Simulation of Summer Diurnal Circulations over the Northwest United States

Matthew C. Brewer Department of Atmospheric Sciences, University of Washington, Seattle, Washington

Search for other papers by Matthew C. Brewer in
Current site
Google Scholar
PubMed
Close
and
Clifford F. Mass Department of Atmospheric Sciences, University of Washington, Seattle, Washington

Search for other papers by Clifford F. Mass in
Current site
Google Scholar
PubMed
Close
Full access

We are aware of a technical issue preventing figures and tables from showing in some newly published articles in the full-text HTML view.
While we are resolving the problem, please use the online PDF version of these articles to view figures and tables.

Abstract

During the summer, strong surface heating combines with the terrain and land–water contrasts of the northwest United States to create a complex array of diurnal circulations. Though observational and modeling studies have described some of these circulations, advances in high-resolution numerical modeling allow for a more comprehensive and three-dimensional examination. To simulate typical summer conditions over the Pacific Northwest, 3-hourly Global Forecast System (GFS) model output for July and August 2009–11 was used to initialize and provide boundary conditions for a high-resolution Weather Research and Forecasting (WRF) Model run. To ensure the realism of the simulation, it was compared to observations from a collection of days representing typical summer conditions. Generally, it was found that the simulated diurnal wind, relative humidity, and temperature were close to the observations. It is shown that regional diurnal circulations over the Pacific Northwest occur on a number of interacting scales, ranging from upslope/downslope winds on local terrain features to larger-scale circulations such as between the Pacific Ocean and the western Oregon and Washington interiors. Such multiscale diurnal circulations occur concurrently, with the interactions producing complex structures, several of which are described in this paper. Wind speeds in the Strait of Juan de Fuca and downstream of the major Cascade Mountain gaps reach maxima between 2100 and 2400 local daylight time (LDT), while most other areas have peak winds earlier in the day. Localized nocturnal low-level wind maxima are described, including one over the northern Willamette Valley and another over the high plateau of eastern Oregon.

Corresponding author address: Matthew C. Brewer, Dept. of Atmospheric Sciences, University of Washington, Box 351640, Seattle, WA 98195-1640. E-mail: mcbrewer83@yahoo.com

Abstract

During the summer, strong surface heating combines with the terrain and land–water contrasts of the northwest United States to create a complex array of diurnal circulations. Though observational and modeling studies have described some of these circulations, advances in high-resolution numerical modeling allow for a more comprehensive and three-dimensional examination. To simulate typical summer conditions over the Pacific Northwest, 3-hourly Global Forecast System (GFS) model output for July and August 2009–11 was used to initialize and provide boundary conditions for a high-resolution Weather Research and Forecasting (WRF) Model run. To ensure the realism of the simulation, it was compared to observations from a collection of days representing typical summer conditions. Generally, it was found that the simulated diurnal wind, relative humidity, and temperature were close to the observations. It is shown that regional diurnal circulations over the Pacific Northwest occur on a number of interacting scales, ranging from upslope/downslope winds on local terrain features to larger-scale circulations such as between the Pacific Ocean and the western Oregon and Washington interiors. Such multiscale diurnal circulations occur concurrently, with the interactions producing complex structures, several of which are described in this paper. Wind speeds in the Strait of Juan de Fuca and downstream of the major Cascade Mountain gaps reach maxima between 2100 and 2400 local daylight time (LDT), while most other areas have peak winds earlier in the day. Localized nocturnal low-level wind maxima are described, including one over the northern Willamette Valley and another over the high plateau of eastern Oregon.

Corresponding author address: Matthew C. Brewer, Dept. of Atmospheric Sciences, University of Washington, Box 351640, Seattle, WA 98195-1640. E-mail: mcbrewer83@yahoo.com

1. Introduction

The timing, intensity, and evolution of lower-tropospheric diurnal circulations are important for aviation, wind energy, agriculture, and boating, among other applications. For example, fire managers must be familiar with the diurnal meteorology of wildfire areas before making critical decisions that affect life and property, and an understanding of diurnal winds sheds light on regional variations in air quality and the transport of pollutants.

Thermally induced diurnal circulations are created when surface radiative forcing interacts with variations in topography and land–water contrasts. These circulations are amplified in summer when solar insolation is strongest and land–water temperature differences are generally greatest. The summer season is also characterized by fewer synoptic systems, fewer clouds, and reduced snow in the mountains, leading to strengthening of the diurnal winds.

The terrain of the northwest United States, its proximity to the Pacific Ocean, and the large amount of coastline in the region produce complex diurnal wind circulations on many scales. The most significant terrain feature of the region is the Cascade Mountains, with a crest approximately 100–200 km inland from the coast (Fig. 1). Elevations in the Cascades reach 2500–3000 m, with higher volcanic peaks. Significant topographic features west of the Cascades include the Olympic Mountains, the Puget Sound lowlands, the Willamette Valley, and the Coast Range. The Columbia basin lies in eastern Washington while the Oregon Plateau is found in eastern Oregon with elevations of ~1000 m.

Fig. 1.
Fig. 1.

The terrain and major geographical features of the northwest United States.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

To illustrate the average summer conditions over the region, Weather Research and Forecasting (WRF) Model output at 12-km grid spacing and averaged over July and August 2009–11 (more information may be found online at http://www.atmos.washington.edu/mm5rt/info.html) is shown in Fig. 2.1 During this period, the eastern Pacific Ocean is dominated by the Pacific high, which produces northerly flow and ocean upwelling along the coast. Such northerly flow is particularly strong along the southern Oregon (Elliot and O’Brien 1977; Bielli et al. 2002) and northern California coasts (Zemba and Friehe 1987; Holt 1996; Burk and Thompson 1996; Taylor et al. 2008), with little change in strength during the day. As the heat low over the Great Basin (covering much of Nevada, western Utah, and southeastern Oregon) strengthens during the day while high pressure is maintained offshore, the low-level geopotential height gradient builds over the Cascades. The Cascades impede marine air from pushing eastward, making the interior substantially warmer than the western lowlands. The contrasting thermal conditions across the Cascades are demonstrated in Fig. 3 by histograms of daily high temperatures for July and August 2007–12 at Seattle–Tacoma International Airport (KSEA) and Hanford, Washington (HHMS). Most summer days in Seattle reach 20°–30°C while at Hanford, on the east side of the mountains, temperature maxima are more frequently between 30° and 40°C.

Fig. 2.
Fig. 2.

(left) Average 925-hPa geopotential height (m; contours) and temperature (°C; color shading), and (right) 10-m wind (vectors and color shading) from WRF output for July and August 2009–11.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Fig. 3.
Fig. 3.

Histograms of daily max temperatures for July and August 2007–12 at (left) KSEA and (right) HHMS.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Terrain features such as the Strait of Juan de Fuca, the Olympic Mountains, Chehalis Gap, Puget Sound, and the Cascade Mountains result in complex diurnal circulations over western Washington. Staley (1957) used hodographs from July and August for 3 yr to describe the diurnal winds over the Puget Sound region. Southerly or light/variable winds in the morning are replaced by northerly flow (known as the sound breeze) in the afternoon. Staley found that stations east of the Puget Sound have a maximum westerly component in the early afternoon and a maximum easterly component at night due to regional mountain–valley and sea–land-breeze circulations. Mass (1982) used vector-averaged winds from July for 3 yr and found that westerlies blow through the Strait of Juan de Fuca the entire day, reaching a maximum in the evening. He found that divergence occurs over the Puget Sound in the afternoon and convergence during the night due to diurnal circulations forced by land–water contrasts and the slopes of regional terrain features, and this is consistent with the results of Staley (1957).

Staley (1959) examined diurnal winds over eastern Washington’s Columbia basin and found outflow during the day and inflow at night. Ellensburg, on the eastern slopes of the Cascades, has a maximum wind from the northwest at 1800 local daylight time (LDT), while at Yakima and Hanford, south and southeast of Ellensburg, respectively, strong northwest winds occur a few hours later. Staley attributed these northwesterly winds to a combination of local drainage flows and cooler air coming through the gaps in the Cascades. Doran and Zhong (1994) used observations and simulations to describe these strong northwesterly winds, suggesting that they develop as the inland thermal low strengthens, drawing in cooler air from west of the Cascades. As heating subsides during the evening, cooler air in the Snoqualmie Pass accelerates down the eastern slopes (katabatic flow) and is channeled by terrain.

Though the synoptic pattern shown in Fig. 2, with high pressure offshore and a thermal low centered over the Great Basin, is dominant in the summer, there are other transient synoptic regimes that occur during this season. Periods of low-level easterly (offshore) flow develop as upper-level ridges move over the region and high surface pressure builds inland (Mass et al. 1986; Chien et al. 1997; Brewer et al. 2012). During these periods, western Oregon and Washington are under a continental rather than marine influence, resulting in above-normal temperatures. Often following easterly flow events, coastally trapped southerlies move northward along the West Coast behind the northward extension of a coastal thermal trough, resulting in a marine push and onshore flow (Mass et al. 1986; Mass and Bond 1996; Nuss et al. 2000). Less frequent synoptic evolutions include the development of a weak offshore trough or the approach of a weak front, both of which contribute to coastal southerly flow.

Although the aforementioned literature provides insights into diurnal winds over the Pacific Northwest, the three-dimensional structures and temporal evolutions of important regional diurnal circulations are not well documented. There are several important regional diurnal circulations that have not been investigated, such as the mesoscale diurnal flows of eastern Oregon. Fortunately, improvements in the resolution and physics of numerical models allow for more accurate and comprehensive simulations of such diurnal circulations.

This paper describes an approach for simulating climatological summer conditions over the Pacific Northwest and applies this technique to enhance knowledge of northwest United States diurnal circulations. Specifically, a high-resolution simulation is used to examine the differing scales of diurnal circulations over the Pacific Northwest and their three-dimensional structures, evolutions, and mutual interactions.

2. Model description

For this research, National Oceanic and Atmospheric Administration/National Weather Service (NOAA/NWS) Global Forecast System (GFS) model output at 1° latitude–longitude resolution was obtained for July and August 2009–11. These data were averaged by hour, and the resulting files were used to initialize and provide boundary conditions for a high-resolution WRF (version 3.5) run. An outer nest of 36-km grid spacing was used, along with three one-way nested domains of 12-, 4-, and -km grid spacing (Fig. 4). The -km run was primarily used in this analysis. Thompson microphysics, new versions of the Rapid Radiative Transfer Model (RRTMG) long- and shortwave radiation schemes, the Yonsei boundary layer scheme, 38 vertical full-sigma levels, Moderate Resolution Imaging Spectroradiometer (MODIS) land use, and a model top of 50 hPa were used in this simulation. The simplified Arakawa–Schubert cumulus parameterization was applied in the outer three domains, but no cumulus parameterization was used for the -km domain.

Fig. 4.
Fig. 4.

Domains for the WRF simulations.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

The model was initialized at 1700 LDT, corresponding to 0000 UTC, and was run for 48 h. A 24-h period for forecast hours 12–35 (0500–0400 LDT) was used to describe the diurnal cycle over the region. The reason for initializing the model at 1700 LDT is twofold. First, sufficient spinup time is needed to produce realistic conditions. Second, if the model is initialized at night, the low-resolution GFS initialization/boundary conditions will not resolve thermal structures in the lower troposphere produced by surface heating the prior day. This is particularly true near complex terrain.

3. Verification

The goal of the WRF run is to simulate typical summer (July and August) diurnal wind, temperature, and moisture variations. To verify the accuracy of the model simulation, it was compared to observations from a collection of days with typical summer synoptic conditions. The question is how to define objectively a “typical” summer day. Figure 5 shows histograms of wind direction for five spatially distributed stations for July and August 2009–11 at the hour of the climatological maximum wind speed at each station.2 Each station has a well-defined wind direction mode at the time of climatological maximum wind. The most frequent wind direction, as well as adjacent directions with more than 5 days, is identified (colored red) in each histogram in Fig. 5. Days outside of the mode are colored blue. Only those days within the mode (red) at all five stations were used for comparison against the WRF run. Specifically, 54% of the days (100 of 186 days) remained after the filtering was applied.

Fig. 5.
Fig. 5.

Histograms of wind direction for July and August 2009–11 at the time of the climatological max wind speed (noted at the top of each graph) for several regional observing sites. The percentage of days within the primary mode (red color) is shown.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

There are some differences among these five stations in the percentage of days characterized by the mode. For example, buoy 46050 (Fig. 5), just offshore of central Oregon, has a narrow mode encompassing about 75% of the days. There is a secondary maximum for southerly to southwesterly winds associated with coastally trapped wind reversals and the occasional trough passage. The four other stations [Ellensburg (KELN); The Dalles, Oregon (KDLS); off the southern tip of Vancouver Island (CWQK); and Hoquiam, Washington (KHQM)] have dominant winds from the west to northwest, with higher percentages of days within the mode. Easterly flow events are evident at KELN on the eastern slopes of the Cascades, where a secondary maximum is found.

The stations used for verification are distributed throughout the region, with most being high quality aviation sites (Fig. 6). Fourteen are from the Automated Surface Observing System (ASOS), while one station (CWQK) is from Environment Canada. Since CWQK does not have temperature and relative humidity measurements, nearby Port Angeles, Washington (KCLM) was used in place of CWQK for those variables.

Fig. 6.
Fig. 6.

Map of stations used for verification.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Figure 7 shows a comparison of model wind speed (green) to the scalar-average observed wind speed (blue) as well as the 16th and 84th percentiles (red) from the observations. The observational average and 16th and 84th percentiles were calculated using the filtered summer days (54% of total days) from July and August 2009–11 as described above. Figure 8 shows the model wind direction (green) compared to the average wind direction at that station (blue), calculated by averaging the zonal u and meridional υ components.

Fig. 7.
Fig. 7.

Comparison of model wind speed (green) to observed average (blue) and 16th and 84th percentiles (red).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Fig. 8.
Fig. 8.

Comparison of model (green) and observed (blue) wind direction.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

In general, model winds are highly realistic. Along the coast, simulated winds are close to the observed, although winds are ~2 m s−1 stronger than observed late in the forecast period at KHQM and Newport, Oregon (KONP), with a wind direction error of ~20° at KHQM. At Astoria, Oregon (KAST), the model winds are slightly weaker than observed in the evening and early morning, with realistic model wind directions for the full 24-h period. The model wind speeds and timing of the strong northwesterly winds in the Strait of Juan de Fuca are highly realistic, as shown by the wind speed and direction at CWQK (Figs. 7 and 8).

Over the western interior, observed and modeled wind directions at KSEA veer from westerly to northerly between the afternoon and evening, and model wind speeds are weaker than observed by about 1 m s−1. At night, when the winds are weaker at KSEA, there is less agreement. In the Willamette Valley, model-simulated northwesterly winds at Portland, Oregon (KPDX) and Eugene, Oregon (KEUG) are close to observed, and the moderate northerly flow that occurs during the afternoon and evening on the east side of the Cascade Mountains near Bend, Oregon (KBDN) is reproduced well. Despite the complex terrain surrounding KDLS and KELN, the intensity and timing of the peak winds at both KELN and KDLS closely match the observations. Farther east, model winds at Walla Walla, Washington (KALW); Spokane, Washington (KGEG); and Baker City, Oregon (KBKE) resemble average summer diurnal winds at these stations. Finally, model winds at Klamath Falls, Oregon (KLMT), are close to the observational average.

Figure 9 shows model-simulated 10-m winds over northwest Washington at 1800 LDT compared to a figure from Mass (1982) that shows the observed average wind vectors at the same time. As observed, the model simulation shows northerly flow of about 4–7 m s−1 over the central sound and westerly (easterly) flow on the east (west) of the sound. Furthermore, there is strong westerly flow in the Strait of Juan de Fuca in both simulated and observed wind fields. South of the Olympic Mountains, moderate westerlies (4–6 m s−1) prevail in the Chehalis Gap in both the simulated and observed maps.

Fig. 9.
Fig. 9.

(top) Vector-average surface wind for 3 yr during July (Mass 1982) and (bottom) 10-m wind speed from the WRF simulation (vectors and color shading), valid at 1800 LDT.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

The WRF run realistically simulates the diurnal temperature cycle over the region (Fig. 10). Near the coast, where the diurnal range is reduced by the influence of the Pacific, the model accurately simulates the diurnal cycle. At KCLM, in the Strait of Juan de Fuca, the model temperatures are higher than observed by about 2°–3°C for most of the forecast period. Simulated temperatures away from the coast are realistic. Simulated morning temperatures at KBKE, in eastern Oregon, are 4°–5°C warmer than observed, though the model daytime temperatures are much closer to the observed results.

Fig. 10.
Fig. 10.

As in Fig. 7, but for temperature.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Generally, most stations, especially away from the coast, manifest a diurnal cycle in relative humidity that resembles the observations (Fig. 11). Along the coast, model relative humidity is higher than observed by about 5%–15% for KHQM, KAST, KONP, and KUIL (Quillayute, Washington). Relative humidity in the Willamette Valley reaches a minimum earlier in the afternoon than observed (see KPDX and KEUG), and this deficiency also appears at two stations east of the Cascade crest (KELN and KDLS). The model relative humidity east of the Cascades is more realistic than on the west side, with the exception of the morning hours at KBKE, where model temperatures did not cool sufficiently in the morning.

Fig. 11.
Fig. 11.

As in Fig. 7, but for relative humidity.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

In summary, the verification results indicate that the WRF run at high resolution is able to simulate the diurnal cycle of all major variables over the Pacific Northwest with substantial fidelity to observations. With these verification results in hand, the remainder of the paper utilizes the WRF simulation to describe the three-dimensional diurnal circulations over the Pacific Northwest.

4. Results

a. 2-m temperature and relative humidity

The -km simulations of 925-hPa geopotential height, 10-m wind, and 2-m temperature for forecast hours 12, 18, 24, and 30 are presented in Fig. 12. High pressure dominates offshore and lower pressure occurs inland, with the height gradients being strongest at 1700 LDT due to pressure falls accompanying solar heating east of the Cascades. The impact of daytime heating is still evident at 2300 LDT, particularly over the basin of eastern Washington and the eastern Oregon highlands.

Fig. 12.
Fig. 12.

The 925-hPa geopotential height (black contours), temperature (color shading; °C), and wind (full barb = 5 m s−1).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

The diurnal range of 2-m temperature over the forecast period (12–35 h into the simulation) is shown in Fig. 13. The range, which varies from 1° to 20°C, was calculated by subtracting the minimum temperature from the maximum temperature at each grid point over the simulated diurnal period. Offshore and coastal areas have little variation in temperature, as do higher-terrain locations in the Cascades, Olympics, and Coast Range. A large range in temperature is noted east of the Cascade crest, such as the Oregon Plateau and the Columbia basin, and in low-lying areas west of the Cascades that are shielded from the influence of the ocean, such as the Willamette Valley.

Fig. 13.
Fig. 13.

Diurnal range of 2-m temperature (°C).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

The diurnal variations in temperature and wind have a profound effect on relative humidity over the region. The simulated evolution of 2-m relative humidity over the Pacific Northwest is shown in Fig. 14. At 0500 LDT, dry conditions occur east of the Cascades, while much higher relative humidity occurs over the Pacific and the western lowlands. There are limited areas east of the Cascades where relative humidity values are greater than 50%, such as the eastern Columbia basin and south-central Oregon near Klamath Falls. At 1100 LDT, relative humidity diminishes at most locations due to diurnal heating, except along the western slopes of the Cascades, where upslope flow leads to higher values. At 1700 LDT, most areas have lower relative humidity values except for the Cascade crest and the tops of the Olympics, where converging daytime upslope flow often leads to enhanced relative humidity and cumulus convection during the summer. At 2300 LDT and into the morning hours (0500 LDT), most areas have an increase in relative humidity as surface temperatures cool. An interesting exception is at the tops of the Olympic Mountains and portions of the north Cascades where relative humidity drops dramatically between 1700 and 0500 LDT as the upslope flow weakens and reverses.

Fig. 14.
Fig. 14.

Evolution of 2-m relative humidity from the WRF run.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Figure 15 shows the range of simulated 2-m relative humidity over the region. Coastal relative humidity varies little, as proximity to the ocean keeps the temperatures and relative humidity stable. Mountain slopes tend to have a small range in relative humidity, while western inland areas away from water (Willamette Valley, lower Columbia basin) have larger ranges. Central Oregon has the highest range of relative humidity (50% and more), which is confirmed by observations at KBDN (Fig. 6 shows location) and near the Oregon–California border at KLMT. This large range in relative humidity is likely due to the large diurnal range in temperature, as shown in Fig. 13.

Fig. 15.
Fig. 15.

Range of 2-m relative humidity (%).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

b. 10-m wind

The evolution of 10-m wind using the WRF simulation is presented in Fig. 16. Offshore, winds increase by 1–3 m s−1 during the day, remaining northerly or northwesterly throughout the diurnal cycle. The strongest winds are over the southern Oregon coastal waters, reaching 12–14 m s−1. Over the Strait of Juan de Fuca, the winds are weakest at 1100 LDT, although they are still moderately strong at 5–6 m s−1 from the northwest. The strait winds reach a peak at ~2200 LDT with northwesterly wind speeds of 12–13 m s−1. Over the western interior of Oregon and Washington, the winds are generally weakest at 0500–0800 LDT. Westerlies associated with the Pacific Ocean sea breeze pass through the gaps in the Coast Range during the day, reaching a peak near 1700 LDT, and weaken during the evening and overnight. In the Willamette Valley, the winds are weak in the morning, with northwest-to-north winds strengthening in the afternoon and evening, reaching a peak of 3–6 m s−1 around 2000 LDT before weakening during the night.

Fig. 16.
Fig. 16.

Simulated wind at 10-m (color shading and vectors; m s−1).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Westerlies passing through the Columbia River Gorge, the only near-sea-level gap through the Cascade Mountains, reach a maximum between 1700 and 2000 LDT. Winds extending out of Snoqualmie Pass and Stevens Pass into central Washington reach a maximum a few hours later, between 2000 and 2300 LDT. Over the western Columbia basin, winds are light in the morning and reach a maximum in the late evening as westerlies extend down the eastern slopes of the Cascades.

Weak southwesterly upslope flow forms in the late morning over the eastern Columbia basin at 1100 LDT and continues into the afternoon and evening. Over the Oregon Plateau, winds are weak in the morning hours. Northwesterly upslope flow strengthens over north-central Oregon between 1400 and 1700 LDT and is strongest in the southern portion of the plateau at 2000–2300 LDT.

Figure 17 shows the hour at which the maximum 10-m wind is reached for each grid point over Oregon and Washington, with a terrain map for reference. The most common colors are blues, purples, and pinks, indicating maximum winds during the afternoon or evening. Lighter blues, which are mostly found on mountain slopes, such as along the eastern slopes of the Olympic Mountains or portions of the western slopes of the Cascades, indicate maximum winds during the early afternoon hours, consistent with upslope flow. Darker blues indicate that peak winds occur in late afternoon and are found offshore and along the western slopes of the coastal and Cascade Mountains. The winds over the lowland areas near Puget Sound and in the Willamette Valley, as well as a large part of eastern Oregon, peak around 2000–2100 LDT, as indicated by purple and pink. Finally, reds indicate maxima around 2200–2400 LDT over the eastern Strait of Juan de Fuca, the lower Columbia basin, and over some of the eastern Oregon highlands. Interestingly, an area of green color (0600–1000 LDT) is found over a portion of eastern Washington, which is consistent with an observed peak wind at Pullman, Washington (KPUW; not shown) at the same time, though a slightly stronger peak in wind speed is reached at KPUW in the late afternoon. There is little diurnal variation in wind speed (~1 m s−1) at KPUW and KGEG, thus rendering the area of green to be less meaningful than other areas of the figure.

Fig. 17.
Fig. 17.

(left) Hour at which the max wind is reached (LDT) and (right) terrain height (m).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Figure 18 shows the diurnal range of the u (top) and the υ (bottom) components of the simulated 10-m wind. West of the Cascades, most of the variation in the u component is over the eastern Strait of Juan de Fuca and the central Strait of Georgia. In western Oregon, there is a moderate range (4–8 m s−1) in the u component within the gaps of the Coast Range. Over the Puget Sound there is a moderate range in the υ-wind component associated with the sound breeze, with weak southerlies in the morning and moderate northerlies in the afternoon. There is little variation in the 10-m υ component offshore, despite speeds of greater than 10 m s−1 over the coastal waters of southern Oregon.

Fig. 18.
Fig. 18.

Diurnal range of u- and υ- components of 10-m wind (m s−1).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

The largest variations in both components, up to 12 m s−1 in some areas, occur east of the Cascades. In the morning, weak easterly upslope flow develops along the eastern slopes of the Washington Cascades. Westerly winds of 6–10 m s−1 develop in the evening, thus explaining the large range in the u component in that area. There is also a large variation (6–12 m s−1) in the u and υ components in eastern Oregon, due to strong nighttime downslope flow into the Snake River valley of Idaho. In central Oregon, there is a large variation (6–10 m s−1) in the υ component as northerlies develop during the day and reverse to weak southerlies at night. This diurnal variation is verified by observations at KBND (Figs. 7 and 8).

c. Lower-tropospheric wind

To provide a three-dimensional view of the evolution of the regional diurnal circulations, Fig. 19 shows winds at 975, 925, 875, and 825 hPa, with no winds shown if the terrain is above the identified pressure level. Over the coastal waters, winds are relatively constant in time at all levels, with speed decreasing with height.

Fig. 19.
Fig. 19.

Wind (colors and vectors) with terrain above the pressure level blocked out.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

In contrast, over the western lowlands of Oregon and Washington, the winds vary substantially in time at all levels. At 1400 LDT, winds are from the north to northwest over western Oregon and Washington and weaken with height. These winds increase in the evening at lower levels and remain weak at 825 hPa. Interestingly, several locations over the Willamette Valley and over the Chehalis Gap (immediately south of the Olympics) develop strong winds aloft at night (975–925 hPa), reaching a peak at 2300 LDT. Winds in the lower troposphere over the Strait of Juan de Fuca reach a maximum at 2000–2300 LDT, with the strongest westerlies at 975 hPa.

In the Cascades, strong westerlies develop in the afternoon within the gaps such as the Columbia River Gorge and the Stampede–Snoqualmie Pass gap. Near and east of the gap exits, winds strengthen during the evening and reach a maximum near 2300 LDT, well after the hours of peak heating.

A major feature in the simulations, yet not discussed in previous literature, is the diurnal northerly flows over the highlands of eastern Oregon. Upslope flow is evident on the slopes of the higher terrain in north-central Oregon at 1700 LDT at 925, 875, and 825 hPa. At 2000 LDT, northerlies at 825 hPa strengthen dramatically and move southward, and strengthen further at 2300 LDT, after which the winds weaken into the morning hours. Interestingly, this feature is barely noticeable at the surface (Fig. 16).

To determine whether the diurnal northerly flow over the Oregon Plateau is realistic, the wind speed and direction at two stations on the plateau (ATFO3 and WTFO3) were compared to the model simulations (Fig. 20). These sites are at the top of mountain peaks, about 500 m above the plateau floor, and are Remote Automated Weather Stations (RAWSs) run by the Bureau of Land Management and the U.S. Forest Service (locations shown in Fig. 6). For much of the morning and afternoon, model wind speeds underestimate the strength of the winds. However, both observations and the model simulations show maximum wind speeds between 2100 and 2300 LDT, and the observed wind direction agrees with the model during the period of strongest winds.

Fig. 20.
Fig. 20.

Comparison of model (green) and observational (blue) wind speed and direction. Wind speeds shown as in Fig. 7.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

To further analyze the three-dimensional structure of the regional wind features, vertical cross sections were created, the locations of which are shown in Fig. 21. Figure 22 shows cross sections over the northern Willamette Valley at 1600 and 2200 LDT. At 1600 LDT, potential temperatures indicate a mixed layer up to ~900 hPa, with weak-to-moderate northwesterly winds throughout the lower atmosphere. Six hours later, northwesterly lower-tropospheric winds strengthen dramatically as the daytime boundary layer collapses and stability increases at the surface.

Fig. 21.
Fig. 21.

Locations of cross sections for Figs. 22 (short line) and 23 (long line).

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Fig. 22.
Fig. 22.

Wind (colors and barbs; full barb = 5 m s−1) and potential temperature (blue lines; K) from the -km WRF simulations at (left) 1600 and (right) 2200 LDT.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

To evaluate the realism of these circulations over the northern Willamette Valley, observations from the Aircraft Communications and Reporting System (ACARS) for aircraft landing–taking off from Portland International Airport (PDX) were examined for typical summer days as defined earlier (Table 1). Winds are weak below 1000 m during the afternoon of 21 July 2011, while during the evening, aircraft measured a low-level jet maximum of 10 m s−1 at 435 m above ground. During the afternoon of 29 July 2011, low-level winds were weak but strengthened that evening to 10 m s−1 around 600 m. Reviewing ACARS data for many days and over several summer seasons revealed that an evening wind maximum is a common feature in the summer over Portland, Oregon, thus confirming the model simulation.

Table 1.

ACARS data for four flights during July 2011.

Table 1.

East–west cross sections of the simulated wind and thermal fields over Washington at two different times are shown in Fig. 23; the cross section location is shown in Fig. 21. At 1600 LDT, northerly flow and stable conditions are evident off the coast of Washington, with a strong horizontal temperature gradient over the coastal region. Daytime boundary layer mixing over western Washington appears to prevent stronger winds over the coastal lowlands despite the large temperature gradient. Six hours later during the evening (2200 LDT), more stable conditions prevail over the lowest 50 hPa of western Washington and strong northwesterlies develop with the stable layer. At this time there is still a modest temperature gradient over the coastal lowlands.

Fig. 23.
Fig. 23.

(left) Wind (colors and barbs; full barb = 5 m s−1) and potential temperature (blue lines; K), and (right) temperature (°C) from the -km WRF simulation.

Citation: Weather and Forecasting 29, 5; 10.1175/WAF-D-14-00018.1

Figure 23 shows that over the Cascade crest, moderate westerlies occur at 1600 LDT in the presence of a strong temperature gradient between the west and east sides of the barrier. Weak winds and a well-mixed boundary layer exist east of the Cascades. Six hours later, strong westerly downslope winds were evident over the eastern slopes of the Cascades, where substantial low-level cooling has occurred. Data from a 400-ft tower at Hanford confirm the realism of the simulated downslope westerly flow, with a wind maximum occurring on average at 2300 LDT at that location (not shown).

5. Discussion and summary

During the summer over the northwest United States, strong surface heating combines with the terrain and land–water contrasts to create diurnal circulations that develop and interact on a variety of scales. This paper provides a three-dimensional description of these complex diurnal circulations over the Pacific Northwest, making use of high-resolution simulations. For this work, GFS model output (3 hourly) was averaged for July and August 2009–11 and used for initial and boundary conditions in a WRF run. This simulation was examined to explore the three-dimensional evolution of winds, temperature, and relative humidity over the region for a typical summer diurnal period.

To analyze the validity of the simulation, the WRF run was compared to a collection of days with typical summer conditions. Wind direction at five stations was used to filter out summer days with atypical synoptic conditions, leaving 54% of the available 186 days (from July and August 2009–11) to compare to the WRF simulation. It was found that the model simulations produced realistic diurnal variations in wind speed, wind direction, relative humidity, and temperature.

The summer synoptic environment over the northwest United States has a large impact on the mesoscale diurnal variations over the region. The east Pacific high produces northerlies along the coast, with cool coastal sea surface temperatures from upwelling, and pushes cool marine air into the western lowlands. During the day, pressure lowers east of the Cascades from solar heating, enhancing the east–west pressure gradient, particularly across the Cascade Mountains (Fig. 12). This gradient contributes to afternoon and early evening westerlies within gaps in the coastal lowlands and the Cascade Mountains.

West of the Cascades, 10-m wind simulations show that sea-breeze westerlies reach a maximum intensity near 1700 LDT, particularly within the gaps of the Coast Range, such as the Chehalis Gap south of the Olympic Mountains. Within the Willamette Valley, north-to-northwesterly flow develops and is enhanced in the afternoon, reaching a peak near 1900 LDT. In the Puget Sound region, simulations confirm the results of Mass (1982) and Staley (1957), which show moderate northerly flow in the afternoon that weakens at night. Within the Strait of Juan de Fuca, strong terrain-channeled northwesterly flow occurs throughout the day, reaching a maximum in the evening near 2200 LDT.

Within the Cascades, surface westerly flow through the near-sea-level Columbia River Gorge is strongest between 1700 and 2000 LDT. In contrast, the gorge winds at ~925 hPa are stronger later in the evening (2000–2300 LDT). An evening wind maximum is also apparent for the downslope wind maximum east of the Snoqualmie and Stevens Passes in central Washington. Doran and Zhong (1994) described the strong northwesterly winds pushing down the eastern slopes of the Cascades as a combination of katabatic flow and cooler air coming through the pass from the west side.

The WRF simulations and confirming observations showed that evening wind maxima occur above the surface over the northern Willamette Valley and the Chehalis Gap, reaching maximum strength near 2300 LDT at approximately 975 hPa. These low-level wind maxima are reminiscent of low-level nocturnal jets/wind maxima that occur all over the world (ReVelle and Nilsson 2008; Baas et al. 2009; Karipot et al. 2009; Kumar et al. 2012), and it has been shown that WRF can realistically simulate these wind phenomena (Storm and Basu 2010; Michelson et al. 2010; Colle and Novak 2010). During the day, horizontal winds often weaken in the boundary layer due to the vertical mixing of air slowed by surface drag. As the surface cools during the evening, a low-level inversion or stable layer forms, inhibiting vertical motion and decoupling the lower troposphere from the surface. Such decoupling from surface drag allows existing pressure–temperature gradients to accelerate air above the surface to form low-level wind maxima (Arya 2001); such evening gradients were evident in the simulations (Fig. 23).

In the case of the well-documented low-level jet in the Great Plains, as well as other larger-scale nocturnal low-level jets over relatively flat terrain, there is a balance between friction, the Coriolis force, and the pressure gradient force within the boundary layer during the day. In the evening as the near-surface stable layer forms, the loss of drag results in an inertial oscillation that drives supergeostrophic low-level winds (Markowski and Richardson 2010). However, because of the smaller scale of the low-level wind features over the Willamette Valley and Chehalis Gap and their relatively short longevity, the Coriolis force is small and thus accelerations of the wind are primarily due to the pressure gradient force.

Over the Oregon Plateau, the simulations indicated strong northerly flow during the evening. These winds are hardly noticeable at the surface, but are found immediately aloft (lowest 50 hPa) above the surface with wind speeds near 10–12 m s−1. Such Oregon Plateau northerlies are confirmed by observations at the tops of local mountain peaks. It appears that this northerly flow is driven by strong temperature–pressure gradients that develop during the day. These gradients are able to accelerate low-level flow during the evening when the drag lessens as increased stability decouples the lower atmosphere from the surface.

An important aspect of the diurnal winds over the region is that diurnal circulations of various scales interact simultaneously over the region, producing a complex and highly three-dimensional flow and thermal evolution. There are local sea-breeze circulations between the Pacific Ocean and the coastal lowlands, regional sea-breeze winds between the western Washington interior landmass and the cool waters of the Straits of Georgia and Juan de Fuca, diurnal circulations between the inland bodies of water and the surrounding land, and slope flows on the substantial terrain of the region. There are also diurnal circulations between the interior of eastern Washington and the western Washington lowlands, and between the heated interior and the Pacific Ocean, as well as regional diurnal flows over the eastern Oregon Plateau. Even more complexity is produced by the varying stability that results in a diurnal cycle of coupling–decoupling of the free atmosphere from the surface. The skill of modern high-resolution modeling systems in generally duplicating these complex, three-dimensional diurnal circulations is a testament to improvements in modeling systems during the past decades.

This paper described the regional diurnal wind features over the Pacific Northwest, and also briefly discussed some of individual circulations over the region. Further analysis, needed to describe the individual circulations in more detail, will be found in subsequent work.

Acknowledgments

This research was supported by USDA Forest Service Joint Venture Agreement PNW 10-JV-11261987-033, a GRIN fellowship from the Joint Fire Science Program administered by the Bureau of Land Management (Award L13AC00303), and the National Science Foundation (Award AGS-1041879).

REFERENCES

  • Arya, S. P., 2001: Introduction to Micrometeorology. 2nd ed. Academic Press, 418 pp.

  • Baas, P., Bosveld F. C. , Baltink H. K. , and Holtslag A. A. M. , 2009: A climatology of nocturnal low-level jets at Cabauw. J. Appl. Meteor. Climatol., 48, 16271642, doi:10.1175/2009JAMC1965.1.

    • Search Google Scholar
    • Export Citation
  • Bielli, S., Barbour P. , Samelson R. , Skyllingstad E. , and Wilczak J. , 2002: Numerical study of the diurnal cycle along the central Oregon coast during summertime northerly flow. Mon. Wea. Rev., 130, 9921008, doi:10.1175/1520-0493(2002)130<0992:NSOTDC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brewer, M. C., Mass C. F. , and Potter B. E. , 2012: The West Coast thermal trough: Climatology and synoptic evolution. Mon. Wea. Rev., 140, 38203843, doi:10.1175/MWR-D-12-00078.1.

    • Search Google Scholar
    • Export Citation
  • Burk, S. D., and Thompson W. T. , 1996: The summertime low-level jet and MBL structure along the California coast. Mon. Wea. Rev., 124, 668686, doi:10.1175/1520-0493(1996)124<0668:TSLLJA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chien, F. C., Mass C. F. , and Kuo Y. H. , 1997: Interaction of a warm-season frontal system with the Coastal Mountains of the western United States. Part I: Prefrontal onshore push, coastal ridging, and alongshore southerlies. Mon. Wea. Rev., 125, 17051729, doi:10.1175/1520-0493(1997)125<1705:IOAWSF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., and Novak D. R. , 2010: The New York Bight jet: Climatology and dynamical evolution. Mon. Wea. Rev., 138, 23852404, doi:10.1175/2009MWR3231.1.

    • Search Google Scholar
    • Export Citation
  • Doran, J. C., and Zhong S. , 1994: Regional drainage flows in the Pacific Northwest. Mon. Wea. Rev., 122, 11581167, doi:10.1175/1520-0493(1994)122<1158:RDFITP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Elliott, D. L., and O’Brien J. J. , 1977: Observational studies of the marine boundary layer over an upwelling region. Mon. Wea. Rev., 105, 8698, doi:10.1175/1520-0493(1977)105<0086:OSOTMB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Holt, T. R., 1996: Mesoscale forcing of a boundary layer jet along the California coast. J. Geophys. Res., 101, 42354254, doi:10.1029/95JD03231.

    • Search Google Scholar
    • Export Citation
  • Karipot, A., Leclerc M. Y. , and Zhang G. , 2009: Characteristics of nocturnal low-level jets observed in the north Florida area. Mon. Wea. Rev., 137, 26052621, doi:10.1175/2009MWR2705.1.

    • Search Google Scholar
    • Export Citation
  • Kumar, M. S., Anandan V. K. , Rao T. N. , and Reddy P. N. , 2012: A climatological study of the nocturnal boundary layer over a complex-terrain station. J. Appl. Meteor. Climatol., 51, 813825, doi:10.1175/JAMC-D-11-047.1.

    • Search Google Scholar
    • Export Citation
  • Markowski, P., and Richardson Y. , 2010: Mesoscale Meteorology in Midlatitudes. John Wiley and Sons, 407 pp.

  • Mass, C. F., 1982: The topographically forced diurnal circulations of western Washington State and their influence on precipitation. Mon. Wea. Rev., 110, 170183, doi:10.1175/1520-0493(1982)110<0170:TTFDCO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., and Bond N. A. , 1996: Coastally trapped wind reversals along the United States West Coast during the warm season. Part II: Synoptic evolution. Mon. Wea. Rev., 124, 446461, doi:10.1175/1520-0493(1996)124<0446:CTWRAT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., Albright M. D. , and Brees D. J. , 1986: The onshore surge of marine air into the Pacific Northwest: A coastal region of complex terrain. Mon. Wea. Rev., 114, 26022627, doi:10.1175/1520-0493(1986)114<2602:TOSOMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Michelson, S. A., Djalalova I. V. , and Bao J. , 2010: Evaluation of the summertime low-level winds simulated by MM5 in the Central Valley of California. J. Appl. Meteor. Climatol., 49, 22302245, doi:10.1175/2010JAMC2295.1.

    • Search Google Scholar
    • Export Citation
  • Nuss, W. A., and Coauthors, 2000: Coastally trapped wind reversals: Progress toward understanding. Bull. Amer. Meteor. Soc., 81, 719743, doi:10.1175/1520-0477(2000)081<0719:CTWRPT>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • ReVelle, D. O., and Nilsson E. D. , 2008: Summertime low-level jets over the high-latitude Arctic Ocean. J. Appl. Meteor. Climatol., 47, 17701784, doi:10.1175/2007JAMC1637.1.

    • Search Google Scholar
    • Export Citation
  • Staley, D. O., 1957: The low-level sea breeze of northwest Washington. J. Meteor., 14, 458470, doi:10.1175/1520-0469(1957)014<0458:TLLSBO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Staley, D. O., 1959: Some observations of surface-wind oscillations in a heated basin. J. Meteor., 16, 364370, doi:10.1175/1520-0469(1959)016<0364:SOOSWO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Storm, B., and Basu S. , 2010: The WRF model forecast-derived low-level wind shear climatology over the United States Great Plains. Energies, 3, 258276, doi:10.3390/en3020258.

    • Search Google Scholar
    • Export Citation
  • Taylor, S. V., Cayan D. R. , Graham N. E. , and Georgakakos K. P. , 2008: Northerly surface winds over the eastern North Pacific Ocean in spring and summer. J. Geophys. Res., 113, D02110, doi:10.1029/2006JD008053.

    • Search Google Scholar
    • Export Citation
  • Zemba, J., and Friehe C. A. , 1987: The marine boundary layer jet in the Coastal Ocean Dynamics Experiment. J. Geophys. Res., 92, 14891496, doi:10.1029/JC092iC02p01489.

    • Search Google Scholar
    • Export Citation
1

For several years, the University of Washington’s Department of Atmospheric Sciences has run WRF at 36-, 12-, 4-, and 1.3-km grid spacings.

2

For each station in Fig. 5, there are a few days when the wind speed is zero and hence would have no wind direction; those days are not represented in this figure. Bins of 10° were used for wind direction.

Save
  • Arya, S. P., 2001: Introduction to Micrometeorology. 2nd ed. Academic Press, 418 pp.

  • Baas, P., Bosveld F. C. , Baltink H. K. , and Holtslag A. A. M. , 2009: A climatology of nocturnal low-level jets at Cabauw. J. Appl. Meteor. Climatol., 48, 16271642, doi:10.1175/2009JAMC1965.1.

    • Search Google Scholar
    • Export Citation
  • Bielli, S., Barbour P. , Samelson R. , Skyllingstad E. , and Wilczak J. , 2002: Numerical study of the diurnal cycle along the central Oregon coast during summertime northerly flow. Mon. Wea. Rev., 130, 9921008, doi:10.1175/1520-0493(2002)130<0992:NSOTDC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brewer, M. C., Mass C. F. , and Potter B. E. , 2012: The West Coast thermal trough: Climatology and synoptic evolution. Mon. Wea. Rev., 140, 38203843, doi:10.1175/MWR-D-12-00078.1.

    • Search Google Scholar
    • Export Citation
  • Burk, S. D., and Thompson W. T. , 1996: The summertime low-level jet and MBL structure along the California coast. Mon. Wea. Rev., 124, 668686, doi:10.1175/1520-0493(1996)124<0668:TSLLJA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chien, F. C., Mass C. F. , and Kuo Y. H. , 1997: Interaction of a warm-season frontal system with the Coastal Mountains of the western United States. Part I: Prefrontal onshore push, coastal ridging, and alongshore southerlies. Mon. Wea. Rev., 125, 17051729, doi:10.1175/1520-0493(1997)125<1705:IOAWSF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., and Novak D. R. , 2010: The New York Bight jet: Climatology and dynamical evolution. Mon. Wea. Rev., 138, 23852404, doi:10.1175/2009MWR3231.1.

    • Search Google Scholar
    • Export Citation
  • Doran, J. C., and Zhong S. , 1994: Regional drainage flows in the Pacific Northwest. Mon. Wea. Rev., 122, 11581167, doi:10.1175/1520-0493(1994)122<1158:RDFITP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Elliott, D. L., and O’Brien J. J. , 1977: Observational studies of the marine boundary layer over an upwelling region. Mon. Wea. Rev., 105, 8698, doi:10.1175/1520-0493(1977)105<0086:OSOTMB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Holt, T. R., 1996: Mesoscale forcing of a boundary layer jet along the California coast. J. Geophys. Res., 101, 42354254, doi:10.1029/95JD03231.

    • Search Google Scholar
    • Export Citation
  • Karipot, A., Leclerc M. Y. , and Zhang G. , 2009: Characteristics of nocturnal low-level jets observed in the north Florida area. Mon. Wea. Rev., 137, 26052621, doi:10.1175/2009MWR2705.1.

    • Search Google Scholar
    • Export Citation
  • Kumar, M. S., Anandan V. K. , Rao T. N. , and Reddy P. N. , 2012: A climatological study of the nocturnal boundary layer over a complex-terrain station. J. Appl. Meteor. Climatol., 51, 813825, doi:10.1175/JAMC-D-11-047.1.

    • Search Google Scholar
    • Export Citation
  • Markowski, P., and Richardson Y. , 2010: Mesoscale Meteorology in Midlatitudes. John Wiley and Sons, 407 pp.

  • Mass, C. F., 1982: The topographically forced diurnal circulations of western Washington State and their influence on precipitation. Mon. Wea. Rev., 110, 170183, doi:10.1175/1520-0493(1982)110<0170:TTFDCO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., and Bond N. A. , 1996: Coastally trapped wind reversals along the United States West Coast during the warm season. Part II: Synoptic evolution. Mon. Wea. Rev., 124, 446461, doi:10.1175/1520-0493(1996)124<0446:CTWRAT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., Albright M. D. , and Brees D. J. , 1986: The onshore surge of marine air into the Pacific Northwest: A coastal region of complex terrain. Mon. Wea. Rev., 114, 26022627, doi:10.1175/1520-0493(1986)114<2602:TOSOMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Michelson, S. A., Djalalova I. V. , and Bao J. , 2010: Evaluation of the summertime low-level winds simulated by MM5 in the Central Valley of California. J. Appl. Meteor. Climatol., 49, 22302245, doi:10.1175/2010JAMC2295.1.

    • Search Google Scholar
    • Export Citation
  • Nuss, W. A., and Coauthors, 2000: Coastally trapped wind reversals: Progress toward understanding. Bull. Amer. Meteor. Soc., 81, 719743, doi:10.1175/1520-0477(2000)081<0719:CTWRPT>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • ReVelle, D. O., and Nilsson E. D. , 2008: Summertime low-level jets over the high-latitude Arctic Ocean. J. Appl. Meteor. Climatol., 47, 17701784, doi:10.1175/2007JAMC1637.1.

    • Search Google Scholar
    • Export Citation
  • Staley, D. O., 1957: The low-level sea breeze of northwest Washington. J. Meteor., 14, 458470, doi:10.1175/1520-0469(1957)014<0458:TLLSBO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Staley, D. O., 1959: Some observations of surface-wind oscillations in a heated basin. J. Meteor., 16, 364370, doi:10.1175/1520-0469(1959)016<0364:SOOSWO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Storm, B., and Basu S. , 2010: The WRF model forecast-derived low-level wind shear climatology over the United States Great Plains. Energies, 3, 258276, doi:10.3390/en3020258.

    • Search Google Scholar
    • Export Citation
  • Taylor, S. V., Cayan D. R. , Graham N. E. , and Georgakakos K. P. , 2008: Northerly surface winds over the eastern North Pacific Ocean in spring and summer. J. Geophys. Res., 113, D02110, doi:10.1029/2006JD008053.

    • Search Google Scholar
    • Export Citation
  • Zemba, J., and Friehe C. A. , 1987: The marine boundary layer jet in the Coastal Ocean Dynamics Experiment. J. Geophys. Res., 92, 14891496, doi:10.1029/JC092iC02p01489.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    The terrain and major geographical features of the northwest United States.

  • Fig. 2.

    (left) Average 925-hPa geopotential height (m; contours) and temperature (°C; color shading), and (right) 10-m wind (vectors and color shading) from WRF output for July and August 2009–11.

  • Fig. 3.

    Histograms of daily max temperatures for July and August 2007–12 at (left) KSEA and (right) HHMS.

  • Fig. 4.

    Domains for the WRF simulations.

  • Fig. 5.

    Histograms of wind direction for July and August 2009–11 at the time of the climatological max wind speed (noted at the top of each graph) for several regional observing sites. The percentage of days within the primary mode (red color) is shown.

  • Fig. 6.

    Map of stations used for verification.

  • Fig. 7.

    Comparison of model wind speed (green) to observed average (blue) and 16th and 84th percentiles (red).

  • Fig. 8.

    Comparison of model (green) and observed (blue) wind direction.

  • Fig. 9.

    (top) Vector-average surface wind for 3 yr during July (Mass 1982) and (bottom) 10-m wind speed from the WRF simulation (vectors and color shading), valid at 1800 LDT.

  • Fig. 10.

    As in Fig. 7, but for temperature.

  • Fig. 11.

    As in Fig. 7, but for relative humidity.

  • Fig. 12.

    The 925-hPa geopotential height (black contours), temperature (color shading; °C), and wind (full barb = 5 m s−1).

  • Fig. 13.

    Diurnal range of 2-m temperature (°C).

  • Fig. 14.

    Evolution of 2-m relative humidity from the WRF run.

  • Fig. 15.

    Range of 2-m relative humidity (%).

  • Fig. 16.

    Simulated wind at 10-m (color shading and vectors; m s−1).

  • Fig. 17.

    (left) Hour at which the max wind is reached (LDT) and (right) terrain height (m).

  • Fig. 18.

    Diurnal range of u- and υ- components of 10-m wind (m s−1).

  • Fig. 19.

    Wind (colors and vectors) with terrain above the pressure level blocked out.

  • Fig. 20.

    Comparison of model (green) and observational (blue) wind speed and direction. Wind speeds shown as in Fig. 7.

  • Fig. 21.

    Locations of cross sections for Figs. 22 (short line) and 23 (long line).

  • Fig. 22.

    Wind (colors and barbs; full barb = 5 m s−1) and potential temperature (blue lines; K) from the -km WRF simulations at (left) 1600 and (right) 2200 LDT.

  • Fig. 23.

    (left) Wind (colors and barbs; full barb = 5 m s−1) and potential temperature (blue lines; K), and (right) temperature (°C) from the -km WRF simulation.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 716 448 130
PDF Downloads 206 51 5