Early morning, 25 June 2003 (0200 UTC, 0600 LT)

I was working the midwatch in Bahrain when we got a call from several Cyclone-class Coastal Patrol (PC) boats.

They wanted a sea-state forecast for a transit along the Oman coast into the Persian Gulf. The PCs had refueled in Djibouti and left yesterday. They recently exited the Gulf of Aden. From past experience, they were concerned about waves raised by the Somali Jet.

[ship at 15N 52E]

I pulled up the current forecast. Our sea-state forecast for this afternoon calls for 8-10 foot seas along the coast with 15-18 foot seas farther offshore in the core of the Somali Jet. So far, so good. The PCs have a 12-foot sea-state operating limit.

How about the synoptic picture? The 0000 UTC global GFS run was just becoming available, so I checked the latest forecast. Synoptic conditions aren't really expected to change much over the next day or so.

As for low-level winds, the analysis shows southwesterly winds across the northern Arabian Sea and the Gulf of Aden. The low-level barbs suggest that the Somali Jet will graze the coast of Oman.

However, winds are forecast to shift to a more westerly direction, with the core of the Somali Jet nearly 500 kilometers off the coast of Oman. That ought to keep the higher wave heights well offshore. This is a pretty typical pattern for this time of year and our sea-state forecasts for the AOR reflect it.

The model shows winds in the Gulf of Aden and along the Oman coast strengthening slightly over the next few hours. But that shouldn't be a problem and should not really build the seas.

I expect that the PCs will be well within their 12-foot sea-state operating limit, as long as they stay close to shore.

Just to be safe, I checked the COAMPS forecast. The PCs are close to the edge of the model domain. But at least we can see the Oman coast; that's where any trouble will be.

The COAMPS forecast shows more variation, but overall it looks a lot like the GFS. I notice that it tightens up the pressure gradient along the coast and moves the higher winds closer to shore, especially near Ra's Mirbat. On the other hand, COAMPS shows decreasing winds through the next few hours. Either way, the winds aren't any higher than those progged by the GFS, so if the PCs can stay close to shore they should be OK.

[ship at 16N 55E]

After a rest, I'm back at my desk. In late afternoon I get word from the Coastal Patrols. They're reporting 35-knot winds as they approach the Ra's Mirbat. Seas are building with wave heights approaching 10 feet. They're wondering if conditions are still looking good to proceed. I pull up the latest obs. They don't really help. There's only a half dozen ship obs, and they're mostly farther offshore. What about the satellite winds?

I grab the latest scatterometer winds. It's not perfect, only a once-a-day peek that occurs first thing in the morning, so the winds could have strengthened. Still, I can see that the PCs are headed into the area with the strongest winds. Not good, and I still don't know what's going to happen. What can the models tell me?

The latest COAMPS run is just coming out. The analysis still looks fine, although the wind forecast might be a little low in the model for the PCs' location. With so few observations going into the analysis, it's easy for the forecast to lag behind. Nonetheless, winds and swell should decrease to the east along the Oman coast in the next few hours.

Yikes! I just got off the horn with the PCs. They're getting beat up out there with waves of 12-15 feet. Winds are still running 30 knots. What's going on? The captain was not at all happy when we talked and wants to know when they'll be out of this stuff.

It looks like a coastal jet has formed along the mountainous coastline of Oman. The jet is maintaining the strong winds exiting the Gulf of Aden. The global model hints at this, but it doesn't really keep things that strong. COAMPS has a better depiction of the jet, but also shows weaker winds.

Fortunately, both the GFS and COAMPS show that the coastal jet should diminish farther to the east. The seas should subside as well, particularly as the PCs turn the corner and head north out of the line of wave propagation.

I get on the horn and and tell the PCs to proceed east and things will improve! I can tell they are skeptical, but they stay on course. Later I hear that the coastal jet persisted throughout the night with winds in excess of 30 knots for at least 300 nautical miles.

Low-level coastal jets occur along some coastlines due to the presence of a low-level baroclinic structure. These winds may exceed 35 knots, leading to high waves. Thus, coastal jets can significantly impact marine interests along the coast and immediately offshore. Coastal jets also lead to significant low-level vertical wind shear, thus presenting a hazard to aviation in the coastal zone.

So what exactly does a coastal jet look like? This example comes from California in August of 1999. We will use this case to illustrate the features of coastal jets throughout the module. This loop shows surface winds for a 24-hour period.

Without a doubt, one difference between the surface winds over land and those over the water is the difference in friction between land and water. Although lower friction offshore contributes to higher windspeeds, a more fundamental difference in the winds is found in the pressure field. This plot shows sea level pressure over the California coast during a coastal jet event. Note that the pressure distribution is distorted by the reduction to sea level for higher terrain. This distortion generates complex details in mountainous regions, particularly over the Sierra Nevada on the eastern edge of the plot.

The factors that contribute to the strong cross-coast pressure gradient are most evident in a cross section of the structure across the coast.

The sloping isentropes marking the top of the marine boundary layer comprise a baroclinic structure. This baroclinic structure leads to the strongest cross-coast pressure gradient at the surface, near the coast. For reasons we will discuss later, this structure yields the jet seen in this figure, the so-called California coastal jet.

The thermal structure described previously plays the dominant role in determining the presence of the coastal jet; however, topographic effects may also contribute to its intensity.

This figure summarizes the relationship between the low-level structure and the coastal jet. Because the inversion capping the marine boundary layer slopes downward toward the coast, the low-level temperature gradient reaches a maximum there. As a result, the sea level pressure gradient also reaches a maximum near the coast. Consequently, the low-level wind increases, leading to a coastal jet.

The large-scale conditions conducive to development of a coastal jet start with persistent synoptic flow oriented parallel to the coast. This animation shows composite charts of sea level pressure and surface winds. Note the subtropical high pressure systems located over the world's oceans. Anticyclonic circulation around these high pressure centers typically provide the synoptic flow that preconditions a coastal jet. Furthermore, we have seen that coastal jets develop when a well-developed, cool marine boundary layer forms that is capped by a strong inversion. Subsidence beneath a subtropical high pressure center, combined with a cool sea surface temperature leads to an inversion over the marine boundary layer. With time, the boundary layer cools and becomes well-mixed, which strengthens the capping inversion. Due to warmer temperatures over land, the inversion slopes down to the coast. This leads to a baroclinic structure that results in an acceleration of low-level flow near the coast.

Using these conditions, we can examine the climatologies of sea level pressure, surface winds, and sea surface temperatures to predict areas that are prone to coastal jets.

Now, let’s look at the synoptic setting along the West Coast of the Unites States. The California coastal jet is associated with the larger-scale baroclinic structure of the atmosphere over the region during the summer. This becomes evident when we consider climatological analyses at 500 hPa, 850 hPa, and the surface.

In the preceding pages, we established that a thermal gradient exists across the coast, and we have looked at the synoptic climatology associated with coastal jets. How do these conditions result in a low-level coastal jet?

Recall that the vertical distance or thickness between pressure surfaces is directly related to the mean temperature in the vertical layer.

Therefore, the strong horizontal temperature gradient at the surface is associated with a strong horizontal pressure gradient force, labeled PGF, from high pressure over the cold ocean to low pressure over the warm land. This is reflected by the steep slope of the 1000-hPa pressure surface.

As we move upward to 850 hPa, cooler temperatures over cold water result in a lesser thickness over the ocean, while warmer temperatures over land lead to a greater thickness onshore. Combined, the lesser thickness over the ocean and the greater thickness over land result in a more gentle slope for the 850-hPa surface than for the 1000-hPa surface.

Continuing upward to 500 hPa, the lesser thickness of the atmosphere over the ocean and the greater thickness of the atmosphere over land result in a reversal of slope. Consequently, the 500-hPa surface slopes down toward the ocean. This is the origin of the 500-hPa trough offshore that we previously observed in the 500-hPa climatology.

If the pressure gradient force and the Coriolis force are in geostrophic balance, then the magnitude of the geostrophic wind is directly proportional to the magnitude of the pressure gradient force and Coriolis force. Since the pressure gradient force is maximized at the surface near the coast, it stands to reason that the strongest winds should be observed near the surface. Due to friction effects, the observed wind maximum is typically a few hundred meters or so above the surface. The smaller pressure gradient at 850 hPa results in lighter winds at that level. Winds at 500 hPa reverse direction in response to the reversal in horizontal pressure gradient between 850 and 500 hPa. In California, this results in 500-hPa winds with a southerly component.

The coastal jet owes its basic existence to the persistence of the baroclinic structure at low levels, as described in the previous section. The detailed structure of the marine inversion becomes key in defining the jet structure: the slope of the strong inversion dominates in determining the thickness between the layers. Putting the two layers (surface-to-850 hPa and 850-to-500 hPa) back-to-back, it is evident that there is a thermal gradient aloft that turns the wind down coast. However, it is really the low-level thermal gradient that transforms this flow into a jet, and this low-level thermal structure is definitely a coastal effect.

This cross section from our California case illustrates the previous point. At low levels, isentropes slope down toward land, while at 500 hPa, isentropes slope in the opposite direction, down toward the ocean.

Along many coastlines throughout the world, a cool marine boundary layer coupled with warm air over land yields a strong sea breeze. At night, the land cools and the thermal gradient reverses, resulting in a land breeze. In the case of the coastal jet, the land/sea breeze plays only a minor role. Why is that?

These plots show surface temperatures during daytime and nighttime for two different coastal locations.

What keeps the ocean cooler than land so that a coastal jet may develop? Cool sea surface temperatures have two origins: cold ocean currents and coastal upwelling.

When winds blow with persistence over an ocean surface, the Coriolis force acts to move surface water at an angle to the wind direction: to the right in the northern hemisphere and to the left in the southern hemisphere. When winds push water offshore, cold water rises from beneath to replace it. This upwelling substantially reduces sea surface temperatures, which in turn reduces temperatures in the marine boundary layer.

In many cases, coastal upwelling further cools a cold ocean current. For example, persistent southerly flow along the coast of southwest Africa in the warm season causes coastal upwelling that further cools the cold water of the Benguela current.

This plot shows satellite-derived sea surface temperatures for the scenario case that started this module. You can clearly see cooler water resulting from upwelling along the coast of Oman. Information like this is available in real time and gets incorporated into NWP models.

An important aspect of coastal jets is their along-coast variability, which influences the distribution of clouds, winds, waves, and upwelling. This model simulation off the California coast depicts this along-coast variation.

These two figures show detailed observations of sea level pressure and wind south of Point Arena on the coast of Northern California.

So, what drives the thinning of the marine boundary layer and the acceleration of the flow? We can explain this phenomenon through hydraulic considerations, analogous to flow through a pipe.

As the flow passes a point or cape, it effectively moves offshore, opening a void between the flow and the coast. When this occurs, the flow reacts by spreading laterally to fill that void. We refer to this spreading as an expansion fan. As the flow spreads laterally, the marine boundary layer thins. The flow must now squeeze through a thinner boundary layer, that is essentially a smaller pipe. Consequently, the flow speeds up.

Conversely, as the flow approaches a point or cape, it effectively piles into a wall as it is blocked by the terrain. In response, the marine boundary layer abruptly thickens in what is termed a hydraulic jump. At this hydraulic jump, the pipe becomes larger and the flow slows down.

The hydraulic analogy only works if the flow can be constrained to a pipe. This figure shows a mountainous coastline with a strong inversion at the top of the marine boundary layer. Thus, we can see that the flow is constrained at its base by the ocean surface and its top by the inversion. On the coastal side, we know that the inversion slopes steeply down toward the coast. With the base of the inversion below mountaintop level, the coastal mountains essentially act as a wall. The seaward side of the flow is constrained by a large-scale pressure gradient directed onshore to prevent flow away from the coastal channel. These conditions give us our "pipe."

Now that we've explored the nature of coastal jets, let's turn our attention to forecast considerations. In this section, we will reexamine the case that started this module. In that case, a coastal jet along the Oman coast hampered the passage of a group of Coastal Patrol boats in transit from Djibouti to the Persian Gulf.

The basic structure that supports a coastal jet is well depicted by synoptic scale numerical models. This loop shows the GFS forecast of sea level pressure over the Arabian Sea through 48 hours.

This plot shows the Somali Jet that results from the synoptic conditions depicted on the previous charts. This broad, large-scale low-level jet is the primary structure that is necessary before mesoscale coastal effects become important. Notice that along the Oman coast, the wind strengthens in the first 12 hours, but then stays remarkably constant through 48 hours, showing little diurnal variation.

Diurnal modulation of the cross-coast thermal gradient is expected due to the inland heating and lack of temperature change over the ocean. This contributes to some modulation of the coastal jet, but the background cross-coast pressure gradient tends to remain strong at all hours of the day. The cool MBL is of sufficient depth (up to 1000 m) to maintain a thermal gradient across the coast even with surface nocturnal cooling at night. This loop of forecast cross sections from Oman clearly shows a persistent cool marine boundary layer off the coast through 24 hours.

The 850-hPa height analysis can often be used to identify a ridge axis that cuts across the coastline, as shown in this figure from California. To the north of this ridge axis, no coastal jet occurs. To the south of the ridge axis, the offshore 850-hPa geostrophic flow occurs and a coastal jet becomes very well established.

Along-coast variations in the coastal jet wind speeds are an important aspect to forecast.

Resolution and physics appear to strongly affect a model's skill at depicting coastal jets. In this example from California, the 15-km MM5 keeps the coastal jet close to shore and accelerates winds to greater than 32 knots at the surface. On the other hand, the Eta model with 32-km resolution depicts a weaker jet located farther offshore with winds to 25 knots or so. The Eta subsequently fails to depict the fine-scale variations in wind speed induced by coastal capes and points.

Depending on the model physics, some current global models appear able to capture some of the mesoscale complexity of coastal jets. Compare the analysis for the 18-km COAMPS with the 12-hour forecast for the global GFS. While COAMPS places high winds closer to shore, both COAMPS and GFS depict wind maxima downwind of points and capes. Maximum wind speeds are actually a bit higher in the GFS forecast. Thus it appears that forecast skill for different models depends on more than just resolution. It also depends on the model physics.

In the absence of a dense network of surface ship and buoy observations, newly-available products derived from polar-orbiting satellites may provide mesoscale details of the surface wind field. Satellite-derived data is especially important in data sparse regions overseas. This figure shows a QuikSCAT-derived wind field showing details of a coastal jet off the coast of Chile.

In addition, the multiple channels of polar satellite information may be used to improve the diagnosis and monitoring of coastal jets. This figure shows a true-color MODIS image for the Northern California coast. Note the clear areas along the coastline downwind of Cape Mendocino and Point Arena. Remember that hydraulic theory predicts an expansion fan with a lowering inversion downstream of capes and points. The descending flow warms adiabatically, which dissipates the clouds.

Now compare the MODIS image with this QuikSCAT wind field for the same time period. The QuikScat image shows the presence of a low-level jet with local wind maxima downwind of both Cape Mendocino and Point Arena. Both the MODIS and QuikSCAT images reveal different manifestations of the coastal jet. QuikSCAT reveals variations in the wind field, while MODIS reveals variations in the fog and stratus. Taken together, the two images paint a more complete picture of the atmosphere than either image taken separately.

In summary, forecasting the coastal jet should focus on the sea-level pressure gradient and low-level thermal structure. Sea-level pressure analyses and careful monitoring of the magnitude of the cross-coast and along-shore pressure gradients are critical to forecasting the location of the strongest jet winds. Mesoscale numerical models, complemented by surface land, ship, and buoy observations as well as satellite observations in data-sparse regions are useful tools to use in forecasting coastal jets.

Coastal mountain ranges contribute to the intensity of coastal jets by enhancing the low-level flow immediately adjacent to the coastline. Knowledge of the local effects of terrain features on the distribution of strongest winds is critical in predicting the sea state along a coastline comprised of complex terrain. The most important mesoscale effect of the terrain on an existing jet is an acceleration of surface winds immediately downstream of capes or regions where coastal ranges jut out into the ocean.