.8 m s−1 at Motodomari and 4.9 m s−1 at Kutsugata. The advection of the warmer air seems to have stopped at around 0300 UTC 22 May at the time when the wind direction changed to the northeast, which relaxed the atmospheric stability in the atmospheric boundary layer. Along with this change, the northeasterly wind speed at Motodomari decreased rapidly and underwent a transition similar to that at Kutsugata. It is thus inferred that the SAR-revealed low-level wind jet in the lee of the island represents evidence of downslope winds over the ocean.

Fig. 3.

Vertical profiles of temperature (T, black line), potential temperature (θ, gray line), and wind vectors (arrows), which were obtained from rawinsonde soundings at (a) Sapporo and (b) Wakkanai (see Fig. 1) at 0000 UTC 22 May 2009.

Fig. 3.

Vertical profiles of temperature (T, black line), potential temperature (θ, gray line), and wind vectors (arrows), which were obtained from rawinsonde soundings at (a) Sapporo and (b) Wakkanai (see Fig. 1) at 0000 UTC 22 May 2009.

Fig. 4.

Time series of potential temperature (shaded contours) and wind vector profiles at (a) Sapporo and (b) Wakkanai. Also shown are (c) wind speed and direction and (d) temperature at Kutsugata (gray line and arrows) and Motodomari (black line and arrows). The time of the SAR observation (Fig. 2) is indicated by solid triangles.

Fig. 4.

Time series of potential temperature (shaded contours) and wind vector profiles at (a) Sapporo and (b) Wakkanai. Also shown are (c) wind speed and direction and (d) temperature at Kutsugata (gray line and arrows) and Motodomari (black line and arrows). The time of the SAR observation (Fig. 2) is indicated by solid triangles.

It should be noted that although we suggested the contribution of stratification-induced wave breaking to the downslope winds, there is another possible process to consider: mountain wave breaking and dissipation at a critical level (e.g., Durran and Klemp 1987; Durran 2003). In fact, the vertical wind profiles at Wakkanai (Fig. 4b) indicate critical levels at around 2 km during the intense winds at Motodomari (from 0000 UTC 21 May to 0000 UTC 22 May), suggesting their contribution to the low-level jet. To clarify an actual process, numerical simulations activated by actual atmospheric measurements are needed.

An episode of similar intense downslope winds has been reported upon in the literature (Fujita and Onoda 1984). In May 1983, a research vessel was anchored in Oshidomari Port, located in the north of the island (see Fig. 1), to escape the strong southwesterly winds. Nevertheless, it suffered continuously from the strong southwesterly winds for the subsequent 18 h with maximum hourly winds of 26 m s−1. The rawinsonde soundings at Wakkanai indicated the stably stratified south-southwesterly with a temperature inversion in the 500–600-m layer. Thus, the SAR-derived wind field on 22 May 2009 (Fig. 2) might be a typical example of the downslope winds behind Rishiri Island.

4. Statistical characteristics of atmospheric island wakes

a. Ocean surface wind patterns dependent on the nondimensional mountain height

In the previous section, we showed, as a case study, that a wind jet developed in the lee of Rishiri Island for the stably stratified flow. To investigate the effects of atmospheric stratification on surface wind patterns, we used the 115 ERS-½ AMI data images taken from August 1991 through February 2003. Although the data acquisition frequency represents year-to-year differences due to the change of the satellite repetition period, we used all images covering most or all of Rishiri Island. The observation times were 0120–0126 UTC (1246–1247 UTC) for descending (ascending) observations. The rawinsonde soundings and AMeDAS observations close to the SAR acquisition were also used. Unfortunately, because only one AMeDAS station (Kutsugata) was available during the period, we cannot analyze the wind distributions on the island as we did in the previous case study.

The images are classified into the following four classes in terms of atmospheric island wakes: type A, an island’s effect on surface winds is hardly seen (17%); type B, a weak wind area (wind shadow) is formed in the lee of the island, while a low-level jet is developed on both sides of the wind shadow (27%), with the one on the left being especially strong; type C, a wind jet is formed in the lee, as in the PALSAR data on 22 May 2009, in Fig. 2 (19%); and type D, a weak wind area (wind shadow) is formed in the lee, but no jets develop (36%). In the present study, a homogeneous region with an average wind speed of 2 m s−1 larger (smaller) than the ambient one was defined as the wind jet (wind shadow). Figures 58 represent typical examples of each type of island wake. The AMeDAS hourly mean winds close to the SAR observation are overlaid with the SAR-derived wind fields. In addition, the temperature, potential temperature, and wind profiles observed at 0000 (1200) UTC for the descending (ascending) data at Wakkanai are depicted.

Fig. 5.

(left) The examples of type-A SAR-derived wind speed maps around Rishiri Island at (a) 0125 UTC 4 Sep 1997, (b) 0124 UTC 16 Jul 1998, (c) 0124 UTC 31 May 2001, and (d) 0123 UTC 31 Jan 2002, on which the AMeDAS wind vectors are overlaid with black arrows, along with (right) vertical profiles of temperature (T, black line), potential temperature (θ, gray line), and wind vectors (arrows) observed at Wakkanai at 0000 UTC.

Fig. 5.

(left) The examples of type-A SAR-derived wind speed maps around Rishiri Island at (a) 0125 UTC 4 Sep 1997, (b) 0124 UTC 16 Jul 1998, (c) 0124 UTC 31 May 2001, and (d) 0123 UTC 31 Jan 2002, on which the AMeDAS wind vectors are overlaid with black arrows, along with (right) vertical profiles of temperature (T, black line), potential temperature (θ, gray line), and wind vectors (arrows) observed at Wakkanai at 0000 UTC.

Fig. 6.

As in Fig. 5, but for type-B winds at (a) 0121 UTC 5 May 1993, (b) 0125 UTC 26 Jul 1995, (c) 0125 UTC 8 Jun 1992, and (d) 0125 UTC 7 May 1998.

Fig. 6.

As in Fig. 5, but for type-B winds at (a) 0121 UTC 5 May 1993, (b) 0125 UTC 26 Jul 1995, (c) 0125 UTC 8 Jun 1992, and (d) 0125 UTC 7 May 1998.

Fig. 7.

As in Fig. 5, but for type-C winds at (a) 0122 UTC 28 Apr 1995, (b) 0122 UTC 27 Feb 1999, (c) 0122 UTC 12 Jun 1999, and (d) 0125 UTC 28 Sep 2000.

Fig. 7.

As in Fig. 5, but for type-C winds at (a) 0122 UTC 28 Apr 1995, (b) 0122 UTC 27 Feb 1999, (c) 0122 UTC 12 Jun 1999, and (d) 0125 UTC 28 Sep 2000.

Fig. 8.

As in Fig. 5, but for type-D winds at (a) 0122 UTC 27 Oct 1993, (b) 0124 UTC 14 Oct 1999, (c) 0122 UTC 8 Jan 2000, and (d) 0125 UTC 6 Apr 2000.

Fig. 8.

As in Fig. 5, but for type-D winds at (a) 0122 UTC 27 Oct 1993, (b) 0124 UTC 14 Oct 1999, (c) 0122 UTC 8 Jan 2000, and (d) 0125 UTC 6 Apr 2000.

Figures 6c and 6d of type B indicate gap winds developed over the Soya Strait between Hokkaido and Sakhalin (see Fig. 1a) under easterly wind conditions. Whereas wind shadows are formed in the lee (west) of Rishiri Island, wind jets are formed southwestward, just to the south of the shadow, generating strong wind shear. The wind fields around Rebun Island (see Fig. 1b) show a more detailed pattern depending on the topography. The strong winds are formed in the lee of Rebun Island, within which the weak wind streaks appear just behind the relatively high topography. The temperature profiles feature strong inversions at around the 500–1000-m layer, which seem to cap the lower layers. It is thus suggested that the stably stratified easterly flow goes around the mountain and generates wind jets and shadows in the lee of the island. Figures 6a and 6b show similar wakes under the southwesterly ambient wind condition. Although wind shadows are seen in the northeast of the island, the wind jets are just to the left of the shadow, extending northeastward from the northern tip of the island. Other wind jets extend from the northern tips of Wakkanai and Rebun Island as well, which also suggests strong topographical forcing. In fact, the flows are stably stratified at around the 200–700-m layer in the cases of Figs. 6a and 6b, as indicated by the potential temperature profiles.

A type C wake (Fig. 7) represents low-level wind jets in the lee of Rishiri Island, as in the case on 22 May 2009 (Fig. 2). Note that although the wind speed map in Fig. 2 is indeed connected with the wind acceleration on the leeward slope, as shown in section 3, it also indicates weak wind areas just to the right of the low-level jet and on the opposite windward side of the island. In addition, the bow-shaped configuration of the jet suggests reversed flows in the leeward wake. These wake configurations characterize a type-B wake, as seen in Figs. 6a and 6b, implying that the case shown in Fig. 2 corresponds to a transition between types B and C. Thus, although it was somewhat difficult to distinguish between types B and C, images with strong wind areas along the leeward direction from the summit of Rishiri Island were classified as type C in the present study. The temperature profiles of type C show inversions at around the 200–800-m layers on 28 April 1995 and 12 June 1999, (Figs. 7a and 7c) like those of type B. Although inversions are not present in the other cases (Figs. 7b and 7d), the potential temperatures increase with height, indicating stably stratified flow conditions. In the SAR images corresponding to the temperature inversions (Figs. 6a, 7a, and 7c), a number of finescale wind jets also appear off the eastern coast of Hokkaido, aligned parallel to each other. Their alignment corresponds to the upstream topography undulation, which corroborates the stably stratified flow condition.

The type-D images (Fig. 8) depict weak wind wakes extending over 30 km in the lee of the island, opposite to the type-C pattern. The temperature profiles indicate nearly neutral stratifications at least up to the mountain height, which are different from those of types B and C. A mottled pattern seen in the SAR images is another indicator of the atmospheric conditions. This pattern has been reported to be the signature of atmospheric convective cells (Mitnik 1992; Alpers et al. 1998), which occur due to heating from the ocean if the atmospheric temperature is cooler than the underlying sea surface temperature (SST). Assuming the temperatures at the lowest layer of the soundings are representative of the air temperatures around Rishiri Island, these values of in (a) 8.4°, (b) 7.9°, (c) −6.6°, and (d) 0.6° are indeed cooler than the daily mean SSTs of (a) 14.3°, (b) 17.1°, (c) 7.1°, and (d) 4.5°, which are derived from the satellite-based global daily 0.1°-grid SST data (Kawai et al. 2006). Thus, they meet the necessary conditions for convective cell development. It is thus suggested by the SAR signal patterns that the type-D atmospheric boundary layers are well mixed and different from those of types B and C.

We examine statistically the relationship between each wake type and atmospheric condition. Figure 9a depicts a scatter diagram for the 115 SAR observations as functions of Nh and U, where the four types are shown with difference symbols The contours of are overlaid with dashed lines. Figure 9b represents histograms of each type, as a function of . As h is constant in this case, the x axis of Fig. 9a (Nh) is proportional to N. The data whose potential temperature gradient is less than 0 are plotted as Nh = 0 in Fig. 9a and those with larger than 5.0 are counted in the 4.75–5.0 range in Fig. 9b. The type of each SAR observation is listed in the  appendix (Table A1) along with the corresponding meteorological measurements, such as U, N, and .

Fig. 9.

(a) Inertia (U, horizontal wind speed) buoyancy (Nh, the Brunt–Väisälä frequency multiplied by mountain height) diagram for the SAR-indentified ocean surface wind patterns (types A–D) around Rishiri Island. Type A is plotted with black circles, type B with green triangles, type C with red triangles, and type D with blue squares. Dashed lines indicate contours of the nondimensional mountain height, . (b) Histograms of occurrence frequency for each type as a function of . Gray (black) parts indicate those for upstream (downstream) cases (mean wind directions clockwise from northwest through southeast).

Fig. 9.

(a) Inertia (U, horizontal wind speed) buoyancy (Nh, the Brunt–Väisälä frequency multiplied by mountain height) diagram for the SAR-indentified ocean surface wind patterns (types A–D) around Rishiri Island. Type A is plotted with black circles, type B with green triangles, type C with red triangles, and type D with blue squares. Dashed lines indicate contours of the nondimensional mountain height, . (b) Histograms of occurrence frequency for each type as a function of . Gray (black) parts indicate those for upstream (downstream) cases (mean wind directions clockwise from northwest through southeast).

Type-A wakes are found in the wide range of , having weak wind speeds, as seen in Fig. 9a. This tendency is more significant in the lower layer. For type A, 85% of the wind speeds averaged over the 0–600-m layer are weaker than 6 m s−1. Their average speed of 4.0 ± 1.7 m s−1 is lower than those of other types: 7.7 ± 2.9 m s−1 for type B, 10.6 ± 3.1 m s−1 for type C, and 9.3 ± 3.5 m s−1 for type D. Note that the error bounds denote standard deviations. In general, SAR signals from oceanic phenomena, such as currents, slicks, and internal waves, become relatively stronger than those from atmospheric ones under weak wind conditions, which might be a reason for the lack of wake signals in type-A images (Fig. 5). Type-B wakes are confined on larger than 2.0. As inferred from Fig. 9a, this is mainly attributed to stable stratification conditions (large Nh), which frequently include a temperature inversion. In fact, the average of the potential temperature gradients of type B is 6.1 ± 2.9 K km−1, which is larger than those of other types: 3.8 ± 2.0K km−1 for type A, 4.6 ± 2.3 K km−1 for type C, and 2.5 ± 1.3 K km−1 for type D. Furthermore, 18 (51%) of the total 35 soundings that indicate a temperature inversion in the 1000–900-hPa layer belong to type B, which is larger than those of the other types: 9 (26%) of type A, 6 (17%) of type C, and 2 (6%) of type D. The average of the wind speeds of the 18 cases is 7.8 ± 2.7 m s−1. Whereas that of the nine type-A cases is weak (4.5 ± 1.9 m s−1), those of the six type-C and two type-D cases are strong (12.3 ± 3.4 and 11.6 ± 2.1 m s−1, respectively). These wind speeds reasonably explain the difference of the wake patterns. Type D tends to be distributed in smaller than 1.75 with its mode of 1.0–1.25. As mentioned previously, the atmospheric stratifications of type D are not stable (the average potential temperature gradient is 2.5 ± 1.3 K km−1), which contributes mainly to the relatively smaller . Type C is roughly plotted between the clusters of types B and D (Fig. 9a), resulting in the mode of being in the 1.75–2.0 range. This is reasonable because the stratifications of type C are not so stable compared with those of type B, whereas the wind speeds of type C are, on the whole, stronger than those of type B.

A well-known wake pattern, in addition to the present four types, is the Kármán-type vortex streets that form in the lee of an isolated island. Indeed, vortex streets behind Rishiri Island were confirmed by satellite cloud images. Although no complete vortex streets were seen in the SAR images used, some wavy wakes, which are hypothesized to be a vortex-shedding regime, were imaged. Figure 10 shows an example of the SAR-derived wavy wakes at (a) 1247 UTC 5 June 1995 and (b) 1254 UTC 3 May 2008, as well as the measurements from the rawinsonde soundings at Wakkanai at 1200 UTC. The easterly flows were very stable in both cases, being capped by temperature inversions. It has been reported that the transition into the vortex shedding is caused by instability in the wake flow itself (Schär and Smith 1993b; Schär and Durran 1997), where the sufficient increase of the reversed flow in the wake can make the flow unstable. In fact, areas of relatively strong wind appear along the center of the wakes in Fig. 10; suggesting the existence of reversed flows. However, neither wind direction nor temporal evolution related to a spatiotemporal instability can be generally determined by the SAR image itself. In the present study, the vortex-shedding patterns were tentatively classified into type B.

Fig. 10.

As in Fig. 5, but for (a) 1247 UTC 5 Jun 1995 and (b) 1254 UTC 3 May 2008 and the vertical profiles at 1200 UTC.

Fig. 10.

As in Fig. 5, but for (a) 1247 UTC 5 Jun 1995 and (b) 1254 UTC 3 May 2008 and the vertical profiles at 1200 UTC.

As mentioned above, the strong low-level jets of type B tend to develop on the left side of the island relative to the direction of the traveling wind. This phenomenon, known as the “corner effect,” can be explained as follows: an incident flow weakens or stagnates on the windward side of the mountain due to blocking by topography, letting the Coriolis force acting toward the right of the flow weaken. Then the pressure gradient force surpasses the Coriolis force, letting the flow bend to the left in the Northern Hemisphere (Arakawa 2004). Numerical simulations that incorporated the earth’s rotation indeed showed that most of the low-level flows went around and strengthened on the left side of an obstacle (Thorsteinsson 1988). In the present study, assuming a horizontal scale L of 16 km, the wind speed W—corresponding to a Rossby number of 1 (Ro = U/fL, where f is the Coriolis parameter)—is estimated to be 1.6 m s−1. Thus, the Coriolis force might act on the flow stagnated on the windward side, letting it turn to the left. Numerical simulations are however needed to elucidate an actual dynamical process contributing to the phenomenon.

It should be noted that because the Wakkanai rawinsonde station is located northeast of Rishiri Island (Fig. 1b), the calculated low-level stabilities might be affected by the island, especially for downstream conditions, such as southwest incident flows. To reduce this possible source of influence, we calculate histograms for the upstream cases only, which are shown in gray in Fig. 9b, where we select the data whose mean wind direction lower than 1700 m is from northwest to southeast in a clockwise direction. Though the number of samples is less, the distribution is on the whole similar to those of the total histograms and the downstream histograms parts shown in black. Therefore, although the possibility of the island effect on each observation cannot be excluded, its influence on the result is, as a whole, insignificant.

b. Seasonal dependence

We examine the seasonal dependence of the four types of wake patterns. Figure 11 depicts the frequencies of occurrence per month. Type A, characterized as the weak low-level flow, does not show clear seasonality, occurring equally throughout the year. On the other hand, types B–D, on which island forcing appears, indicate characteristic seasonal features and a connection between them. Whereas type B occurs in boreal summer from April to October and has no occurrence in winter, type D occurs mainly in winter, overlapping slightly with type B. Type C occurs through the year, overlapping with types B and D. This relationship between the seasonal dependences of the three types seems to be further evidence that type C belongs to a transitional state between types B and D, which has been seen in the histograms with respect to (Fig. 9b).

Fig. 11.

Histograms of occurrence frequency for each type (A–D) as a function of month.

Fig. 11.

Histograms of occurrence frequency for each type (A–D) as a function of month.

The apparent difference between types B and D can be explained by the seasonal dependence of the atmospheric stability. In summer, a characteristic wind known as the “yamase” occurs frequently in the northern part of Japan. The yamase is a cold and moist wind frequently accompanied by low-level clouds, blowing from the Pacific into the east coast of northern Japan from spring to autumn (e.g., Takai et al. 2006). Some type-B wind patterns (Figs. 6c and 6d) are considered to be a response to the yamase wind. Takai et al. (2006) have produced a composite image of scatterometer ocean surface winds for yamase conditions defined as anomalous low temperatures along the eastern coast of Japan. This image showed that the easterly winds around Hokkaido developed a low-level jet between Hokkaido and Sakhalin, which is analogous to Figs. 6c and 6d. It is thus indicated that one of the main mechanisms responsible for type B is an intrusion of the low-level cold easterly originating from the yamase, which tends to be capped by a temperature inversion.

In the winter, a northwesterly monsoon (a cold and dry air outbreak from the continent) is dominant. Under the monsoon, the cold air is heated by relatively warmer SSTs and is well mixed. Actually, the potential temperature profiles of type D (Fig. 8) are approximately constant up to a height of 1 km. Furthermore, as mentioned, the mottled pattern in the SAR images is an indicator of the well-mixed atmospheric boundary layer and of the type-D wind pattern.

5. Discussion

a. Comparison with the -dependent flow regimes

The configurations of the SAR-derived wind types and their transition dependent on suggests that types D, C, and B correspond to the -dependent regimes described in the introduction: small-amplitude waves, wave breaking, and flow splitting, respectively. In the small-amplitude wave regime, laboratory experiments demonstrated that for the flow separated around the top of the mountain, forming a boundary layer below in the leeward slope of the model hill (Hunt and Snyder 1980). Their flow field is consistent with those of type D. When exceeds a critical value , wave breaking occurs above the leeward slope. Note that has been reported to be in the range of 1.15–1.20 for an axisymmetric mountain (β = 1) (e.g., Bauer et al. 2000; Epifanio 2003), although it is dependent on β as well as the vertical shape of the mountain. Wave breaking leads to the acceleration of the low-level lee-slope flow. Thus, type C, which is characterized by the low-level jet in the lee and is roughly distributed in the of 1.0–2.5, is associated with the wave-breaking regime. Meanwhile, it should be noted that Smith et al. (1997) indicated wave breaking (wave dynamics) as a factor contributing to the development of a long straight (weak) wake observed in the lee of St. Vincent. Its configuration resembles some of the type-D wakes. In fact, the atmospheric conditions under which the weak wakes were observed are similar to those of some type-D cases shown in Fig. 8: neutral stratification in the lower layer with moderate stratification capped under an inversion on top of the atmospheric boundary layer. It is thus implied that not only a boundary layer separation but also wave dynamics contributed to some of the type-D wakes. The altitude at which an inversion occurs and the degree of stratification underneath might be the distinguishing factors between types C and D. An analysis taking the vertical shear of the atmospheric stability into account will be needed for such further rigorous classification.

With the further increase of , the flow splits, passing around the mountains, and lee vortices appear when exceeds another critical value . Type B is indeed characterized by wakes in the lee of the mountain with strong jets developing on both sides of the wake, being consistent with the lee vortex and flow-splitting regime. Concerning , it has been reported that whereas it is larger than for an elongated ridge (β > 1), it is nearly equal to for an axisymmetric mountain (β = 1) (e.g., Epifanio 2003). This implies that the state of type C is very sensitive to the change of around . In fact, type C is situated in a transition from type B to type D, overlapping with them (Fig. 9).

On the other hand, the transition point, type C’s mode of 1.75–2.0, is somewhat larger than the reported critical value . Here, we consider another nondimensional parameter governing the behavior of the flow: the vertical aspect ratio δ = U/Na, where a is an along-stream length scale. When δ is small, the flow is hydrostatic, while for δ ~ 1, the flow is nonhydrostatic. The hydrostatic flow was implicitly assumed in the present analysis. However, defining a to be 2 km, a half-width of Rishiri Island (see Fig. 1c), most δ are not small (δ > 0.2), so the flow cannot be assumed to be fully hydrostatic. It has been pointed out that the critical values between the flow regimes become larger with the increase of δ (Laprise and Peltier 1989). Therefore, the large vertical aspect ratio might be a possible explanation for the relatively large transition points, of 1.75–2.0. At the same time, vertical wind shear can be another contributing factor. Positive wind shear, which is frequently seen in the wind profiles in Figs. 58, is known to delay the onset of wave breaking.

b. Intensification of flow over the mountain

Wind speed intensifications are estimated for type C from the SAR-derived wind speeds as the difference between the wind speeds of the low-level jet in the lee of the mountain and those at the upstream range away from the mountain. According to Smith (1980), for an isolated bell-shaped mountain defined as

 
formula

where hm and a are the mountain height and horizontal scale, a linear theory gives a zonal velocity perturbation at the ground as

 
formula

where N is the Brunt–Väisälä frequency. It reaches its maximum at x = a/21/2 where umax = hmN2/33/2. Figure 12 depicts the plots of the estimated wind increases as a function of umax. In the present case, as hm is constant, umax is proportional to N. The wind intensification ranging from 1 to 10 m s−1 is approximately proportional to N, being on the order of umax. Despite the rough estimate, the present study demonstrates that the wind speed of the low-level jet in the lee of the mountain intensifies with the increasing of N. However, the theoretical flow (u′) approaches the maximum on the leeward slope and decays rapidly outside the mountain, which is apparently inconsistent with the SAR measurements. In addition, because the SAR measurements are area averages over the jets, their maximums are expected to exceed the estimation by the linear theory. It is thus suggested that the nonlinear effect such as breaking plays a significant role in the development of the low-level jets of type C, as discussed in the previous section.

Fig. 12.

Wind speed increases over the low-level jets of type C as a function of Nh2/33/2, where N is the Brunt–Väisälä frequency and h is the mountain height, which correspond to those estimated based on the linear theory of Smith (1980).

Fig. 12.

Wind speed increases over the low-level jets of type C as a function of Nh2/33/2, where N is the Brunt–Väisälä frequency and h is the mountain height, which correspond to those estimated based on the linear theory of Smith (1980).

6. Summary and conclusions

We investigated the characteristics of ocean surface winds in the lee of Rishiri Island, linking atmospheric stability and flow speeds and using a number of high spatial resolution wind images derived by SAR and in situ atmospheric observations. We obtained the following results:

  1. From the SAR-derived wind map, we found evidence that a low-level jet formed in the lee of the island under the ambient stably stratified flow. The time series of wind measurements on the islands and the rawinsonde soundings suggested that the jet was connected to downslope winds because the wind speeds observed in the leeward foot of the island rapidly intensified, which was concurrent with an increase in atmospheric stability in the atmospheric boundary layer.

  2. The 115 SAR wind images were classified into the following four types in terms of wind pattern around the island: type A, which showed no significant island wakes and accounted for 17% of the total; type B, which represented wind shadows in the lee of the island accompanying low-level jets on both sides of the shadow area, the left jet being generally stronger than the right one, and accounted for 27%; type C, which depicted low-level jets that formed in the lee of the island and accounted for 19%; and type D, which indicated wind shadows in the lee of the island, but showed no jets that differ from type B, and accounted for 36%.

  3. Type A corresponded to weak ambient flow conditions. On the other hand, types B–D with their significant island wakes exhibited a dependence on the nondimensional mountain height . Whereas type B occurred under the large (>2.0) flows, type D tended to occur under relatively small (<1.75) and type C, in general, occurred within their transition range . These wake types and their -dependent transitions were qualitatively consistent with the common regimes for hydrostatic flow past topography with uniform upstream wind speed and the Brunt–Väisälä frequency N: small-amplitude waves (quasi-linear and weakly nonlinear ranges), wave breaking, and flow splitting and lee vortices.

  4. Intensification of the low-level jets of type C that had passed over the mountain was approximately proportional to N within the cases of the present study, which was roughly consistent with the linear lee wave theory of Smith (1980) However, the importance of nonlinearity, such as wave breaking, was strongly implied due to the discrepancies between the linear theory and the observations in terms of the position and extent of the low-level jets.

  5. Occurrence frequencies of each type showed seasonal dependences that were associated with seasonal characteristics of atmospheric stability. Type B came up mainly in boreal summer from May to September, when the intrusion of the cold easterly known as the yamase frequently occurs, allowing the low-level flow to stabilize. Type D came up mainly in winter from October to March, when well-mixed low-level flows are common due to a cold-air monsoon flowing over the relatively warmer sea surface. Type C tended to occur throughout the year, falling upon a transition between types B and D.

The present study demonstrated by using actual atmospheric measurements that the behavior of the wind field around the island, being sensitive to , changed drastically. Surface winds in the lee of the island rose and fell with the monotonic increasing of around the critical values , which are considered to correspond to the wave breaking and flow splitting. We could not determine the exact critical number that defines explicitly the transition boundaries between wind regimes because the rawinsonde soundings we used were not strictly representative of the atmosphere over Rishiri Island in a spatiotemporal sense. In addition, we considered neither the critical level of incident flows nor the vertical shear of atmospheric stability due to the limitations of the representativeness of the soundings. Although these problems are beyond the scope of the present study, numerical simulations as well as simultaneous and intensive atmospheric measurements are considered of value for solving them. The general results of the present study could provide reliable observational evidence for future model studies. In any case, understanding the vertical structure of atmospheric conditions is critical to predicting the low-level jets revealed by the present study.

Acknowledgments

The authors thank the editor and the three anonymous reviewers for their constructive comments, which were helpful in improving our paper.

APPENDIX

Summary of SAR Observations and Meteorological Measurements

Table A1.

Summary of SAR observations and meteorological measurements. Displayed in each case is the SAR information: the observation time, satellite, sensor, and rawinsonde measurements, mean wind speeds U, Brunt–Väisälä frequencies N, nondimensional mountain height , classified types, and corresponding figure numbers.

Summary of SAR observations and meteorological measurements. Displayed in each case is the SAR information: the observation time, satellite, sensor, and rawinsonde measurements, mean wind speeds U, Brunt–Väisälä frequencies N, nondimensional mountain height , classified types, and corresponding figure numbers.
Summary of SAR observations and meteorological measurements. Displayed in each case is the SAR information: the observation time, satellite, sensor, and rawinsonde measurements, mean wind speeds U, Brunt–Väisälä frequencies N, nondimensional mountain height , classified types, and corresponding figure numbers.

REFERENCES

REFERENCES
Alpers
,
W.
,
U.
Pahl
, and
G.
Gross
,
1998
:
Wind fields in coastal areas studied by ERS-1 synthetic aperture radar imagery and numerical modeling
.
J. Geophys. Res.
,
103
,
7875
7886
.
Alpers
,
W.
,
A.
Ivanov
, and
J.
Horstmann
,
2009
:
Observations of bora events over the Adriatic Sea and Black Sea by spaceborne Synthetic Aperture Radar
.
Mon. Wea. Rev.
,
137
,
1150
1161
.
Arakawa
,
S.
,
2004
:
Local Winds (in Japanese)
.
Senzando Press, 157 pp
.
Bauer
,
M. H.
,
G. J.
Mayr
,
I.
Vergeiner
, and
H.
Pichler
,
2000
:
Strongly nonlinear flow over and around a three-dimensional mountain as a function of the horizontal aspect ratio
.
J. Atmos. Sci.
,
57
,
3971
3991
.
Brighton
,
P. W. M.
,
1978
:
Strongly stratified flow past three-dimensional obstacles
.
Quart. J. Roy. Meteor. Soc.
,
104
,
289
307
.
Colle
,
B. A.
,
K. A.
Loescher
,
G. S.
Young
, and
N. S.
Winstead
,
2006
:
Climatology of barrier jets along the Alaskan coast. Part II: Large-scale and sounding composites
.
Mon. Wea. Rev.
,
134
,
454
477
.
Drazin
,
P. G.
,
1961
:
On the steady flow of a fluid of variable density past an obstacle
.
Tellus
,
8
,
239
251
.
Durran
,
D. R.
,
2003
:
Downslope winds
.
Encyclopedia of the Atmospheric Sciences, J. Holton, J. Pyle, and J. Curry, Eds., Cambridge University Press, 644–650
.
Durran
,
D. R.
, and
J. P.
Klemp
,
1987
:
Another look at downslope winds. Part II: Nonlinear amplitude beneath wave-overturning layers
.
J. Atmos. Sci.
,
44
,
3402
3412
.
Epifanio
,
C. C.
,
2003
:
Lee vortices
.
Encyclopedia of the Atmospheric Sciences, J. Holton, J. Pyle, and J. Curry, Eds., Cambridge University Press, 1150–1160
.
Epifanio
,
C. C.
, and
R.
Rotunno
,
2005
:
The dynamics of orographic wake formation in flows with upstream blocking
.
J. Atmos. Sci.
,
62
,
3127
3150
.
Etling
,
D.
,
1990
:
Mesoscale vortex shedding from large islands: A comparison with laboratory experiments of rotating stratified flows
.
Meteor. Atmos. Phys.
,
43
,
145
151
.
Ferber
,
G. K.
, and
C. F.
Mass
,
1990
:
Surface pressure perturbations produced by an isolated mesoscale topographic barrier. Part II: Influence on regional circulation
.
Mon. Wea. Rev.
,
118
,
2597
2606
.
Fujita
,
Y.
, and
H.
Onoda
,
1984
:
South-westerly strong winds on the lee side of Rishiri Island (in Japanese with an English abstract)
.
J. Atmos. Res.
,
36
,
121
129
.
Horstmann
,
J.
,
W.
Koch
,
S.
Lehner
, and
R.
Tonboe
,
2000
:
Wind retrieval over the ocean using synthetic aperture radar with C-band HH polarization
.
IEEE Trans. Geosci. Remote Sens.
,
38
,
2122
2131
.
Hunt
,
J. C. R.
, and
W. H.
Snyder
,
1980
:
Experiments on stably and neutrally stratified flow over a model three-dimensional hill
.
J. Fluid Mech.
,
96
,
671
704
.
Isoguchi
,
O.
, and
H.
Kawamura
,
2007
:
Coastal wind jets flowing into Tsushima Strait and their effect on wind-wave development
.
J. Atmos. Sci.
,
64
,
564
578
.
Isoguchi
,
O.
, and
M.
Shimada
,
2009
:
An L-band ocean geophysical model function derived from PALSAR
.
IEEE Trans. Geosci. Remote Sens.
,
47
,
1925
1936
.
Kawai
,
Y.
,
H.
Kawamura
,
S.
Takahashi
,
K.
Hosoda
,
H.
Murakami
,
M.
Kachi
, and
L.
Guan
,
2006
:
Satellite-based high-resolution global optimum interpolation sea surface temperature data
.
J. Geophys. Res.
,
111
,
C06016
,
doi:10.1029/2005JC003313
.
Laprise
,
R.
, and
W. R.
Peltier
,
1989
:
On the structural characteristics of steady finite-amplitude mountain waves over the bell-shaped topography
.
J. Atmos. Sci.
,
46
,
586
595
.
Lilly
,
D. K.
, and
J. B.
Klemp
,
1979
:
The effects of terrain shape on nonlinear hydrostatic mountain waves
.
J. Fluid Mech.
,
96
,
671
704
.
Loescher
,
K. A.
,
G. S.
Young
,
B. A.
Colle
, and
N. S.
Winstead
,
2006
:
Climatology of barrier jets along the Alaskan coast. Part I: Spatial and temporal distributions
.
Mon. Wea. Rev.
,
134
,
437
453
.
Long
,
R. R.
,
1953
:
Some aspects of the flow of stratified fluids, 1. A theoretical investigation
.
Tellus
,
5
,
42
58
.
Mass
,
C. F.
, and
G. K.
Ferber
,
1990
:
Surface pressure perturbations produced by an isolated mesoscale topographic barrier. Part I: General characteristics and dynamics
.
Mon. Wea. Rev.
,
118
,
2579
2596
.
Miranda
,
P. M. A.
, and
M. A.
Valente
,
1997
:
Critical level resonance in three-dimensional flow past isolated mountains
.
J. Atmos. Sci.
,
54
,
1574
1588
.
Mitnik
,
L. M.
,
1992
:
Mesoscale coherent structures in the surface wind field during cold air out breaks over the Far Eastern seas from satellite side looking radar
.
Mer (Paris)
,
30
,
287
296
.
Monaldo
,
F. M.
,
D. R.
Thompson
,
R. C.
Beal
,
W. G.
Pichel
, and
P.
Clemente-Colón
,
2001
:
Comparison of SAR-derived wind speed with model predictions and ocean buoy measurements
.
IEEE Trans. Geosci. Remote Sens.
,
39
,
2587
2600
.
Monaldo
,
F. M.
,
D. R.
Thompson
,
W. G.
Pichel
, and
P.
Clemente-Colón
,
2004
:
A systematic comparison of QuikSCAT and SAR ocean surface wind speeds
.
IEEE Trans. Geosci. Remote Sens.
,
42
,
283
291
.
Pan
,
F.
, and
R. B.
Smith
,
1999
:
Gap winds and wakes: SAR observations and numerical simulations
.
J. Atmos. Sci.
,
56
,
905
923
.
Reinecke
,
P. A.
, and
D. R.
Durran
,
2008
:
Estimating topographic blocking using a Froude number when the static stability is nonuniform
.
J. Atmos. Sci.
,
65
,
1035
1048
.
Sandvik
,
A. D.
, and
B. R.
Furevik
,
2002
:
Case study of a coastal jet at Spitsbergen—Comparison of SAR- and model-estimated wind
.
Mon. Wea. Rev.
,
130
,
1040
1051
.
Schär
,
C.
, and
R. B.
Smith
,
1993a
:
Shallow-water flow past isolated topography. Part I: Vorticity production and wake formation
.
J. Atmos. Sci.
,
50
,
1373
1400
.
Schär
,
C.
, and
R. B.
Smith
,
1993b
:
Shallow-water flow past isolated topography. Part II: Transition to vortex shedding
.
J. Atmos. Sci.
,
50
,
1401
1412
.
Schär
,
C.
, and
D. R.
Durran
,
1997
:
Vortex formation and vortex shedding in continuously stratified flows past isolated topography
.
J. Atmos. Sci.
,
54
,
534
554
.
Shimada
,
M.
,
1999
:
Verification processor for SAR calibration and interferometry
.
Adv. Space Res.
,
23
,
1477
1486
.
Shimada
,
M.
,
2002
:
User’s guide to NASDA’s SAR products version 3
.
NDX-000291, 27 pp
.
Shimada
,
M.
,
O.
Isoguchi
,
T.
Tadono
, and
K.
Isono
,
2009
:
PALSAR radiometric and geometric calibration
.
IEEE Trans. Geosci. Remote Sens.
,
47
,
3915
3932
.
Shimada
,
T.
, and
H.
Kawamura
,
2004
:
Wind jets and wind waves off the Pacific coast of northern Japan under winter monsoon captured by combined use of scatterometer, synthetic aperture radar, and altimeter
.
J. Geophys. Res.
,
109
,
C12027
,
doi:10.1029/2004JC002450
.
Shimada
,
T.
, and
H.
Kawamura
,
2006
:
Wind-wave development under alternating wind jets and wakes induced by orographic effects
.
Geophys. Res. Lett.
,
33
,
L02602
,
doi:10.1029/2005GL025241
.
Shimada
,
T.
,
H.
Kawamura
, and
M.
Shimada
,
2003
:
An L-band geophysical model function for SAR wind retrieval using JERS-1 SAR
.
IEEE Trans. Geosci. Remote Sens.
,
41
,
518
531
.
Shimada
,
T.
,
O.
Isoguchi
, and
H.
Kawamura
,
2008
:
Numerical simulations of wind wave growth under a coastal wind jet through the Kanmon Strait
.
Wea. Forecasting
,
23
,
1162
1175
.
Smith
,
R. B.
,
1980
:
Linear theory of stratified hydrostatic flow past an isolated mountain
.
Tellus
,
32
,
348
364
.
Smith
,
R. B.
,
1985
:
On severe downslope winds
.
J. Atmos. Sci.
,
42
,
2597
2603
.
Smith
,
R. B.
, and
V.
Grubišić
,
1993
:
Aerial observations of Hawaii’s wake
.
J. Atmos. Sci.
,
50
,
3728
3750
.
Smith
,
R. B.
,
A. C.
Gleason
,
P. A.
Gluhosky
, and
V.
Grubišić
,
1997
:
The wake of St. Vincent
.
J. Atmos. Sci.
,
54
,
606
623
.
Smolarkiewicz
,
P. K.
, and
R.
Rotunno
,
1989
:
Low Froude number flow past three-dimensional obstacles. Part I: Baroclinically generated lee vortices
.
J. Atmos. Sci.
,
46
,
1154
1164
.
Stoffelen
,
A.
, and
D.
Anderson
,
1997
:
Scatterometer data interpretation: Estimation and validation of the transfer function CMOD4
.
J. Geophys. Res.
,
102
,
5767
5780
.
Takai
,
H.
,
H.
Kawamura
, and
O.
Isoguchi
,
2006
:
Characteristics of the yamase winds over the oceans around Japan observed by the scatterometer-derived ocean surface winds
.
J. Meteor. Soc. Japan
,
84
,
365
373
.
Thorsteinsson
,
S.
,
1988
:
Finite amplitude stratified air flow past isolated mountains on an f-plane
.
Tellus
,
40A
,
220
236
.
Walter
,
B. A.
, and
J. E.
Overland
,
1982
:
Response of stratified flow in the lee of the Olympic Mountains
.
Mon. Wea. Rev.
,
110
,
1458
1473
.
Winstead
,
N. S.
, and
Coauthors
,
2006
:
Using SAR remote sensing, field observations, and models to better understand coastal flows in the Gulf of Alaska
.
Bull. Amer. Meteor. Soc.
,
87
,
787
800
.
Xie
,
S.-P.
,
W. T.
Liu
,
Q.
Liu
, and
M.
Nonaka
,
2001
:
Far-reaching effects of the Hawaiian Islands on the Pacific ocean–atmosphere system
.
Science
,
292
,
2057
2060
.