1. Introduction
Understanding the winds within a valley and their interactions with the larger-scale forcings is of interest for a number of reasons. For example, the dispersion of pollutants in a valley depends strongly on local valley circulations (e.g., Whiteman 1989; Fast et al. 2006); nocturnal minimum surface temperatures depend strongly on the near-surface wind speed (e.g., Estournel and Guedalia 1985; Steeneveld et al. 2006) and hence on the strength of the valley wind; land surface–atmosphere exchanges over mountainous regions are closely linked to slope and valley flows; and the effects of these flows on mesoscale fluxes need to be parameterized in numerical weather prediction and climate models (e.g., Noppel and Fiedler 2002; Weigel et al. 2007; Rotach and Zardi 2007).
Three major mechanisms that can produce within-valley winds are thermal forcing, pressure-driven channeling, and downward momentum transport (Whiteman and Doran 1993). Thermal forcing refers to valley winds that are generated by locally developed along-valley pressure gradients produced hydrostatically from temperature differences along the valley’s axis, and between the valley and the adjacent plains (e.g., Wagner 1938; Vergeiner and Dreiseitl 1987; Whiteman 1990; Rampanelli et al. 2004). Such forcings produce upvalley winds during the day and downvalley winds at night. These winds are independent from the above-valley wind direction. They occur frequently in areas with large diurnal cycles in surface sensible heat fluxes. Along-valley and valley–plain temperature differences can be produced by at least three factors: radiative differences, differences in the conversion rate of net radiation to sensible heat flux, and by the valley volume effect (Whiteman 2000). While the first two factors are attributed primarily to differences in land surface properties, they may also be influenced by the thermally induced flows themselves (De Wekker et al. 1998). Much attention has been given to the valley volume effect, which can be quantified in terms of a topographic amplification factor (TAF) (e.g., Wagner 1938; Steinacker 1984; McKee and O’Neil 1989). According to the volume argument, energy fluxes into the valley atmosphere are distributed through a smaller volume than over the plain, amplifying the temperature change in the valley volume relative to the plain. The basic premise of the volume argument is that the control volume is thermodynamically closed. Several studies (Kuwagata and Kimura 1995, 1997; Rampanelli et al. 2004; Weigel et al. 2006) have, however, pointed to the role of thermally induced cross-valley circulations, which extend well above the valley top, in heating the valley atmosphere.
Pressure-driven channeling traditionally refers to a valley wind forced by the along-valley component of the large-scale geostrophic pressure gradient. This mechanism was initially proposed by Fiedler (1983) and was used by Gross and Wippermann (1987) to explain the frequent along-valley winds in Germany’s shallow and wide Rhine Valley. Because of the 90° offset between the geostrophic wind direction and the geostrophic pressure gradient, pressure-driven channeling can produce a valley wind blowing in the opposite direction of the along-valley component of the above-valley wind. In the north–south-oriented Rhine Valley, for example, a geostrophic northwesterly flow aloft may induce a southerly flow within the valley. Pressure-driven channeling in curved valleys has been investigated by Kossmann and Sturman (2003).
Downward momentum transport, or forced channeling, refers to the downward transport of horizontal momentum from the above-valley synoptic wind (e.g., Banta and Cotton 1981; Banta 1986). In a wide and shallow valley, this mechanism may lead to a valley wind similar to the geostrophic wind direction aloft. In a narrow and deep valley, this mechanism may produce winds aligned with the valley axis, the direction being determined by the along-valley component of the wind aloft.
These mechanisms interact to produce complex flow patterns that depend on the geometry of the valley, land surface properties, the thermal structure of the atmosphere, the magnitude and spatial distribution of radiative forcing, and the larger-scale synoptic situation. Several studies (e.g., Barr and Orgill 1989; Doran 1991; Gudiksen et al. 1992; Coulter and Gudiksen 1995) have documented the effects of ambient winds, atmospheric stability, cloudiness, and atmospheric moisture on the structure and strength of nocturnal drainage flows. The nonlinear interaction between nocturnal drainage flows and mountain waves has been investigated by Poulos et al. (2000, 2007). The ability of high-resolution numerical models to simulate these thermally induced flows over complex terrain has been demonstrated by several studies in recent years. Zängl et al. (2001), for example, investigated the very strong daytime valley winds in the Kali Gandaki Valley in Nepal (Egger et al. 2000), Zhong and Fast (2003) presented a model intercomparison for flows in the Salt Lake Valley and compared the simulations with measurements from the Vertical Transport and Mixing Experiment (VTMX) campaign (Doran et al. 2002), and De Wekker et al. (2005) used high-resolution simulations to investigate the convective boundary layer structure in the Riviera Valley in Switzerland.
Complex nighttime valley flow structures of thermally driven valley flows were observed in California’s Owens Valley during the Terrain-Induced Rotor Experiment (T-REX). In one case the flow in the north–south-oriented valley was characterized by a three-layer structure with a midlevel (2–2.5 km MSL) upvalley flow after midnight local time, sandwiched between a low-level northerly downvalley flow and an above-ridge-top northwesterly flow. The second case was characterized by a strong and deep downvalley jet, with average wind speeds exceeding 15 m s−1 for several hours and with weak winds aloft. The difference in valley flow structure between these two cases is surprisingly large considering the relatively similar upper-level (500 hPa) synoptic forcing with westerly to northwesterly flow in both cases.
The aim of this paper is to document the nighttime flow structure and evolution in the Owens Valley and to investigate the physical mechanisms leading to the differing flow evolution for the two cases. In our approach, we combine measurements from the T-REX field campaign (Grubisic et al. 2008) with the output of large-eddy simulations (LES). For the modeling, we use the Advanced Regional Prediction System (ARPS; Xue et al. 2000, 2001, 2003). The model has been used successfully by Chow et al. (2006) and Weigel et al. (2006) to study the daytime flow evolution in the medium-sized Riviera Valley, and by Smith and Skyllingstad (2005) for the simulation of nocturnal katabatic flows in an idealized configuration. For this study, the model setup is similar to that used by Chow et al. (2006).
2. Experiment description
a. T-REX and the Owens Valley
A comprehensive description of the T-REX field campaign can be found in Grubisic et al. (2008). While a major focus of the T-REX field campaign was the evolution of lee waves and rotors under strongly forced conditions, another related objective was to study the structure and evolution of the complex terrain boundary layer under relatively clear and quiescent conditions. The observational campaign was carried out in Owens Valley (Fig. 1a) from 1 March to 30 April 2006. During this 2-month period, 15 intensive observation periods (IOPs), for observing mountain wave/rotor/boundary layer interactions and 5 enhanced observing periods (EOPs) for documenting the nonrotor mountain boundary layer evolution were carried out.
Owens Valley has a length of about 150 km, a base width of about 15–30 km, and an upvalley orientation from approximately south-southeast to north-northwest. For the most part the slope of the valley floor is very small (slope angle < 0.1°). The lowest point in the valley is Owens Lake, which has an elevation of 1086 m MSL.1 The valley floor rises to 1150–1200 m in Haiwee in the south and to about 1150 m in Independence and 1250 m in Bishop in the north. Thus, within the lowest 100 m above ground Owens Valley is a terminal basin. This may influence the near-surface flow in the Independence area, but not the higher levels considered in this paper. While the surrounding peaks of the Sierra Nevada mountain range to the west and the White Mountains to the northeast reach more than 4000 m, the peaks of the Inyo Mountains in the southeast are about 3000 m high. Because of the semiarid climate a large fraction of the net radiation is converted to sensible heat, which can lead to strong thermally induced flows.
The dataset obtained during the EOPs includes radio soundings, remote sensing measurements from wind profilers, radio acoustic sounding systems (RASS), aerosol and Doppler lidars, sonic and profile measurements from three 30-m and several smaller flux towers, and standard meteorological observations from an array of surface stations. Most systems were located along two valley transects just south of Independence. In this study, the southernmost wind profiler/RASS system and the 1.5-hourly radiosonde ascents from Independence airport are used to document the evolution of the valley flow, and flux data from the three 30-m Integrated Surface Flux Facility (ISFF) flux towers are used to evaluate the model surface energy budget. The locations of the systems are shown in Fig. 1b. The wind profiler/RASS is part of the National Center for Atmospheric Research’s (NCAR) mobile integrated sounding system (MISS; Parsons et al. 1994; Cohn et al. 2005) and was located south of Independence near the center of the valley. It produced a vertical profile of wind and temperature every 30 min. Wind profiler measurements at night are sometimes contaminated by echoes from migrating birds (e.g., Wilczak et al. 1995), and there was some evidence of similar contamination here. Three filtering techinques were applied to mitigate these signals. First, the raw Doppler spectra signals were filtered during collection with the widely used statistical averaging method algorithm (SAM; Merritt 1995); second, NCAR Improved Moments Algorithm (NIMA) processing (Morse et al. 2002) removed additional contaminated data; third, a final filter was applied to the EOP 5 data whereby samples with signal-to-noise ratio (SNR) greater than 15 dB, or standard deviation of SNR (over the 30-min averaging period) greater than 6 dB above 2.8 km, were removed. After application of these filtering techniques, there appeared to be very few bird echoes remaining in the data. The wind profiler/RASS data are well suited for model evaluation, as they represent an average over a relatively large volume (the width of the radar beam is 9°). For the present analysis, the RASS virtual temperature measurements are converted to potential temperature using surface temperature and pressure measurements from the collocated surface station (differences between measured virtual temperature and actual dry air temperature in the range of 0.4–0.9 K were ignored for this calculation). The sonic and radiation measurements from the 30-m towers were processed as described in Oncley et al. (2007).
Our investigation into the nocturnal flow evolution in Owens Valley is based on two T-REX EOPs with very different valley flow evolution: EOP 2 (29/30 March 2006) with its three layer structure and EOP 5 (29/30 April 2006) with its unusually strong downvalley jet.
b. Numerical experiments
ARPS, a comprehensive regional to storm-scale prediction system, is used to predict the nighttime flow evolution in Owens Valley. It is used in large-eddy simulation mode with a 1.5 turbulent kinetic energy (TKE) turbulence closure (Deardorff 1980; Moeng 1984). Radiation processes are parameterized using a sophisticated radiation package developed at the National Aeronautics and Space Administration (NASA) Goddard Space Flight Center [shortwave radiation is based on models of Chou (1990, 1992) and longwave radiation is based on Chou and Suarez (1994) and Tao et al. (1996)]. Topographic shading after Colette et al. (2003) was used. The land surface soil–vegetation model is based on the force–restore method (e.g., Noilhan and Planton 1989) and is described in detail in Xue et al. (2001) and Ren and Xue (2004). ARPS uses 13 soil types (including water and ice) and 14 vegetation classes. Land use, vegetation, and soil-type data for all Owens Valley grids are obtained from U.S. Geological Survey (USGS) 30-arc-s global data. Topographic height is from the USGS 30-arc-s global dataset for the 9- and 3-km grid and from the USGS 3-arc-s U.S. dataset for the 1-km and 350-m grid. Turbulent sensible heat, latent heat, and momentum fluxes in the surface layer are based on similarity theory (Monin and Obukhov 1954). The stability functions for unstable and neutral conditions are based on Byun (1990), and those for the free convection and stable case on Deardorff (1972).
ARPS is run in a one-way nesting mode. The outermost simulation domain (9-km grid spacing) is initialized from the National Oceanic and Atmospheric Administration’s (NOAA) 12-km North American Mesoscale (NAM) analysis dataset and is then successively nested down to grids of finer horizontal spacings (see Table 1 and Fig. 1). In addition to the standard 1-km grid, a large 1-km grid is used to investigate the sensitivity to domain size (see below). The selection of grid spacing, time steps, and other parameters such as computational mixing coefficients was made using experience gained through previous simulations over complex terrain (Chow et al. 2006; Weigel et al. 2006). NAM analyses were available at 6-hourly intervals; linear interpolation was used to provide lateral boundary conditions at intermediate times. All simulations start at 1200 UTC on the previous day and were integrated for 36 h.
Initialization of the land surface properties such as soil temperature, soil moisture, and snow cover using the NAM dataset resulted in poor simulation results (the near-surface flow at the central flux tower did not produce the observed valley wind shift near the end of the EOP time period). The NAM data contained large patches of snow approximately 0.5 m deep north of Independence and at the northern end of the valley while no snow was actually present. This inconsistency is due to the 12-km resolution terrain used by NAM that effectively smooths out Owens Valley. To compensate for the lack of initialization data with a fine enough resolution to resolve Owens Valley, the following procedure was adopted. On the 9-km domain, NAM soil moisture, soil temperature, and snow cover are used. However, a minimum snow-level elevation (3000 m for EOP 2; 3200 m for EOP 5) is imposed below which the snow depth is set to zero. On the 3-km and finer grids, idealized distributions are specified. Soil moisture is set to a fixed saturation rate of 20%, which corresponds roughly to the average of the gravimetric soil moisture measurements from 23 sites located in the central part of Owens Valley (Daniels et al. 2006). Initial soil temperature is set equal to the surface air temperature, and snow depth is set to 0.5 m above 3000 m (3200 m) and zero below for EOP 2 (EOP 5). According to visual observations in the field and based on snow depth and snow water equivalent information from the National Operational Hydrologic Remote Sensing Center, the snow line was in the 2500–3000-m range for EOP 2 and around 3000 m for EOP 5. Because snow cover in the model is given by a step function (either zero or 0.5 m) we choose a model snow line that is on the upper end of the estimated true snow line.
c. Sensitivity tests
Several sensitivity tests were undertaken to investigate the influence of the model setup on the simulation results (see also Schmidli and Poulos 2006; Daniels et al. 2006). The tests included variations in the soil moisture initialization, snow cover, computational mixing, grid resolution, and domain size and are briefly summarized below. A more detailed sensitivity analysis of thermally induced flows for a similar model setup using ARPS, but for a different region, can be found in Chow et al. (2006).
The evolution of the simulated flow was found to be sensitive to soil moisture in the entire catchment area of the valley. For example, increasing the soil moisture saturation rate on the 1- and 3-km grids from 20% to 40% delayed the onset of EOP 2 nighttime downvalley flow by about 2 h. This comes as no surprise, as several previous studies have documented the sensitivity of thermally induced flows to soil moisture (e.g., Ookouchi et al. 1984; Banta and Gannon 1995; Chow et al. 2006). Changing the snow cover (no snow versus snow line at 2000 m) had a minor influence on the evolution of the slope flows during the evening transition period, and did not significantly influence the evolution of the along-valley flow, likely due to the relatively small size of the snow-covered areas. Changing the computational mixing coefficient showed that a moderate amount of fourth-order computational mixing improved the agreement with observations, in comparison to using only the minimal amount required for computational stability.
The evolution of the valley atmosphere is well represented on the 1-km grid, but not on the 3-km grid. This is not surprising, as important topographic features such as the Inyo and White Mountains are not adequately resolved on the 3-km grid (in some places the model altitude is more than 1 km lower than the true altitude because of the smoothing of the topography). Increasing the model resolution from 1 km to 350 m improves the near-surface fields, but it has only a minor influence on the simulated evolution of the downvalley flow. Finally, the large 1-km grid (LRG1km in Fig. 1a) produces very similar results, as long as the same soil moisture is used on the 1- and 3-km grids.
In summary, while the details of the flow evolution, especially during the evening transition and near the surface, were found to be sensitive to the model setup (particularly to soil moisture), we have chosen a model configuration that is well justified based on observed conditions in the field. In addition, the qualitative features of the valley flow, such as the three-layer structure observed during EOP 2, were found to be robust and to compare well with the observational data.
3. Observed evolution of the valley atmosphere
Here we give a brief overview of the observed evolution of the nighttime boundary layer and valley atmosphere during EOP 2 and EOP 5. The upper-level synoptic situation for EOP 2 is characterized by the passage of a weak shortwave high pressure ridge (Fig. 2) and some high cirrus clouds between 0800 and 1400 UTC (not shown). The 500-hPa winds in the Owens Valley region remain fairly constant, at about 15–20 m s−1, turning from northwesterly to westerly directions during the course of the night. The upper-level synoptic situation for EOP 5 is similar; the region is again under a weak high pressure ridge. The 500-hPa winds are, however, significantly weaker (less than 10 m s−1) and from slightly more northerly directions. The skies are clear throughout the night. The low-level (850 hPa) synoptic situation for the two EOPs is characterized by a very weak north–south pressure gradient over the study region for EOP 2 and a somewhat stronger east–west pressure gradient for EOP 5.
Figure 3 depicts the evolution of the vertical distribution of wind and potential temperature observed at the MISS site in the valley center for EOP 2 and EOP 5. The most striking feature for EOP 2 is the layered structure of the valley atmosphere consisting of a low-level downvalley flow, a midlevel upvalley flow (at ∼3000 m), and westerly winds above the ridge line of the Sierra Nevada. The low-level flow is downvalley throughout the night. Note that the flow is already downvalley well before sunset (0200 UTC) because of strong synoptic forcing on 29 March. The downvalley flow is mostly fairly weak with velocities in the range of 3–6 m s−1; between 0500 and 0700 UTC, however, the velocities increase to 6–9 m s−1. The low-level flow switches to upvalley about 3 h after sunrise (1700 UTC). The midlevel upvalley flow layer first appears at about 0100 local time (0900 UTC) at heights between 3000 and 3500 m. At first, the flow is fairly weak, but it increases in strength throughout the night. The upvalley flow layer also steadily increases in depth as its lower boundary descends toward the surface where it merges with the daytime upvalley flow around 1700 UTC. The potential temperature measurements show the gradual cooling of the valley atmosphere throughout the night with minimum temperatures attained about 1 h after sunrise.
For EOP 5, the evolution of the valley wind is very different. The main feature is the strong and deep downvalley jet with winds exceeding 15 m s−1. The downvalley flow starts shortly after 0500 UTC and rapidly increases in strength and attains speeds over 15 m s−1 by 1000 UTC. It remains strong until the early morning hours (1500 UTC). With the onset of the daytime heating, the strong downvalley flow is slowly eroded, leading to weak and variable winds around noon. Because of the strong nighttime winds, temperature measurements from the RASS are less numerous for EOP 5. Nevertheless, one can see an indication of near-surface cooling before the onset of the strong downvalley flow and more or less constant conditions for the following 5 h.
4. Model evaluation
In this section a comparison of the results from the numerical simulations with various measurements is made to demonstrate that the model reproduces the key features of the structure and evolution of the valley atmosphere. Unless mentioned otherwise, all results refer to the 1-km-resolution simulation. We start out with a comparison with the wind profiler and RASS measurements, followed by a more detailed evaluation of the valley flow evolution, and the local thermal forcing. The forcings producing the layered structure for EOP 2 and the strong downvalley jet for EOP 5 are then analyzed in section 5.
a. Time evolution of vertical profiles of wind and temperature
Figure 4 shows the simulated evolution of wind and temperature corresponding to the observations in Fig. 3. The model reproduces well the characteristic features of the two EOPs, such as the downward growth and increase in strength of the midlevel upvalley flow layer for EOP 2, and the strong downvalley jet for EOP 5. The simulated peak downvalley flow for EOP 2 occurs, however, about an hour later than observed, and the onset of the upvalley flow is delayed by about 1–2 h. For EOP 5, the main model bias is again the delayed morning transition. Also, the maximum strength of the downvalley jet is underestimated by about 2 m s−1. Overall, however, the model reproduces the differences between the two cases very well. Note that the apparent difference in the near-surface temperature evolution between the model and observations is due to the fact that the lowest two gate ranges of the RASS are at 120 and 180 m above ground, respectively. Hence, the strong surface inversion is not seen in the RASS data.
b. Radiation and surface energy budget
An accurate simulation of the valley wind depends on an adequate representation of the surface energy balance (i.e., the net radiative forcing at the surface and its partitioning into sensible, latent, and ground heat flux). Nighttime average (0300–1300 UTC) measured and modeled fluxes at the three ISFF flux towers (see Fig. 1b) are compared in Table 2. The latent heat flux is not included as it is negligible for the two cases during the nighttime (averaged, measured, and modeled fluxes are typically less than 2 W m−2). Note that even for a perfect simulation substantial differences would be expected as the model values represent area averages while the measurements are much more locally determined. Average outgoing longwave radiation, for instance, differed by up to 8 W m−2 between the three radiometers located at the central tower within a few meters of each other. Also, measurement of the weak turbulent fluxes and the extrapolation of the flux measurements taken at 5 m above ground to actual surface fluxes are associated with considerable uncertainties (e.g., Oncley et al. 2007). The observed surface energy residuum varies between −8 and −28 W m−2.
For the southern and western tower, the difference in observed and modeled net radiation is similar for the two EOPs; the modeled net radiation is about 14% too large. This difference is mainly due to an underestimation of the incoming longwave radiation (ΔLWin is −11 W m−2 for EOP 2 and −14 W m−2 for EOP 5); the difference in outgoing longwave radiation is less than 5 W m−2 for both EOPs. For the central tower, the difference in observed and measured net radiation is considerably larger and due to differences in both incoming and outgoing longwave radiation.
The simulated sensible heat flux is close to the observed flux for EOP 5, but the simulated value is more than a factor of 2 larger than the observed flux for EOP 2. The largest differences between simulated and observed fluxes are found during the stable weak wind conditions in the early morning hours of EOP 2 (not shown). This is likely due to the parameterization of the surface fluxes in ARPS. In the current implementation the drag and heat transfer coefficient are not allowed to drop below 25% of their respective values for a neutral atmosphere (see also Poulos and Burns 2003). This may explain part of the cold bias visible for EOP 2 after 0900 UTC (in Fig. 7). The simulated ground heat flux, on the other hand, is close to the observed flux for EOP 2, at least for the southern and western tower, but it is 2.5–4 times larger than the observed flux for EOP 5.
This brief comparison points at various avenues for improvement, such as a more accurate description, initialization, and simulation of the land surface, or a more accurate parameterization of turbulent fluxes in the atmospheric surface layer. For the focus of the present study, however, which is on the evolution of the valley atmosphere and not on the surface fields, the skill of the model is adequate.
c. Structure of valley atmosphere
In Fig. 5, modeled vertical profiles of potential temperature, wind speed, wind direction, and specific humidity q for EOP 2 are compared with radiosonde data from the Independence airport site (see Fig. 1). The temperature structure of the atmosphere at sunset is characterized by a well-mixed layer that extends to about 4 km, a stable layer between 4 and 4.3 km, and a 5-K inversion between 4.3 and 4.7 km (not shown). During the course of the night, the valley atmosphere cools and the inversion descends and increases in strength. By sunrise, the valley atmosphere is stably stratified up to the base of the inversion, which has descended to about 3.2 km. The strong elevated inversion present in the early morning hours is part of the synoptic-scale flow. It is also visible in the soundings from Desert Rock in Nevada and Vandenberg Air Force Base in California (not shown). The model reproduces well the general cooling of the valley atmosphere. It does, however, exhibit a cold bias near the surface (below 1.4 km) and a weaker elevated inversion. The latter is not surprising as the vertical grid spacing at 3.2 km is about 200 m and the NAM input data are even less well resolved. At sunrise, the base of the model-simulated inversion is a few hundred meters below the observed inversion base. This fact together with the lower height of the Sierra Nevada ridge line in the model topography might explain the smaller depth of the simulated upvalley flow layer (cf. Figs. 3, 4). The inversion is associated with a sharp decrease in specific humidity. Although this sharp decrease is not captured by the model, the general decrease of specific humidity above the inversion is simulated well.
The wind structure at sunset is characterized by a moderate downvalley flow below 2.5 km, strong northwesterly flow above the Sierra Nevada ridge line, and very weak winds in between. A layer of weak upvalley flow is first visible in the 0927 UTC sounding. In the following hours, this layer increases in depth and strength. Again this evolution is captured well by the model. As already seen in the wind profiler data (Figs. 3, 4), the simulated wind speed of the downvalley flow is, however, too high.
A comparison with radiosonde data for EOP 5 is shown in Fig. 6. Again, the overnight cooling of the atmosphere is reproduced well. There is, however, an almost-height-independent bias of about −1.5 K. With regard to the wind structure, there is a directional error during the evening transition (0210 UTC), but good agreement with regard to the nighttime downvalley jet (0900 and 1402 UTC). The simulated specific humidity is slightly too large.
Overall the differences between the 1000- and 350-m simulation results are very small. The vertical profiles of potential temperature are almost identical for the two resolutions. And although the simulated wind profiles differ during the evening transition they are very similar after the onset of the downvalley flow. Thus while the 350-m resolution may improve the spatial detail of the simulated downslope flows, the simulated downvalley flow is almost identical to the 1000-m runs. Thus for the current purpose it is sufficient to focus on the 1000-m simulation results.
As the diurnal valley winds are driven by horizontal temperature contrasts, a good simulation of the temperature evolution of the valley atmosphere is crucial for an accurate simulation of the valley winds. While there are not sufficient observations to evaluate the along-valley variations in valley cooling rates, we can evaluate the evolution of the temperature at a single point in the core of the downvalley jet. In Fig. 7, the simulated temperature evolution at 1400 m (234 m AGL) at the MISS site is compared with the RASS and radiosonde measurements. This height corresponds roughly to the core of the simulated downvalley flow for both EOP 2 and EOP 5. Note the good agreement between the RASS and the radiosonde measurements most of the time, despite a horizontal separation of about 10 km. For EOP 2, the simulated cooling closely follows the observations until about 0900 UTC after which the model fails to reproduce the observed reduction in cooling rate. This is likely due to missing high-level clouds in the model and to the generally too high surface sensible heat flux under stable weak-wind conditions. Overall, however, the cooling and heating of the atmosphere is well reproduced by the model for both cases. At 2600 m, which is in the upvalley flow layer in EOP 2, the correspondence between observed and simulated cooling and heating rates is even better (not shown), and the model nicely reproduces the total nighttime cooling of about 2.5 K for EOP 2 and 5 K for EOP 5.
5. Nocturnal valley flow dynamics
a. Overview
We start with an overview of the simulated flow evolution. Figure 8 shows the simulated wind field at 1400 m—the height of the core of the downvalley flow—and on the lowest model level, respectively. The largest differences between the two EOPs occur around sunset (0200 UTC). While there is no significant along-valley flow at this height for EOP 2, there is still upvalley flow in the Owens Valley for EOP 5. For both cases, however, downslope flows have started to form. Three hours later, at 0500 UTC, the two cases are qualitatively similar. There are widespread downslope flows and in the northern part of the valley the downvalley flow has started. At 0800 UTC, the leading edge of the downvalley flow has reached the Owens Lake region and the downslope flows have weakened. Thereafter the downvalley flow remains fairly steady for EOP 5 and slowly decreases in strength for EOP 2 (not shown). Although the wind speeds differ considerably, the qualitative evolution of the low-level flow is quite similar for these two EOPs.
A more detailed comparison of the spatial structure of the downvalley flow at 0800 UTC, which corresponds to the beginning of the quasi-stationary phase, is given in Fig. 9. Note the similar spatial structure for the two EOPs. Because of the curvature of the Owens Valley in the Independence region, the downvalley flow is pushed to the western side of the valley. In addition, the broadening of the valley just north of Independence leads to a zone of stagnant flow or even counterflow on the eastern side of the valley (note the upvalley wind in the vertical cross sections at y = 20 km). Indeed, a horizontal vortex is visible in the EOP 2 case. While these local features are important for the local surface climate, the following analysis will focus mainly on along-valley averaged quantities.
b. Along-valley acceleration and layering
All terms on the rhs of Eq. (1) can induce an along-valley wind. The classic thermally driven valley wind is the result of locally produced horizontal temperature gradients (term A). Above the ridge top, however, horizontal temperature gradients may also be produced by large-scale temperature advection into one part of the valley. Pressure-driven channeling refers to a valley wind forced by term B. If the valley floor is steeply inclined, term C may dominate, and the along-valley flow will have characteristics of a buoyancy-driven slope flow. Strong flow aloft may force an along-valley flow by advective or turbulent downward transport of high momentum air from above the valley (terms D and E). Finally, horizontal advection (included in term D) may redistribute the along-valley momentum generated by the other processes. In a real valley, all forcings can be important at the same time.
Several simplifications can be made for the analysis of the two T-REX EOPs. For the most part, the valley floor of Owens Valley is almost horizontal—the average slope of the valley between Bishop (∼1250 m) and the valley mouth at Haiwee (1145 m) is only about 0.05°—therefore the nonhydrostatic pressure gradient (term C) can be neglected. The vertical transport of above-ridge-top along-valley momentum can also be excluded as a significant forcing factor of the along-valley wind for the two EOPs. The low values of turbulent kinetic energy and observed and modeled wind speed in the region below the Sierra Nevada crest line (see Figs. 3 –6) are not compatible with significant downward transport of high-momentum air from aloft. Thus the primary forcing of the low-level along-valley flow during EOP 2 and EOP 5 is due to the low-level along-valley pressure gradient, which for the present analysis has been decomposed into two terms, the thermal wind term that represents local forcings (term A) and the midlevel pressure gradient (term B) that represents mainly larger-scale forcings. These generating forces are opposed by surface friction and mediated through the turbulent momentum flux divergence term, which tends to decelerate the downvalley flow.
The evolution of the along-valley pressure forcing at three different heights is shown in Fig. 11. It can be seen that the evolution of the low-level pressure forcing [Δp1400, as in Eq. (4)], corresponding to the sum of the terms A and B in Eq. (1), is dominated by the evolution of the midlevel pressure forcing (Δp2000), except during the evening and morning transition periods during which the thermal forcing changes rapidly. At sunset (0200 UTC), the along-valley pressure forcing is close to zero at all three heights for both EOPs. In the following 2–3 h, the low-level pressure forcing (Δp1400) rapidly increases in strength (due to thermal and midlevel forcing). Thereafter it decreases slowly for EOP 2 but it continues to increase strongly for EOP 5. These differing forcing histories explain the resulting moderate downvalley wind for EOP 2 and the strong downvalley wind for EOP 5.
This figure points to the origin of the difference between the two EOPs. For EOP 2, the midlevel pressure gradient, after an initial increase, decreases throughout the night, eventually reversing its sign and leading to the onset of the midlevel upvalley flow. The continued increase of the upvalley forcing produces the downward growth and strengthening of the upvalley flow layer. For EOP 5, on the other hand, the midlevel pressure gradient continues to increase, thus producing the strong downvalley jet. Thus, while the low-level thermal forcing for the two cases is similar, the very different evolution of the superimposed midlevel forcing leads to the contrasting evolution of the valley flow.
Note that the along-valley component of the synoptic pressure difference above the peaks of the Sierra Nevada mountains (i.e., at 4400 m) is negative for both EOPs (−0.6 hPa for EOP 2 and −0.3 hPa for EOP 5). Thus, pure pressure-driven channeling induced by the upper-level synoptic pressure distribution would lead to southerly upvalley flow in the valley for both EOPs, albeit of different magnitudes. This corroborates that the key to the differences between the two EOPs lies in the different midlevel forcing.
The analysis above considers only two points along the valley axis to calculate the pressure gradient. Figure 12 shows snapshots of the along-valley variation of low- and midlevel pressure. For both cases the along-valley pressure gradients are very weak at 0200 UTC (not shown in Fig. 12, but seen in Fig. 11). At 0500 UTC, concurrent with the onset of the downvalley flow, there is a strong low-level pressure gradient in the northern part of the valley. Once the downvalley flow has established itself (0800 and 1100 UTC; Fig. 8), the low-level downvalley forcing remains located mainly in the northern part of the valley. The pressure gradient in the southern part of the valley remains weak and tracks the evolution of the midlevel pressure gradient, especially for EOP 2. This shows that the thermal forcing (regions with a large difference between the low- and midlevel pressure) is restricted mainly to the northern part of the valley. The midlevel pressure distribution is fairly uniform, characteristic of its nonlocal, larger-scale origin. For comparison, Fig. 12 also shows the relative magnitude of the (horizontal) advection term in Eq. (1). Note that by using u∂u/∂s = ∂/∂s(u2/2) and the definition of dynamic pressure pd = v2/2, the horizontal advection term is equal to the along-valley gradient of dynamic pressure. Comparison of the static and total (sum of static and dynamic) pressure distribution at 1400 m shows that the momentum advection term is at times not negligible, but it also does not significantly change the above conclusions.
c. Origin of the midlevel forcings
Figure 13 illustrates that the midlevel flow and forcing within the Owens Valley is the result of the interaction of the midlevel synoptic-scale flow with the Sierra Nevada mountain range. The interaction of this midlevel flow, which is southwesterly for EOP 2 and northerly for EOP 5, with the mountain range produces southerly flow for EOP 2 and northerly flow for EOP 5 not only within the valley but also in the adjacent areas. This corroborates the idea that the midlevel flow and forcing within the valley is not due to some local process. Further analysis of the synoptic evolution shows that the evolution of the midlevel along-valley forcing for EOP 5, with its maximum at 1200 UTC (see Fig. 11), is due to the passage of a cold front. The associated cold air advection into the northern end of the valley above about 2000 m leads to the stronger downvalley pressure forcing at 2000 m in comparison with 2600 m.
Inspection of the topography in Fig. 13 shows why Owens Valley is so susceptible to external influences well below the Sierra Nevada crest height. At 2600 m, only the Sierra Nevada and White Mountains portion of the valley sidewalls remains. The valley is thus poorly defined and open to influences from the north and the east. In fact, the valley starts to be poorly defined already at heights above about 2200 m (see Fig. 1). Therefore, the midlevel along-valley pressure gradient within the valley is easily dominated by external influences. In addition the valley is relatively wide, which further reduces the relative strength of local thermal forcings.
To isolate the large-scale forcings from the local thermal forcing we undertook an EOP 2 simulation for which the radiative heat transfer and surface heat fluxes were turned off on all grids. Figure 14 compares the resulting along-valley flow with the standard simulation. It can be seen that above about 2500 m, the flow structure is very similar for the two simulations. The nocturnal downvalley jet, however, is missing for the simulation with no thermal forcing. Note the upvalley flow along the eastern sidewall at y = 20 km in the reference simulation related to the local vortex (Fig. 9). This comparison confirms that the upvalley flow during EOP 2 is due to the interaction of the large-scale flow with the Sierra Nevada mountains and not due to local thermal forcing.
6. Conclusions
Numerical experiments are performed using ARPS aimed at understanding the differences in observed flow evolution in California’s Owens Valley for two nights that were expected to be dominated by local thermally driven valley winds due to the presence of a weak high pressure ridge. The performance of ARPS has been evaluated through comparison with data from the T-REX field campaign. The model is able to reproduce the large differences between the two nights and the main features of the valley wind system in Owens Valley. Therefore it is used to investigate the physical mechanisms leading to the observed flow evolution.
It is found that the Owens Valley is very susceptible to large-scale forcings originating from levels well below the crest height of the Sierra Nevada (∼4000 m). In contrast to a narrow and deep valley within a larger mountain range, Owens Valley is increasingly open to influences from the north and the east at heights above about 2200 m (which is well below the height of the western sidewall). It is shown that the low-level flow evolution is strongly influenced by these midlevel forcings as the corresponding midlevel along-valley pressure gradient is directly superimposed on the low-level thermal forcing.
Figure 15 summarizes the upper-level synoptic setting and the valley flow structure for the two EOPs. In EOP 2, strong upper-level westerlies and the associated geostrophic pressure gradient produce a midlevel pressure forcing that opposes the low-level thermal forcing. This results in the observed three-layer structure consisting of a layer of upvalley flow sandwiched between the thermally driven downvalley flow and the synoptic northwesterly flow. In EOP 5, a weaker upper-level forcing of the same sign is overpowered by a more active and opposing midlevel forcing. Hence, the resulting midlevel pressure forcing enhances the low-level thermal forcing, producing a strong downvalley jet with wind speeds exceeding 15 m s−1.
The susceptibility of the Owens Valley wind system to external influences points to the importance of large-scale forcing for explaining the observed valley flow structure and more generally for interpreting observations taken in the Owens Valley including near-surface observations.
Acknowledgments
The ground-based instruments teams from the university community and the National Center for Atmospheric Research’s Earth Observing Laboratory (NCAR EOL) are recognized for their outstanding efforts at the Terrain-Induced Rotor Experiment (T-REX). The primary sponsor of T-REX is the U.S. National Science Foundation (NSF). The observational data were downloaded from the T-REX Data Archive, which is maintained by NCAR EOL. Comments by W. Brown and T. Weckwerth on an earlier version of this manuscript are appreciated. The authors are indebted to the Center for Analysis and Prediction of Storms, University of Oklahoma, for access to and support of the ARPS model. Analyses and graphics were produced with the open source software packages python, matplotlib, and NCAR’s PyNGL. The support of the Swiss National Science Foundation (Grant PA002-111427) for the first author and the support of NSF Grants ATM-0453595 and ATM-0645784 (Physical Meteorology Program: S. Nelson, Program Director) (FKC and MHD) are gratefully acknowledged.
REFERENCES
Banta, R. M., 1986: Daytime boundary-layer evolution over mountainous terrain. Part II: Numerical studies of upslope flow duration. Mon. Wea. Rev., 114 , 1112–1130.
Banta, R. M., and W. R. Cotton, 1981: An analysis of the structure of local wind systems in a broad mountain bassin. J. Appl. Meteor., 20 , 1255–1266.
Banta, R. M., and P. T. Gannon, 1995: Influence of soil moisture on simulations of katabatic flow. Theor. Appl. Climatol., 52 , 85–94.
Barr, S., and M. M. Orgill, 1989: Influence of external meteorology on nocturnal valley drainage winds. J. Appl. Meteor., 28 , 497–517.
Byun, D. W., 1990: On the analytical solutions of flux-profile relationships for the atmospheric surface layer. J. Appl. Meteor., 29 , 652–657.
Chou, M-D., 1990: Parameterization for the absorption of solar radiation by O2 and CO2 with application for climate studies. J. Climate, 3 , 209–217.
Chou, M-D., 1992: A solar radiation model for climate studies. J. Atmos. Sci., 49 , 762–772.
Chou, M-D., and M. J. Suarez, 1994: An efficient thermal infrared radiation parameterization for use in general circulation models. NASA Tech. Memo. 104606, Vol. 3, 102 pp.
Chow, F. K., A. P. Weigel, R. L. Street, M. W. Rotach, and M. Xue, 2006: High-resolution large-eddy simulations of flow in a steep Alpine valley. Part I: Methodology, verification, and sensitivity experiments. J. Appl. Meteor. Climatol., 45 , 63–86.
Cohn, S. A., W. O. J. Brown, and V. Grubisic, 2005: The Mobile Integrated Sounding System (MISS): Description and lessons from the Sierra Rotors Project. Preprints, 13th Symp. on Meteorological Observations and Instrumentation, Savannah, GA, Amer. Meteor. Soc., 7.2. [Available online at http://ams.confex.com/ams/pdfpapers/94055.pdf.].
Colette, A., F. K. Chow, and R. L. Steet, 2003: A numerical study of inversion-layer breakup and the effects of topographic shading in idealized valleys. J. Appl. Meteor., 42 , 1255–1272.
Coulter, R. L., and P. Gudiksen, 1995: The dependence of canyon winds on surface cooling and external forcing in Colorado’s Front Range. J. Appl. Meteor., 34 , 1419–1429.
Daniels, M. H., F. K. Chow, and G. S. Poulos, 2006: Effects of soil moisture initialization on simulations of atmospheric boundary layer evolution in Owens Valley. Extended Abstracts, 12th Conf. on Mountain Meteorology, Santa Fe, NM, Amer. Meteor. Soc., 7.2. [Available online at http://ams.confex.com/ams/pdfpapers/114757.pdf.].
Davies-Jones, R., 2003: An expression for effective buoyancy in surroundings with horizontal density gradients. J. Atmos. Sci., 60 , 2922–2925.
Deardorff, J. W., 1972: Parametrization of the planetary boundary layer for use in general circulation models. Mon. Wea. Rev., 100 , 93–106.
Deardorff, J. W., 1980: Stratocumulus-capped mixed layers derived from a 3-dimensional model. Bound.-Layer Meteor., 18 , 495–527.
De Wekker, S. F. J., S. Zhong, J. D. Fast, and C. D. Whiteman, 1998: A numerical study of the thermally driven plain-to-basin wind over idealized basin topographies. J. Appl. Meteor., 37 , 606–622.
De Wekker, S. F. J., D. G. Steyn, J. D. Fast, M. W. Rotach, and S. Zhong, 2005: The performance of RAMS in representing the convective boundary layer structure in a very steep valley. Environ. Fluid Mech., 5 , 35–62.
Doran, J. C., 1991: The effects of ambient winds on valley drainage flows. Bound-Layer Meteor., 55 , 177–189.
Doran, J. C., J. D. Fast, and J. Horel, 2002: The VTMX 2000 campaign. Bound.-Layer Meteor., 83 , 537–551.
Egger, J., S. Bajrachaya, U. Egger, R. Heinrich, J. Reuder, P. Shayka, H. Wendth, and V. Wirth, 2000: Diurnal winds in the Himalayan Kali Gandaki Valley. Part I: Observations. Mon. Wea. Rev., 128 , 1106–1122.
Estournel, C., and D. Guedalia, 1985: Influence of geostrophic wind on atmospheric nocturnal cooling. J. Atmos. Sci., 42 , 2695–2698.
Fast, J. D., K. J. Allwine, R. N. Dietz, K. L. Clawson, and J. C. Torcolini, 2006: Dispersion of perfluorocarbon tracers within the Salt Lake valley during VTMX 2000. J. Appl. Meteor., 45 , 793–812.
Fiedler, F., 1983: Einige Charakteristika der Strömung im Oberrheingraben. Wissenschaftliche Berichte des Instituts f¨r Meterologie und Klimaforschung der Universität Karlsruhe, Band 4, 113–123.
Gross, G., and F. Wippermann, 1987: Channeling and countercurrent in the upper Rhine Valley: Numerical simulations. J. Climate Appl. Meteor., 26 , 1293–1304.
Grubišic, V., and Coauthors, 2008: The terrain-induced rotor experiment: A field campaign overview including observational highlights. Bull. Amer. Meteor. Soc., 89 , 1513–1533.
Gudiksen, P. H., J. M. Leone, C. W. King, D. Ruffieux, and W. D. Neff, 1992: Measurements and modeling of the effects of ambient meteorology on nocturnal drainage flows. J. Appl. Meteor., 31 , 1023–1032.
Kossmann, M., and A. P. Sturman, 2003: Pressure-driven channeling effects in bent valleys. J. Appl. Meteor., 42 , 151–158.
Kuwagata, T., and F. Kimura, 1995: Daytime boundary layer evolution in a deep valley. Part I: Observations in the Ina Valley. J. Appl. Meteor., 34 , 1082–1091.
Kuwagata, T., and F. Kimura, 1997: Daytime boundary layer evolution in a deep valley. Part II: Numerical simulation of the cross-valley circulation. J. Appl. Meteor., 36 , 883–895.
Mahrt, L., 1982: Momentum balance of gravity flows. J. Atmos. Sci., 39 , 2701–2711.
McKee, T. B., and R. D. O’Neil, 1989: The role of valley geometry and energy budget in the formation of nocturnal valley winds. J. Appl. Meteor., 28 , 445–456.
Merritt, D. A., 1995: A statistical averaging method for wind profiler Doppler spectra. J. Atmos. Oceanic Technol., 12 , 985–995.
Moeng, C-H., 1984: A large-eddy simulation model for the study of planetary boundary-layer turbulence. J. Atmos. Sci., 41 , 2052–2062.
Monin, A. S., and A. M. Obukhov, 1954: Basic laws of turbulent mixing in the ground layer of the atmosphere. Akad. Nauk SSSR Geofiz. Inst. Tr., 151 , 163–187.
Morse, C. S., R. K. Goodrick, and L. B. Cornman, 2002: The NIMA method for improved moment estimation from Doppler spectra. J. Atmos. Oceanic Technol., 19 , 274–295.
Noilhan, J., and S. Planton, 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117 , 536–549.
Noppel, H., and F. Fiedler, 2002: Mesoscale heat transport over complex terrain by slope winds—A conceptual model and numerical simulations. Bound.-Layer Meteor., 104 , 73–97.
Oncley, S. P., and Coauthors, 2007: The Energy Balance Experiment EBEX-2000. Part I: Overview and energy balance. Bound.-Layer Meteor., 123 , 1–28.
Ookouchi, Y., M. Segal, R. C. Kessler, and R. A. Pielke, 1984: Evaluation of soil moisture effects on the generation and modification of mesoscale circulations. Mon. Wea. Rev., 112 , 2281–2292.
Parsons, D., and Coauthors, 1994: The integrated sounding system: Description and preliminary observations from TOGA COARE. Bull. Amer. Meteor. Soc., 75 , 553–567.
Poulos, G. S., and S. P. Burns, 2003: An evaluation of bulk Ri-based surface layer flux formulas for stable and very stable conditions with intermittent turbulence. J. Atmos. Sci., 60 , 2523–2537.
Poulos, G. S., J. E. Bossert, R. A. Pielke, and T. B. McKee, 2000: The interaction of katabatic flow and mountain waves. Part I: Observations and idealized simulations. J. Atmos. Sci., 57 , 1919–1936.
Poulos, G. S., J. E. Bossert, T. B. McKee, and R. A. Pielke, 2007: The interaction of katabatic flow and mountain waves. Part II: Case study analysis and conceptual model. J. Atmos. Sci., 64 , 1857–1879.
Rampanelli, G., D. Zardi, and R. Rotunno, 2004: Mechanisms of upvalley winds. J. Atmos. Sci., 61 , 3097–3111.
Ren, D., and M. Xue, 2004: A revised force-restore model for land surface modeling. J. Appl. Meteor., 43 , 1768–1782.
Rotach, M. W., and D. Zardi, 2007: On the boundary-layer structure over highly complex terrain: Key findings from MAP. Quart. J. Roy. Meteor. Soc., 133 , 937–948.
Schmidli, J., and G. S. Poulos, 2006: High-resolution modeling of the nighttime boundary layer evolution in the Owens valley: Comparison to observations. Extended Abstracts, 12th Conf. on Mountain Meteorology, Santa Fe, NM, Amer. Meteor. Soc., 7.1. [Available online at http://ams.confex.com/ams/pdfpapers/114527.pdf.].
Smith, C. M., and E. D. Skyllingstad, 2005: Numerical simulation of katabatic flow with changing slope angle. Mon. Wea. Rev., 133 , 3065–3080.
Steeneveld, G. J., B. J. H. van de Wiel, and A. A. M. Holtslag, 2006: Modeling the evolution of the atmospheric boundary layer coupled to the land surface for three contrasting nights in CASES-99. J. Atmos. Sci., 63 , 920–935.
Steinacker, R., 1984: Area-height distribution of a valley and its relation to the valley wind. Contrib. Atmos. Phys., 57 , 64–71.
Tao, W-K., S. Lang, J. Simpson, C-H. Sui, B. Ferrier, and M-D. Chou, 1996: Mechanism of cloud–radiation interaction in the tropics and midlatitude. J. Atmos. Sci., 53 , 2624–2651.
Vergeiner, I., and E. Dreiseitl, 1987: Valley winds and slope winds—Observations and elementary thoughts. Meteor. Atmos. Phys., 36 , 264–286.
Wagner, A., 1938: Theorie und Beobachtung der periodischen Gebirgswinde. Gerl. Beitr. Geophys., 52 , 408–449.
Weigel, A. P., F. K. Chow, M. W. Rotach, R. L. Street, and M. Xue, 2006: High-resolution large-eddy simulations of flow in a steep Alpine valley. Part II: Flow structure and heat budgets. J. Appl. Meteor. Climatol., 45 , 87–107.
Weigel, A. P., F. K. Chow, and M. W. Rotach, 2007: The effect of mountainous topography on moisture exchange between the surface and the free atmosphere. Bound.-Layer Meteor., 125 , 227–244.
Whiteman, C. D., 1989: Morning transition tracer experiments in a deep narrow valley. J. Appl. Meteor., 28 , 626–635.
Whiteman, C. D., 1990: Observations of thermally developed wind systems in mountainous terrain. Atmospheric Processes over Complex Terrain, Meteor. Monogr., No. 23, Amer. Meteor. Soc., 5–42.
Whiteman, C. D., 2000: Mountain Meteorology: Fundamentals and Applications. Oxford University Press, 355 pp.
Whiteman, C. D., and J. C. Doran, 1993: The relationship between overlying synoptic-scale flows and winds within a valley. J. Appl. Meteor., 32 , 1669–1682.
Wilczak, J. M., and Coauthors, 1995: Contamination of wind profiler data by migrating brids: Characteristics of corrupted data and potential solutions. J. Atmos. Oceanic Technol., 12 , 449–467.
Xue, M., K. K. Droegemeier, and V. Wong, 2000: The Advanced Regional Prediction System (ARPS)—A multi-scale nonhydrostatic atmospheric simulation and prediction model. Part I: Model dynamics and verification. Meteor. Atmos. Phys., 75 , 161–193.
Xue, M., and Coauthors, 2001: The Advanced Regional Prediction System (ARPS)—A multi-scale nonhydrostatic atmospheric simulation and prediction model. Part II: Model physics and applications. Meteor. Atmos. Phys., 76 , 143–165.
Xue, M., D. H. Wang, J. D. Gao, K. Brewster, and K. K. Droegemeier, 2003: The Advanced Regional Prediction System (ARPS), storm-scale numerical weather prediction and data assimilation. Meteor. Atmos. Phys., 82 , 139–170.
Zängl, G., J. Egger, and V. Wirth, 2001: Diurnal winds in the Himalayan Kali Gandaki Valley. Part II: Modeling. Mon. Wea. Rev., 129 , 1062–1080.
Zhong, S. Y., and J. Fast, 2003: An evaluation of the MM5, RAMS, and Meso-Eta models at subkilometer resolution using VTMX field campaign data in the Salt Lake Valley. Mon. Wea. Rev., 131 , 1301–1322.
(a) Topography of the larger Owens Valley region including the Sierra Nevada mountain range west of the valley and the Inyo and White Mountains to the east (MM: Mount Montgomery Pass, BP: Big Pine, IND: Independence, OL: Owens Lake, HW: Haiwee Pass). The white boxes outline the standard (STD1km) and large (LRG1km) 1-km simulation domains for the Owens Valley. The thick black line depicts the 2000-m contour, the thin lines depict the 800- and 3200-m contour. (b) Actual smoothed topography of the standard 1-km grid. Location of the surface stations is also shown (circles: NCAR ISFF flux towers; asterisk: MISS; up triangle: launch site of the radiosondes at Independence airport). The white box outlines the 350-m simulation domain, and the black lines denote the location of the vertical cross sections shown in Figs. 9 and 14, respectively. Data extracted from points A and B are used in section 5b.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Synoptic situation for the two EOPs. Geopotential height and wind barbs at 1200 UTC 30 Mar for EOP 2 and 30 Apr for EOP 5 from the NAM analysis. The asterisk denotes the location of Independence.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Time–height cross section of observed winds and potential temperature at MISS. The bar indicates the wind speed (m s−1), and the vertical axis is in kilometers. The contour interval for potential temperature is 2.5 K. The two horizontal lines denote the approximate height of the Sierra Nevada and the Inyos, and the two vertical lines denote local sunset and sunrise. Note that the contour lines of potential temperature are broken because of missing data.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
As in Fig. 3, but for the numerical simulation.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Observed radiosonde (black) and simulated (gray) profiles of potential temperature (pt), wind speed, wind direction, and specific humidity q for EOP 2. Simulated profiles are for the 1000-m (solid line or large symbols) and 350-m (dashed line or small symbols) resolution. The horizontal line denotes the approximate height of the Sierra Nevada ridge line.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
As in Fig. 5, but for EOP 5.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Measured (black) and simulated (gray) time series of potential temperature at 1400 m at the MISS site (Independence airport for the radiosonde).
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Temporal evolution of the simulated wind field at 1400 m (where ground level is below 1400 m) and on the lowest model level (where ground level is higher than 1400 m), respectively. The color bar denotes the wind speed (m s−1). Wind vectors are plotted at every fourth grid point. The asterisk denotes the location of the MISS site. Topography contours are shown at 1400 (thick), 2000, and 3200 m. The horizontal and vertical axes indicate distance (km).
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Snapshot of the downvalley flow at 0800 UTC (350-m grid). (top) Simulated wind field at 1400 m (as in Fig. 8). (bottom) Simulated along-valley wind component (thick lines; contour interval 2 m s−1) and potential temperature (thin lines; contour interval 2 K) in the cross sections indicated in Fig. 1b.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Simulated time series of the thermally induced along-valley pressure gradient forcing (ΔpT) in the 1400–2000-m layer between points A and B in Fig. 1b and the actual pressure gradient forcing (Δp = Δp1400−Δp2000). Note that for a layer depth of 600 m, a horizontal pressure difference of 0.2 hPa corresponds to a temperature difference of about 1 K.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Simulated time series of along-valley pressure forcing (Δpz) at heights of 1400, 2000, and 2600 m MSL between points A and B in Fig. 1b (pressure at B relative to pressure at A).
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Simulated along-valley pressure variation at the height of the downvalley flow (1400 m) and of the upvalley flow (2600 m). The values are averaged over the valley width as defined by the 1400-m contour; (p + pd) refers to the sum of static and dynamic pressure (see text). The y axis denotes the model y coordinate (see Fig. 8).
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Simulated wind field at 2600 m at 1200 UTC (3-km grid). The bar denotes the wind speed (m s−1). Wind vectors are plotted at every third grid point. The asterisk denotes the location of the MISS site. Topography contours are shown at 2600 (thick), 0, 800, 2000, and 3200 m. Axis units are in kilometers.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Comparison of simulated along-valley wind component (thick lines; contour interval 2 m s−1) and potential temperature (thin lines; contour interval 2 K) for EOP 2 at 1400 UTC in the west–east cross sections shown in Fig. lb for the reference simulation and a simulation with no radiation and no surface heat flux (350-m grid).
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Schematic of the valley flow structure for the two EOPs showing the three flow layers with the upper-level synoptic flow blowing parallel to the isobars.
Citation: Journal of Applied Meteorology and Climatology 48, 1; 10.1175/2008JAMC1852.1
Nested grid configurations, with dimensions and time step sizes. In the vertical direction, the minimum grid spacing is Δzmin at the surface, and the average spacing is Δzavg; Δt denotes the large time step and Δτ is the small time step.
Components of surface energy budget averaged over 0300–1300 UTC. The observed values for sensible heat flux are from the 5-m level; f denotes the fraction of the simulated to the observed value. All values are in watts per meter squared.
All altitudes are above mean sea level.