We're aboard the U.S.S Ronald Reagan, an aircraft carrier bound for home. After two weeks in port, getting some work done at the Bremerton Shipyards, near Seattle, we were heading south for San Diego.

Nothing against Seattle, but I miss San Diego. I like my weather warm, my food hot, and my drinks cold. Beats cold drizzle and hot coffee in my book.

We were only supposed to be in port for a week, but maintenance found some unexpected work to do, so we spent an extra week roaming Seattle. Now we're trying to get ahead of the first storm of the winter and to conduct some training flights as well.

It's 1200Z on 27 October 1999. A look at the satellite loop tells the story. A cyclone 1000 kilometers to the west is heading toward Vancouver Island. We're presently located off the Oregon Coast, getting close to California. Ceilings are already dropping and we're getting sporadic rain. It looks like we won't launch planes until the front passes.

We'd rounded Cape Mendocino and were off Point Arena when we received word of a fishing boat in trouble. According to the Coast Guard, the trawler "Michael B" had apparently lost power and was adrift. The trawler was located just south of Cape Mendocino, in an area known as the Lost Coast. The Michael B had drifted close to shore and was in danger of running aground. We were the closest assistance available. The Coast Guard has dispatched a cutter to the scene and wants to know if we can assist.

The Duty Officer and the CAG (Commander, Air Group) were meeting to discuss options for the rescue. They need a weather briefing. Now.

I pull up the latest ship and buoy observations. In the rescue area, winds are running about 25 knots, out of the south-southeast. North of Cape Mendocino, winds are stronger, about 35 knots. But that shouldn't present a problem for us.

A look at the Eta model winds shows general agreement with the obs. The forecast winds are a bit lighter and come in at more of an angle to the coastline, oriented south-southwest and running about 20 to 25 knots. However, the model forecast indicates the wind speeds should be decreasing over the next few hours.

Looking at the low-level theta-e and wind fields, we can expect the front to pass through at about 00Z. Then conditions should start to improve.

The CAG dispatches 2 helicopters to evacuate the trawler.

The helicopters are on the scene and things are getting ugly. Winds are a sustained 45 knots with stronger gusts! Seas are huge. How can that be? How in the world did the model miss that?! Now the CAG is angry that I busted the forecast and his pilots are involved in some risky rescue mission. At this point, I can only hope that everything turns out all right.

Why are the winds so strong out ahead of the front? Why are they parallel to the coast when the model says that they should be coming in at about a 45° angle?

Landfalling cyclones and their attendant fronts significantly impact the structure of mesoscale wind and precipitation fields along the west coast of North America. The complex interaction of the wind field with topography can locally enhance nearshore winds. Destabilization of the atmosphere and the subsequent development of focused heavy precipitation can lead to significant flood events in mountain valleys. While the basic theory governing flow interaction with topography is well known, forecasting wind and precipitation remains a difficult task. The variability of wind, moisture, and temperature within a cyclone, when combined with complex topography, alters the significance and subsequent impact of the interaction.

This visible satellite image shows a well-developed cyclone off the Pacific Northwest. A cold-frontal cloud band makes landfall along the Washington / Oregon coast, then trails offshore farther south. In this typical synoptic situation, southwesterly flow exists ahead of the cold front, changing to northwesterly flow behind it. The environment is typically fairly stable ahead of the front, becoming less stable as the front approaches and perhaps even unstable after frontal passage.

Coastal mountain ranges parallel the West Coast from Southern California up through British Columbia and southeast Alaska. These mountains produce topographic effects on the structure and evolution of fronts and cyclones as they approach the coast. We can summarize the basic effects as follows:

1. Alteration of the structure of the wind field. Under stable conditions, the cross-mountain component of low-level flow is blocked. Under less stable conditions the low-level flow moves freely up and over the coastal mountains.
2. Locally enhanced along-coast winds, which may occur under both blocked and unblocked conditions.
3. Localized pressure gradients and ageostrophic circulations that develop and intensify as fronts make landfall. Alteration of the wind field sometimes develops well ahead of the front, but may occur postfrontally as well.
4. A slowing of the landward movement of fronts and their associated precipitation fields.
5. Enhancement of precipitation on the windward slopes of coastal mountains as flow ascends the barrier.

As fronts and cyclones approach coastal mountains, the wind exhibits a strong tendency to turn in an along-mountain direction and often increase in speed. This diagram describes the dynamic response as a front approaches a mountainous coastline from the west. Low pressure sits to the north and west. The flow well offshore is mostly geostrophic, flowing nearly parallel to the isobars and directed onshore. As the front approaches, there can be a variety of responses depending on several factors, including wind speed and wind direction, the height and orientation of the coastal mountains, and the stability of the lower atmosphere.

The Froude number is one parameter that accounts for several of these factors. The Froude number (Fr) is merely the cross-mountain component of the wind (U) divided by the stability (N = Brunt-Vaisala Frequency) times the mountain height (h): Fr = u/ (N * h). From this you can see that increasing wind speed increases the Froude number while increasing stability or mountain height decreases the Froude number.

If the cross-mountain wind component is small and/or the air mass is very stable, the Froude number is less than one. Under these conditions the low-level flow cannot ascend the ridge and is blocked. Then the flow below mountaintop level in the coastal zone is forced to turn toward low pressure and accelerate in the along-coast direction. This flow blocking produces a distinct barrier jet below mountaintop level, similar to that observed in cold air damming or other blocked flow situations. The amount of along-coast acceleration is determined by the magnitude of the pressure gradient along the mountain barrier. Because the flow partially ascends the windward slope before its ascent halts, the blocked flow tends to be cooler than air further offshore. Consequently, the blocked region has slightly higher surface pressure, which results in what appears as a windward ridge.

Windward ridging is especially prevalent when cold interior air can be advected out through coastal gaps into the blocked flow. This event occurs frequently during winter months when cold interior air is drawn toward the coast as pressure falls in response to an approaching cyclone.

As with other flow blocking scenarios, the upstream or offshore distance impacted by the blocking is determined by the Rossby radius associated with the topography. Under stable conditions at midlatitudes, the Rossby radius is typically on the order of 1000 km.

Even when the flow is not totally blocked (Fr > 1), the wind still tends to turn toward low pressure and blow more parallel to the mountain barrier. In this situation, adiabatic cooling in the air being forced up the mountain acts to produce a windward ridge of higher pressure on the upwind side of the mountain barrier. Adiabatic warming of descending air on the downwind side produces a lee trough. In response, the flow slows as it ascends the mountain range and turns across isobars toward low pressure, trying to adjust to the windward ridge. The magnitude of the response depends on the magnitude of the windward ridging, which depends on the background stability and incident flow speed. Moderately strong stability with strong flow across the topography will produce a more pronounced windward ridge.

The along-coast wind response can be highly variable up and down the coast due to variations in both the background pressure gradient as well as variations in the dynamically forced pressure field discussed previously. The maximum dynamic response tends to occur when the Froude number equals one about midway up the mountain. This results in blocked flow at low levels, but not near ridgetop level. Consequently, flow blocking and windward ridging can contribute to along-mountain flow acceleration.

This figure shows a surface analysis for 0000 UTC 7 January 1995 with a typical synoptic setting for a cyclone and attendant fronts approaching the West Coast. As the fronts near the coast, interaction with the coastal topography results in the development of a barrier jet.

Let's take a look at a 27 October 1999 case of a landfalling cyclone and associated frontal systems. We will examine the evolution of the mesoscale structure of the wind and precipitation fields along the coastal mountains. Some questions to consider include:

1. What is the synoptic pattern that produces topographic effects?
2. How does it evolve relative to the coastal topography?
3. What mesoscale effects come into play as the front comes ashore?

Since complex mesoscale flow interactions with topography occur in association with landfalling fronts, it stands to reason that higher-resolution NWP models will simulate these interactions more realistically than coarser-resolution models. In order to gain better insight into the effects of model resolution in forecasting the effects of landfalling fronts, we will now examine a case from 27 October 1999 using coarse- and fine-resolution NWP models. The coarse model is the 40-km Eta, while the finer-resolution model is the 12-km MM5.

This image shows a GOES-IR satellite image overlaid with the 12-hour Eta forecast of sea level pressure, model winds, and buoy observations, valid at 1200 UTC on 27 October. With the front still well offshore, the model forecast winds compare favorably with the observed surface winds. The model-derived winds and buoy observations both show that the flow is deflected across isobars toward lower pressure. The cross-isobar winds probably reflect an ageostrophic response of the wind field as it encounters the coastal topography.

By 1800 UTC, the Eta model depicts the front approaching the Washington/Oregon coastline. Offshore of the northern California coast, we see southwest flow ahead of the front, but a distinct change to southerly and southeasterly flow along the coast. This change reflects the ageostrophic acceleration of winds along the coast, possibly due to flow blocking.

By 0000 UTC the observed front is pushing onshore, with the coastal buoys continuing to show coast-parallel flow. However, the Eta model was much faster at bringing the front onshore and shows it well inland over northern California at this time. The buoy observations along the northern California coast show 45-knot surface winds parallel to the coast suggesting a rather well-developed barrier jet.

Clearly, the coarse-resolution Eta model forecast is too fast on the frontal movement across the coast. This may occur because the model does not adequately handle the dynamic effects of topographic blocking that slow the front down as it approaches the coast and initiates development of the barrier jet. These limitations should be taken into account when using coarse resolution models, such as the 40-km Eta, to forecast landfalling fronts along the West Coast.

Now let's consider the dynamics that produce the topographic interaction and stronger winds right along the coast. First we examine cross sections and model soundings from the MM5 model simulation. The first set of cross sections shows potential temperature and isotachs for the wind component oriented perpendicular to the cross section (in other words, the coast-parallel flow). The cross sections run ENE-WSW, perpendicular to the California coastline.

We start with the 12-hour forecast, when the front is still well offshore. Note that the isentropes, which started out fairly horizontal when the front was much further to the west, now slope upward toward the mountains. This change in slope occurred because prefrontal warm advection depressed the isentropes offshore, while those near shore stayed relatively unchanged. Consequently, isentropes slope upward toward the coast. The result is a pocket of cooler air near the surface right next to the mountains and an increase in stratification over the top of this cool air, shown by the tighter packing of the isentropes. Given the implied horizontal thermal and pressure gradients across the coast, thermal wind balance requires that the prefrontal southerly wind should increase toward the surface. The isotachs in the model show this increase and produce a wind maximum along the coast, but there is no core structure resembling a barrier jet at this time.

By 1500 UTC, the winds normal to the cross-section have strengthened considerably and there is a semblance of a barrier jet structure developing windward of the coastal range. There is still a gradual slope of isentropes downward offshore, consistent with thermal wind balance and the low-level wind maxima. This slope tends to steepen with time as the prefrontal warm air advection increases at low levels and the front moves closer to the coast.

By 1800 UTC, the cross section shows a well-developed barrier jet, with the southerly wind component intensifying to greater than 16 m/s. Notice that the low-level stratification has not increased much over the coastal zone in this case, which limits its longevity. However, as the warm prefrontal air progresses closer to the coast, the slope of the 288K isentrope steepens, which amplifies the barrier jet wind speed.

By 2100 UTC, the cold front lies just offshore. Cool temperatures behind the cold front result in a ridging of the isentropes with the slope of isentropic surfaces now reversed to slope downward toward the coast. Also note that the barrier jet has now been lifted toward the mountaintop level and the isentropes out ahead of the front have flattened.

Although the winds near the base of the mountains continue to be strong ahead of the front, they have weakened when compared with the previous time period. Recall that the isentropic slope was much greater at that time due to the trapping of the cooler air against the mountains.

Now let's examine MM5 model cross sections of the flow perpendicular to the coastal mountains for the same period to further illustrate how it interacts with the mountains.

At 1200 UTC the wind speed at around 950 hPa decreases from 14 m/s far offshore to 6 m/s near the coastal mountains. This speed decrease is most pronounced below mountaintop level, which is consistent with some blocking of the cross-mountain flow even when the front still lies far offshore.

Now we watch the cross-mountain flow evolve as the barrier jet develops. At 1500 UTC note that, in general, the cross-mountain flow increases as the front approaches the coast. However, flow near the base of the mountains remains fixed at about 6 m/s. Thus, the cross-mountain wind speeds show an increasing tendency to decelerate in the coastal region.

By 1800 UTC, when the barrier jet has become well developed, the cross-mountain flow decreases from 22 m/s offshore ahead of the front to less than 10 m/s along the coast in the barrier jet. This deceleration is consistent with flow blocking and the increasing slope of the isentropes in the coastal region.

Is the prefrontal flow blocked? We can determine this by examination of the Froude number (Fr). Recall that Fr=U/(Nh), where U is the cross-mountain flow, N is the Brunt-Vaisala frequency, and h is the mountain height. When the Fr < 1, the flow tends to be blocked, while the winds tend to flow over the mountain barrier for Fr > 1.

This figure shows a 6-hr MM5 forecast of equivalent potential temperature (theta-e), Brunt-Vaisala frequency (N), and 24-meter winds valid at 1800 UTC on 27 October. Recall that this is the time when a pronounced barrier jet was evident on the previous cross sections. In the northern part of the domain, we note the warm, moist theta-e ridge ahead of the approaching cold front, with cooler, drier air (lower theta-e) behind the front.

Sidebar: Computation of the Froude number
If wind speed = 10 m/s, wind angle = 40°, N = 0.015 s-1, and h = 1500 m,
and Fr = U / (Nh)
Fr = 0.47

Now we will attempt to ascertain if the low-level flow will stay blocked as the front approaches the coastline and subsequently passes inland.

At 2100 UTC, conditions have not changed much. The front is still offshore and the flow ahead of the front is still about 25-30 knots toward the coastal topography. The Brunt-Vaisala frequency remains high ahead of the front and drops off dramatically behind it. This suggests that the flow continues to be blocked to maintain the barrier jet along the coast.

While the Brunt-Vaisala Frequency is a useful measure of atmospheric stability and a factor in the computation of the Froude number, it is not commonly used in an operational setting. A more common measure of stability is the atmospheric lapse rate. For example, we all know that the dry adiabatic lapse rate is about 9.8°C/km and that the lapse rate of the standard atmosphere in the U.S. is about 6.5°C/km. What might not be clear is that the Brunt-Vaisala Frequency and the lapse rate are closely related. This figure shows both the MM5 forecast of lapse rate and Brunt-Vaisala Frequency for 2100 UTC. Note that the yellow lapse rate contours closely mimic the color-filled Brunt-Vaisala Frequency contours. We can see that the least stable area in the northwest corner of the plot has a lapse rate exceeding 8°C/km and a Brunt-Vaisala Frequency less than 6 x 10^-3 (0.006). In other words, as the lapse rate approaches 9.8°C/km, the Brunt-Vaisala Frequency approaches 0. Similarly, a lapse rate of 6.5°C/km roughly corresponds with a Brunt-Vaisala Frequency of 10 x 10^-3 (0.010).

On this figure, note the sounding location near Cape Mendocino.

This animation of model soundings shows the evolution of the atmosphere just south of Cape Mendocino as the front makes landfall. Compare the relatively stable prefrontal environment to the much less stable postfrontal environment. The initial sounding in the sequence shows a very stable marine boundary layer capped by an inversion, while the last sounding displays a low-level lapse rate very close to dry adiabatic. Also note the pronounced wind shift with frontal passage, from more southerly to more westerly. These changes are typical of landfalling fronts and lead to the topographic interactions we just discussed.

In conclusion, the interaction of landfalling fronts with coastal topography produces important modifications to coastal winds. The tendency of the coastal mountains to block the onshore advection of warm air results in low-level blocking and the development of a barrier jet along the coastal mountains. Strong differential warm advection that helps to increase stability with moderate onshore flow is highly favorable for flow blocking and a barrier jet. The angle that the front makes with the coast can be important in determining the degree to which flow blocking might occur.

Now let's consider a different sort of event where the synoptic situation is similar, yet the wind field evolves in a different manner. In this case, while coastal topography exerts little effect on flow ahead of the front, it still contributes to a turning and acceleration of winds after frontal passage.

This animation shows IR satellite imagery and sea level pressure forecasts for a cyclone and landfalling front on 14-15 March 2003. The animation shows a well-developed cyclone tracking northeastward toward the coast of British Columbia from a position several hundred kilometers off the California/Oregon coast. A frontal cloud band trails to the southwest of the cyclone. The western edge of the frontal cloud band is sharply defined, and gives a good indication of the surface front location. It appears that the surface front first comes ashore near the California/Oregon border between 0600 and 0900 UTC on 15 March.

The prefrontal winds in this case are rather strong (greater than 30 kt) and coast-parallel, similar to the case with blocked flow.

Six hours later at 0000 UTC on 16 March, nearshore winds continue to parallel the coast and the southerly response extends even farther to the south, with prominent ridging in the sea-level pressure field near the coast and troughing inland of the coastal mountains. The synoptic and mesoscale pressure pattern has remained rather similar throughout the period, which allows the flow to adjust more geostrophically to the pressure distribution. This extended period of mesoscale ridging is particularly important. It allows time for the winds to respond to the mesoscale pressure distribution.

Obviously, the evolution here is somewhat different than what we saw in the earlier case. In both instances, we observed the prefrontal winds to be strong and rather coast parallel. However, the prefrontal structure was different due to the differing orientation of the front relative to the coast. In this case, as the front moved onshore, the winds turned toward the coast for only a brief period, followed by a return to a southerly, coast-parallel direction. So what are the dynamics in this situation both ahead of and behind the front?

Here we see an E-W cross section of the potential temperature and the wind speed perpendicular to the cross section for 0300 UTC, about 3 hours before the front made landfall. A prominent 40 m/s southerly jet occurs ahead of the front at 900 hPa and propagates along with it. As we saw previously, this prefrontal jet occurred even when the front was offshore and is due to the baroclinic structure of the front. Note that the cross section shows little of the isentropic rise toward the coast that we saw in the previous case when the flow was blocked. The synoptic-scale flow that occurs ahead of the front in this case is parallel to the coastal mountains so there is very little interaction with the topography.

By 1200 UTC, the front is past the coastal mountains as shown by the isentropes sloping downward to the east. There is a weak coastal jet near the coastal mountains with a peak southerly component of 10 kt. This feature appears locked to the topography. This weak barrier jet continues for the next 12 hours or so as noted previously.

Now consider the cross-barrier flow. The E-W cross-section above shows the potential temperature and cross-coast wind component at 0300 UTC, the same time that we observed the 40 m/s southerly jet. Note that the low-level cross-barrier flow has negative values in the vicinity of the coastal range, indicating offshore flow. Even though the stratification is probably sufficient to produce blocking, the prefrontal wind component is directed offshore and so it produces very little response.

As the front passes inland, the flow over the topography increases to about 6 m/s.

The flow up over the mountains leads to the subtle windward ridging that we see in the cross section.

To summarize what we've observed in the cross sections:

1. In the prefrontal environment, we observe a 40 m/s, southerly prefrontal jet. Meanwhile, near the base of the mountains, the component of flow parallel to the cross section is directed offshore at about 8 knots.
2. In the postfrontal environment, we see a weak jet-like feature in the component of wind oriented perpendicular to the cross section of about 10 knots. The wind component parallel to the cross section shows flow of about 6 knots up over the mountains. This cross-mountain flow is accomodated by weak stability in the post frontal environment, and produces a windward ridge / lee trough response in the pressure field.

We have examined two cases that exhibited significantly different dynamic situations. In the first case, stably stratified cool air was trapped at low levels on the windward side of the coastal range. Although a windward pressure ridge developed, the flow was blocked by the topography and accelerated ageostrophically, resulting in the development of a strong southerly coast-parallel low-level jet. The blocked low-level flow resulted in slowing of the surface front as it approached the coast. In extreme cases, the front aloft continues toward the coast while the surface front stalls offshore. The resulting split front destabilizes the prefrontal environment, which enhances convective precipitation on the windward side of the coastal range.

In the second case, the prefrontal flow was coast parallel and resulted in minimal topographic interaction. Prefrontal winds were strong and coast parallel because the front had a strong low-level jet that was oriented along the coast due to the direction the front approached the coast. However, the postfrontal flow turned southerly and coast parallel except for a brief period of westerly flow. Only a very weak barrier jet developed in the postfrontal flow in this case but its existence was due to interaction of the cross-coast flow with the topography. The environment was less stable so that the flow was not blocked and was relatively unimpeded as it moved up and over the coastal range. Low-level windward pressure ridging and leeside pressure troughing resulted from upslope cooling and downslope warming as flow crossed the coastal mountains. The persistence of this pattern allows a 10-knot, along-coast flow to develop in association with the pressure gradient due to the windward ridging.

To illustrate the potential for postfrontal interaction let's briefly look at another case of a landfalling front. This case occurred on 14 March 1979 along the southeast Alaska coast and was described by Overland and Bond (1993). As we see in this animation, the landfalling front was aligned nearly parallel to the coast. Prefrontal winds were largely parallel to the coast, while postfrontal winds were more nearly perpendicular to the coast. Postfrontal flow up and over the mountains resulted in a pronounced windward ridge/lee trough.

This figure shows a time series of wind observations at three surface sites along the coast. Frontal passage is denoted by the dashed line, as the front swings into the coast, then progresses northward. This plot shows very strong postfrontal winds in response to a strong pressure gradient that developed 3-6 hours after frontal passage. The largest pressure rises behind the front occurred where the cross-mountain flow was most favorable to produce windward ridging as it flowed up and over the coastal mountains. Adiabatic cooling of the postfrontal air rising up the slope produced a rapid pressure rise. This lead to development of a windward ridge and an associated strong ageostrophic wind response, with wind gusts to 31 m/s in YAK! This response was much stronger than that seen in the case we examined earlier along the California coast.

The previous cases illustrate the complexity of the dynamics associated with landfalling fronts along the West Coast. There can be many different dynamic scenarios, depending on the flow direction, stability, and a variety of other factors, which can alter the significance and subsequent impact of topographic effects. These various factors must be closely monitored in order to assess the potential impacts of topography on the mesoscale flow fields and also the precipitation patterns, which will be examined next, associated with these landfalling fronts.

As we have seen, flow blocking and low-level barrier jets are rather common features in the prefrontal environment of landfalling fronts along mountainous coasts. However, more than just the wind field is affected by interaction of flow and coastal topography. Studies have shown that flow blocking ahead of landfalling fronts impedes frontal propagation so that the surface front tends to deform and become aligned with the topography. This conceptual animation illustrates how that occurs.

As the surface front approaches land, cool air is trapped along the coast and a southerly barrier jet develops. The trapped cool air impedes the landward progress of the surface front when it gets close to the coast. Meanwhile, parts of the front that are farther from the coast still progress landward until they encroach upon the trapped cool air, at which time their movement slows. With time, the surface front tends to deform and mirror the shape of the mountains.

This animation shows in cross section an extreme example of what can happen when a landfalling front encounters cool air trapped against the base of coastal mountains. The animation is based on a landfalling front in southern California described by Neiman et al. (2003).

While the surface front slows or stalls, the upper portion of the front is unimpeded by the blocking and continues to move toward the coast. Eventually, it shears off from and overruns the surface front. As a result, the cool, postfrontal air mass aloft flows over the warm, moist prefrontal air mass below it. This cool intrusion aloft destabilizes the near-coast environment, sometimes enhancing convective rainfall on the windward side of the coastal mountains. This destabilization combines with orographic lifting to produce very heavy rains and subsequent flooding in mountain valleys and rivers.

Eventually, a secondary frontal surge may sweep through. The merger of the primary and secondary fronts frequently leads to a narrow cold frontal rainband. Passage of the merged front typically scours out the trapped air near the surface. West winds behind the secondary front also increase the flow component normal to the mountains, combining with the lower postfrontal stability to increase the Froude number to values much greater than one. As a consequence, the flow at all levels is no longer blocked and passes freely over the coastal mountains.

Precipitation fields are profoundly altered by the interaction of topography with landfalling fronts. Orographic effects modify the rainfall on both windward and leeward slopes.

As this animation clearly shows, precipitation associated with a landfalling front is concentrated on windward slopes, both along the coast and further inland in the Sierra Nevada.

At its most basic, the principal ingredients required for precipitation are moisture and lift. Upslope flow provides lift. Thus it follows that the greatest rainfall occurs where the flow is forced most strongly up a slope (given a similar moisture distribution). Conversely, lee slopes tend to experience descending flow, which inhibits precipitation.

To illustrate how the dynamic interactions of the flow with topography can affect the mesoscale organization of precipitation, let's revisit the case of 27-28 October 1999. This figure shows the 6-hour Eta model forecast of the 850-hPa wind and precipitation for 1800 UTC on 27 October. Note the broad southerly flow ahead of the approaching front as well as the maximum in precipitation exceeding 0.4 inches offshore of the northern California coast, in the vicinity of the front.

Six hours later, at 0000 UTC, the precipitation maximum approaches 0.7 inches in the vicinity of the front along the northern California coast. Also noteworthy are the strong southwest winds at 850 hPa, just ahead of the front. Given this flow and the topography, one might expect to see the largest precipitation amounts in southwest Oregon where the flow is strongest and oriented more across-mountain than further north or south. However, the Eta model appears to predict only coarse features and does not produce any mesoscale details of the precipitation distribution resulting from mesoscale interactions with topography.

Now let's look at a higher resolution MM5 mesoscale model simulation of the 27-28 October 1999 case. The figures show 840-hPa wind and 3-hour accumulated precipitation. The 6-hour forecast valid 1800 UTC on February 27 progs a broad band of precipitation extending southwestward offshore in association with the front, similar to what was depicted in the coarser Eta model run. However, the MM5 model simulation depicts much greater mesoscale detail in the prefrontal precipitation field with maxima approaching or exceeding 1 inch along and immediately inland of the northern California/southern Oregon coastline. This detail reflects the ability of the higher resolution MM5 to more realistically simulate the topographic interactions that were altogether missed by the Eta model.

As the front moves onshore, the MM5 forecast valid at 2100 UTC continues to show mesoscale details of the orographic enhancement of precipitation on windward slopes.

By 0000 UTC on 28 October, the MM5 simulation shows the front progressing inland over northern California, with a continuation of mesoscale orographic enhancement of precipitation now extending eastward to the mountains of the Sierra Nevada.

It is clear that higher-resolution model simulations are much better at depicting the mesoscale details of both the wind and precipitation fields associated with landfalling fronts.

The orographic enhancement of precipitation along the coastal range is fairly easy to understand. For less stable situations where the flow across the mountains is relatively unimpeded, the maximum precipitation should be closely correlated with the area where maximum orographic lifting is occurring. However, if the environment is more stable, blocking of the flow can complicate the precipitation distribution. As this diagram shows, flow blocking may shift the orographic precipitation offshore, away from the windward slope. In this case, isentropes are shifted upward in the blocked region, resulting in the strongest isentropic lift occurring upstream of the topography.