A Climatological Study of Thermally Driven Wind Systems of the US Intermountain West

Jebb Q. Stewart1, C. David Whiteman2, W. James Steenburgh1, and Xindi Bian2

1NOAA/Cooperative Institute for Regional Prediction and Department of MeteorologyUniversity of Utah

Salt Lake City, Utah

2Pacific Northwest National LaboratoryRichland, Washington

Submitted to Bulletin of the American Meteorological Society

Proposal submission date: 06 June 2001Planned article submission date: 1 August 2001

Corresponding Author:W. James Steenburgh

University of UtahDepartment of Meteorology

145 South 1460 East Room 819Salt Lake City, UT 84112-0110

(801) 581-8727e-mail: jimsteen@met.utah.edu

ABSTRACT

The Intermountain West of the United States, because of its dry continental climate, is an

area where local thermally driven winds occur frequently. This paper documents the diurnal

evolution of thermally driven valley, slope, and lake winds for summer fair-weather conditions

in four regions of the Intermountain West where dense wind networks have been operated.

Because of the diverse topography in these regions, the results are expected to be broadly

representative of variations in thermally driven wind climates throughout the Intermountain

West. The regions include the Wasatch Front Valleys of northern Utah, the Snake River Plain of

Idaho, the Southern Nevada basin and range province, and central Arizona where the Salt and

Gila Valleys run out onto a plain. The analyses focus on the interplay of valley, slope, and lake

winds, as well as the frequency and regularity of the winds.

In general, on fair weather days, winds in all four regions exhibit a consistent direction from

day to day at a given hour. A measure of this wind consistency is defined. The nighttime hours

exhibit a high consistency, the daytime hours a moderate consistency, and transition periods a

low consistency. The low consistency during the transition periods reflects day-to-day variations

in the timing of thermally driven wind system reversals. Thermally driven circulations are

similar in the four regions, but the Wasatch Front Valleys are influenced by lake breezes from

the adjacent Great Salt Lake, the Snake River Plain is influenced by wind channeling from

nearby mountains, Southern Nevada has both local and region scale thermally driven winds that

are produced by the distinctive Basin and Range topography, and winds in Central Arizona

experience regional-scale circulations associated with regional scale contrasts in elevation and/or

surface heating.

1. Introduction

The complex topography of the western United States produces a wide variety of thermally

and dynamically driven mesoscale wind systems. Historically, the knowledge and understanding

of such systems has been limited by a variety of factors, including a lack of observational data.

In the Intermountain West (IW, Fig. 1), the semiarid area between the Cascade-Sierra and Rocky

Mountain chains, dry air and soil conditions promote intense daytime solar heating and nighttime

radiation loss resulting in large diurnal temperature changes that produce strong thermally driven

winds (Carter 2000). Thermally driven wind systems in mountainous terrain consist of three

major wind circulations: plain-basin or plain-mountain winds, valley winds, and slope winds.

The different phasing and superposition of the valley and slope wind systems produces

clockwise and counterclockwise diurnal rotations on the right and left banks or sidewalls of a

valley, respectively, when looking up a valley (Hawkes 1947, Whiteman 1990). In some regions

of the IW such as the Wasatch Front Valleys (WFV) of northern Utah where mountains are

located adjacent to lakes (Fig. 2a), a forth diurnal wind system is apparent, the lake-land breeze.

The interaction of these four wind systems creates complex flow patterns that are a part of the

everyday winds in complex terrain. Understanding of these wind systems is valuable to boundary

layer and air pollution meteorologists, fire weather forecasters and operational meteorologists.

Over the past few years high-density data from more than 50 independent meteorological

networks has been gathered as part of MesoWest, a collection of cooperative mesonets in the

western United States (Horel et al. 2000). MesoWest is managed jointly by the NOAA

Cooperative Institute for Regional Prediction at the University of Utah and the Salt Lake City

National Weather Service Forecast Office and presently collects data from over 2500 surface

stations in the western United States. In this paper, we illustrate the diurnal evolution of complex

terrain winds in the IW using MesoWest data from 1997 to 1999.

2. Areas of study

The four study regions were chosen because they have high density observations and

illustrate typical thermally driven wind systems of the IW (Fig. 1). Here we introduce the

regions, in turn, going counterclockwise around the IW.

The Salt Lake, Tooele and Rush Valleys, designated here as the Wasatch Front Valleys or

WFV (Fig. 2a) are bounded by three north-south mountain ranges that extend above 3000 m

mean sea level (MSL) – the Wasatch, Oquirrh and Stansbury Mountains1. The largest of the

valleys, the Salt Lake Valley, is a broad valley containing the extensive Salt Lake City urban

area. This valley, bounded to the east by the Wasatch Mountains and to the west by the Oquirrh

Mountains, drains northward into the Great Salt Lake (GSL, 1300 m). The Salt Lake Valley is

generally considered to be bounded to the south by a transverse mountain range (the Traverse

Mountains) that extends between the Wasatch and Oquirrh Mountains. In reality, however, the

Jordan River which drains the valley flows northward from the Utah Valley through a narrow

gap in the Traverse Range. Several major canyons issue into the Salt Lake Valley from the

Wasatch Mountains. The Tooele and Rush Valleys are a pair of valleys that are bounded by the

Oquirrh Mountains on the east, the Stansbury Mountains on the west, and the Tintic and Sheep

Rock Mountains on the south. The valleys, like the adjacent Salt Lake Valley, drain northward

into the GSL. A transverse mountain range culminating in a peak called South Mountain

separates these two valleys into an upper basin with minimum elevation of 1560 m (the Rush

Valley) and a lower valley (the Tooele Valley) that drains directly into the GSL. Several low

passes on the southeast boundary of the Rush Valley connect the Rush Valley to the Salt Lake

1 Unless otherwise indicated, all elevations specified in the remainder of this paper are relative to mean sea-level.

Valley. The region is semiarid, with vegetation consisting mainly of grasses and low shrubs

(mainly sagebrush). The GSL, with an average depth of only 4.8 m, exhibits little seasonal

temperature lag relative to the mean daily air temperatures in the lake’s surroundings

(Steenburgh et al. 2000).

The Snake River Plain (SRP) is a broad, flat-floored, arc-shaped valley in southern Idaho that

slopes downward to the west from 1500 to 900 m (Fig. 2b). The valley is bounded to the north

by the central Idaho Mountains, which reach elevations above 3000 m, and to the south by

several mountain ranges that reach elevations of around 2100 m. Major canyons issue onto the

SRP from the surrounding mountains, with three especially prominent canyons entering the east

end of the SRP from the north. The SRP is arid or semiarid with vegetation consisting of short

grasses and shrubs (mainly sagebrush) at the lower elevations and low density coniferous forests

at higher elevations.

Terrain in Southern Nevada (Fig. 2c) slopes steeply upward from southwest to northeast,

culminating in the Great Basin, an extensive intermountain basin with basin floor elevations near

1500 m located between the Sierra-Cascade Mountains and the Rocky Mountains. The Southern

Nevada landscape is part of the basin and range province, a geomorphological province

characterized by many short north-south oriented mountain ranges that reach elevations around

3000 m and are separated by broad alluvial basins. The lowest elevations in the Southern

Nevada region are to the southwest in California’s Death Valley (-86 m) and to the southeast

along the Colorado River (360 m). The climate is semiarid except on the highest mountain

ranges. Vegetation is sparse, with short grasses and shrubs at mid elevations and low density

forests at the highest elevations. Barren ground and arid vegetation are found at lower

elevations.

The central Arizona region lies to the west of the Sierra Ancha (Fig 2d). The general terrain

slopes from the Sierra Ancha (1900 m) southwesterward to the Yuma Desert (200 m). The

Mazatzal Mountains are a 1500 m barrier between the Sierra Ancha and Phoenix (PHX). From

the Sierra Ancha and surrounding mountain ranges, several canyons issue into the lower Phoenix

basin, including the Black Canyon north of Phoenix through which Interstate 17 runs. Central

Arizona is arid, with intense sunlight throughout much of the summer. Only sparse grass

vegetation survives the lack of precipitation leaving large areas of barren ground. Short shrubs

and grasses survive at mid elevations, and low density coniferous forests are found at the highest

elevations.

3. Data and methods

Surface observations were provided by MesoWest, a collection of more than 50

independently operated mesonets across the western U.S. (Splitt and Horel 1998; Horel et al.

2000; Tucker 1997). The MesoWest network provides high density observations in regions that

are often not well sampled by the conventional Federal Aviation Administration/National

Weather Service/ Department of Defense networks. Data were collected via phone modems,

internet connections, or radio transmissions and archived at the University of Utah where an

automated quality control process removes erroneous values.

Additional processing steps were implemented to prepare the data for analysis of thermally

driven circulations in the study regions. First, data for the summer months June through August

were extracted from the MesoWest archive for the years 1997-1999. The summer period was

chosen because synoptic flows are typically weak then and valley flows are more pronounced.

Second, any sub-hourly-averaged data (e.g., 5 min, 15 min, or 30 min) were averaged to one

hour, using the convention that the time indicated represented the beginning of the hour of

interest. Third, time periods with weak winds aloft and clear to partly cloudy skies were selected

for analysis since these are times when thermally driven flows are most pronounced. Such

periods were identified as the 12-hour blocks centered on the rawinsonde observation times

(2200 - 0900 LST for the 0400 LST soundings and 1000 - 2100 LST for the 1600 LST

soundings2), for which the rawinsonde 700-hPa wind speeds were 7 m s-1. The 12-hour blocks

were included in the climatology only if the total daily solar radiation was 65% of the

theoretical extraterrestrial solar radiation for that day as computed from a solar model

(Whiteman and Allwine 1986). The solar radiation criteria was considered to applied to the

2200-2100 LST period that corresponds to the rawinsonde time periods above. These criteria are

similar to those used to identify periods of thermally driven flows in previous research in the

Intermountain West (Whiteman et al. 1999). Locations for the 700 hPa wind and solar radiation

data are listed in Table 1. All hourly observations from the 12-hour blocks meeting the criteria

described above were grouped into hourly bins and processed to determine a mean vector wind

and a mean arithmetic wind speed for that hour of the day. The mean hourly vector winds were

then plotted on maps and a video loop was prepared to animate the hourly maps to investigate

the evolution of the hourly winds during a mean thermally driven wind day. Then, a method was

developed to determine the fair weather day-to-fair weather day variance of wind direction for

each hour of the day. This was accomplished by defining wind consistency, defined as the ratio

of the vector mean and arithmetic mean wind speeds for each hour of the day. If the wind blows

from the same direction at a given hour on all fair weather days, the consistency is one; if it is

equally likely from all directions, or blows half the time from one direction and half the time

from the opposite, the consistency is 0. This definition of wind consistency is similar to Panofsky

2 Soundings are launched an hour before 0000 and 1200 UTC soundings. Data from the lower elevations of the soundings (below 700 hPa) are applied to the hour of launch.

and Brier's (1965) definition of wind persistence, the steadiness of the wind over a continuous

time period. We prefer the term wind consistency because we deal with a discontinuous data set

(only fair weather days and observations at a given hour of the day).

4. Results

a. Wasatch Front Valleys

The Wasatch Front Valleys are located south of the GSL where Hawkes (1947), Smidy

(1972), Astling (1986) and Stone and Hoard (1990a, 1990b) investigated the interactions

between the diurnal land/lake and mountain wind systems. These previous studies did not,

however, attempt to isolate the thermally driven component of the flow by considering only

situations with clear skies and weak synoptic-scale forcing.

Winds in the Wasatch Front Valleys are the result of lake/land breezes, slope flows, and

valley flows. During the night (0400 LST), downslope and down-valley winds are observed in

the Salt Lake and Tooele Valleys (Fig, 3a). The down-valley winds are likely reinforced by

offshore flow induced by the GSL. Since the Tooele Valley slopes little near the GSL and

steepens as one moves towards South Mountain, weaker winds (1-3 m s-1) are found near the

GSL, while stronger winds (3-4 m s-1) are found farther up valley. In the Rush Valley, an

elevated basin with few low gaps for winds to exit, the slope of the land is gentle except near the

mountains. This results in light (~1 m s-1) downslope winds over lower portions of the valley

that converge into the basin. At higher elevation sites within the Rush Valley, stronger

downslope winds (1.5 to 4 m s-1) are found. Along the South Mountain divide, 4-5 m s-1

southerly winds illustrate the nocturnal flow of air across the divide from the Rush Valley to the

Tooele Valley. The southwest wind observed along I-15 near the summit of the Traverse

Mountains in the Salt Lake Valley illustrates a similar flow pattern between the Utah and Salt

Lake Valleys (small-scale terrain effects result in southwesterly rather than southerly flow at this

site). These conditions persist until the morning transition period.

The morning transition period starts at sunrise (~0600 LST) and continues through about

1100 LST. During this period, downslope and down-valley winds weaken and eventually

become upslope and up-valley, with the slope flow transition preceding the valley flow transition

by about 2 h. At 1000 LST, the mid-point of the morning transition period, winds over the Salt

Lake Valley and eastern Tooele Valley develop an upslope component (Fig. 3b) while the down-

valley flow begins to weaken. A lake breeze begins to move into the northern Tooele Valley.

This circulation feature is first observed near the GSL shoreline at 0700 LST and advances

southward through the Tooele Valley to South Mountain by 1100 LST (not shown). The

southward lake-breeze penetration is slower over the Salt Lake Valley (Fig. 3b).

During the afternoon regime, circulations within the Wasatch Front Valleys are the result of

interactions between upslope flows, up-valley flows, and the GSL breeze. At 1600 UTC,

observations in the Tooele Valley show a coupling between upslope and up-valley flows, which

are in phase with the GSL breeze and produce a diffluent up-valley wind pattern (Fig. 3c). A

similar but weaker flow pattern in which upslope flows are more dominant is observed over the

Salt Lake Valley. The anomalous westerly flow at Magna (QMG) in the Salt Lake Valley is the

result of the interaction of the GSL breeze with the local terrain. In contrast to the Tooele and

Salt Lake Valleys, there is little evidence of upslope and up-valley winds within the Rush Valley.

Instead, the GSL-breeze crosses the South Mountain divide, producing strong northerly flow

throughout the Rush Valley (Fig. 3c), providing an example of how an external thermally driven

wind system can overwhelm local slope and valley flows.

The evening transition period begins around sunset (~1900 LST). Within three hours of

sunset (2200 LST), downslope and down-valley flows are already well developed (Fig. 3d). In

fact, downslope and down-valley flows reach their strongest magnitude during this period. In

contrast to the morning transition which required 5-6 h to complete, the evening transition is

accomplished in only 2-3 h.

Time series from selected observing sites are presented in Figs. 3e-h. Stations on the west

side of the Tooele Valley exhibit a counterclockwise rotation of the wind with time (e.g., RES,

Fig. 3e), while stations on the east side of the valley feature a clockwise rotation (e.g., TOO, Fig.

3f). This is consistent with the conceptual model of the diurnal turning of the winds over valley

sidewalls presented by Hawkes (1947), since the thermally driven lake/land and up/down valley

circulations are in phase and are superimposed on the up/down slope circulations (Whiteman,

2000).

Stations in the middle of the Tooele Valley, such as Erda Airport (POR, Fig. 3g), are

influenced primarily by coupled lake/land and up/down valley flows. At POR the day-to-day

wind consistency at individual hours is above 90% during the nighttime hours when offshore

down-valley flow dominates, then drops to 20% at ~0900 LST due to day-to-day variations in

the time of the reversal of the lake and valley circulations. Consistency increases to near 75%

late in the afternoon (1600 LST) when the lake breeze and up-valley circulation typically are

well developed.

High consistency in nighttime downslope flow over the Tooele Valley also is evident at

Grantsville Reservoir (RES). At ~2100 LST downslope winds are developed and persist for 9-10

hours with a consistency above 90%. The strongest downslope winds (4-5 m s-1) occur during for

the first two hours of this period right after sunset (Figs. 3e,h).

Finally, there are several major canyons that issue into the Salt Lake Valley from the east.

Of these, Parley's Canyon is known to produce strong outflow into the Salt Lake Valley during

the night and early morning hours. The Parley's Canyon (UT5) time series shows that this

outflow develops typically around 1900 LST, reaches peak magnitude around 0200 LST, and

persists overnight until 0900 LST. Such outflow has a consistency of more than 90%.

b. Snake River Plain

A dense network of meteorological observing stations is found on the eastern Snake River

Plain (SRP) in Idaho at the Department of Energy's Idaho National Engineering Laboratory.

These stations have provided data for several climatological studies focused on transport and

diffusion (e.g., Clawson et al. 1989, Clements 1979, Wilkens 1955, and Yanskey et al. 1966).

The Eastern SRP generally experiences plain-parallel flows from either the southwest or the

northeast. Southwesterly flows are caused by the channeling of the prevailing westerlies by the

surrounding mountains or by daytime, thermally driven, up-plain circulations, while

northeasterly flows are caused by channeling of winds from easterly directions or by nocturnal

down-plain flows (Clawson et al.1981). The strong up-valley flow can persist for many hours

after sunset (Clawson et al. 1981). While the typical eastern SRP wind flows are well

documented, little information exists for flow patterns over the central and western SRP and

prior studies have not attempted to isolate the thermally driven flow component.

Although winds in the SRP are generally oriented along the plain's longitudinal axis, they are

also influenced by up- and downslope flows from the adjacent sidewalls and outflow from

canyons that enter the plain from surrounding mountain ranges. During the nocturnal regime

(0300 LST; Fig 4a), 2-4 m s-1 down-plain flows extend along much of the SRP. Downslope flow

and canyon outflow are observed at several locations near adjacent mountain ranges. In the

eastern SRP, winds in the central portion of the plain exhibit a northerly component, apparently

due to downslope flow from the high terrain in the northern half of the SRP.

The morning transition period begins with solar heating shortly after sunrise (~0600 LST).

By 0900 LST, several stations develop up-plain or upslope flow, although downslope and down-

plain flow persists in the other locations (Fig. 4b). Low level divergence occurs west of IDF as

upslope flows develop on both the north and south sidewalls of the valley (Fig. 4b). The

transition from down- to upslope takes 3-4 h while the down- to up-plain transition takes 5-6 h.

The slope flow transition is quicker in the SRP than in the Wasatch Front Valleys, while the

valley flow reversal is slower.

During the afternoon regime (1500 LST) the up-plain flow develops and strengthens to 3-4

m s-1 along the longitudinal axis and becomes the dominant flow pattern throughout the SRP

(Fig. 4c). Topographic channeling helps to enhance this effect. At the peak of the up-plain flow,

around 1700 LST, it diverges only weakly into the surrounding topography. Stations farther

away from the mountain ranges are the first to develop strong up-plain flow as illustrated by time

series from Howe (HOW), which is located near the Central Idaho Mountains, and Richfield

(RHF), which is closer to the mid-plain axis (Fig. 4e,f). At HOW the south-southeasterly

upslope flow develops around 1000 LST and persists until around 1400 LST when the

southwesterly up-plain flow dominates (Fig. 4e). At RHF the upslope flow also develops at 1000

LST, but is overtaken by the strong up-plain flow at 1200 LST, 2 h earlier than at HOW.

Within an hour of sunset (2100 LST), the evening transition period begins. At 2100 LST,

downslope flows and canyon outflow combine with the developing down-plain flow to produce a

low level convergence zone over the eastern SRP (Fig. 4d). The up/downslope transition takes 1-

2 hours, where as up- to down-plain transition takes 5-6 hours, eventually resulting in the

reestablishment of the evening regime (Fig. 4a).

Time series from selected observing sites in the SRP are presented in Figs. 4e-g. Consistent

with the Hawkes (1947) conceptual model for the diurnal turning of the winds over valley

sidewalls, winds rotate clockwise with time at stations on the north side of the plain (RHF, Fig.

4f) and counterclockwise with time at stations on the south side of the plain (IDF, Fig. 4g).

At RHF and HOW consistency remains above 80% through out the night, like similar

stations in the WFV region (0000-0600 LST; Figs 3g,h; 4e, f). The transition period for slope

flow lasts 3 hours at HOW (Fig 4e) whereas the valley flow transition lasts for 6 hours (e.g.,

RHF; Fig. 4f), with consistency values near or below 40%. During the daytime heating hours

(1200-1800 LST) consistency remains at moderate levels near 60% until the evening transition

period (~2100 LST) when consistency values drop to near 30% at HOW (Fig. 4e) and near 50%

at RHF (Fig. 4f). Once the nocturnal regime is developed, consistency values return to 80%,

completing the cycle.

c. Southern Nevada basin and range province

Winds within the Southern Nevada region are influenced by local thermally driven flows

(i.e., valley and slope flows) as well as regional-scale thermally driven flows (i.e., mountain-

plain circulations). In the northern half of the study area, where the National Oceanic and

Atmospheric Administration’s Air Resources Laboratory (NOAA/ARL) maintains a dense

meteorological network at the Nevada Test Site, the topography is characterized by a complex

system of mountain ranges separated by broad lowland valleys and basins (Fig. 2c). To the

south, the topography includes the Sheep Range, which reaches elevations of up to 3000 m, and

the Spring Range, which includes 3630 m Mt. Charleson. The Las Vegas Valley is located just

northeast of the Spring Range and runs southeastward to Las Vegas and Lake Mead along the

Colorado River.

Winds during the night are dominated by locally forced downslope and down-valley flows

(0300 LST, Fig. 5a). In the southern half of the study region, downslope flows from the

northeast and southwest converge into Las Vegas, while southwesterly downslope flow is

observed northeast of the Spring Range. Winds in the northern portion of the domain show little

mesoscale organization. Instead, a wide variety of wind directions and speeds are observed due

to the complexity of slope and valley flows that are produced by the complex assortment of

topographic features found in this region.

During the morning transition period, widespread southeasterly-southerly flow develops at

most southern Nevada observing sites, with weak confluence evident over the Nevada Test Site

(0900 LST, Fig. 5b). These flows are thermally-driven regional-scale plain-basin or plain-

mountain flows that develop after the nocturnal boundary layer erodes and the daytime

convective boundary layer develops in the Intermountain Basin in response to the pressure

contrast between high pressure over the eastern Pacific Ocean and the low pressure that develops

with surface heating over the interior Intermountain West. As a result, a 3-4 m s-1

southwesterly flow eventually dominates the flow pattern over the Nevada Test Site during the

afternoon regime (1500 LST, Fig. 5c). These large-scale southwesterly winds have a high

consistency that during midday exceeds 70% (e.g., DRA, Fig. 5e). In contrast, southeasterly up-

valley flow is able to develop over the Las Vegas Valley, a deep and relatively large-scale

lowland feature surrounded by very high topography, particularly to the southwest. This results

in the development of an afternoon convergence zone in the lowland region north of the Spring

Mountains. The Nellis Range observing site (A36), which is located in the central Las Vegas

valley near the base of the lower slopes of the Spring Mountains, experiences up-valley flow at

0900 LST, which reaches 3-4 m s-1 during the afternoon (Fig. 5f). The consistency of this up-

valley flow approaches 80% and, in general, afternoon wind consistencies throughout southern

Nevada are higher than those observed over the Wasatch Front Valleys and Snake River Plain

regions (cf. Figs. 3e-h, 4e-g, 5e-g). In addition, some stations, such as Area 2 (A02) at the

Nevada Test Site, have very high consistencies through most of the diurnal cycle, with only

small reductions in consistency during transition periods (Fig. 5g). This suggests that the

transitions from daytime to nighttime flow regimes at this station are regular and relatively rapid.

During the evening transition (2100 LST), the southwesterly flow weakens and nocturnal

drainage flows begin to develop (Fig. 5d). Stations near Las Vegas and within the Las Vegas

Valley are the last to complete the transition, and at 2100 LST are still experiencing strong

southwesterly winds or southeasterly up-valley flow.

d. Central Arizona area

Thermal winds within the central Arizona region can be separated into two categories, those

protected from, and those exposed to the up-valley flow from the Yuma desert. Protected

stations observe a diurnal cycle of wind consistency similar to areas in the WFV and SRP

regions. Exposed stations show similarity to those in the Southern Nevada area with wind

consistency highest during the afternoon. In this region, winds at mountain sites have a high

consistency of nocturnal downslope drainage flow (0300 LST) (Fig. 6a). Sites around Phoenix

show drainage convergence into the low lying areas of the Phoenix basin. Most drainage flows

(downslope and down-valley flows) are 1-2 m s-1. During the night, sites in the two categories

show similar conditions. Within two hours of sunrise (~0530) the morning transition begins with

a transition from downslope to upslope winds and from down-valley to up-valley winds (Fig.

6b). Downslope to upslope transition times (1 h) are comparatively quicker then up- to down-

valley transition times (3-4 h). Immediately after the transition, wind speeds pick up to 3-4 m s-1

around the Phoenix area closer to the Yuma desert while winds along the Sierra Ancha remain at

1-2 m s-1. This is the beginning step of segregation of the two categories.

Throughout the afternoon regime (1500 LST) (Fig. 6c) upslope and up-valley flows

strengthen near Phoenix and close to the west desert reaching 5 m s-1 and higher. Especially

slopes exposed to the southwest. Wind speeds at stations in the Sierra Ancha remain around 1-2

m s-1. Consistencies at Luke Air Force Base (LUF, Fig 6e) reach 70-80% in the afternoon as the

southwesterlies strengthen to 4.5 m s-1. During this time period, the two categories of stations

become apparent. Stations exposed to up-valley winds from the Yuma desert observe have

higher wind consistencies (~80%) in the afternoon than stations not exposed to this wind, which

develop local upslope and up-valley flows with consistencies of 30-50%.

As soon as the sun sets (~ 1930 LST), the evening transition period begins as upslope winds

reverse to downslope winds in the Sierra Ancha (Fig. 6d). An hour after later sunset the down-

valley wind transition begins. The slope wind transition requires about one hour whereas the

valley wind transition takes 3-4 hours. Northern stations are the first to stabilize after the

transition begins while stations near Phoenix observe westerly winds that last 1 hour longer and

then stabilize. After the transition, drainage wind speeds across the area are 1-2 m s-1 with some

mountain station winds reaching 3 m s-1.

The time series in Fig 6 provide additional information on the two categories of winds in the

Central Arizona region, with stations close to the desert exhibiting different wind patterns than

stations in the Sierra Ancha. At EEK in the mountains the downslope winds have high wind

consistency (90 %) and strengthen from 1 to 3 m s-1 throughout the nocturnal period (Fig. 6f).

During the transition from downslope to upslope, wind consistency drops and stays below 50%

during the afternoon regime, with light (1-2 m s-1) upslope flows. This is the same pattern for

many stations along the Sierra Ancha. At STT close to the desert, there is high wind consistency

(~80%) overnight with winds strengthening from 3 to 4 m s-1 (Fig. 6g). During the transition

times consistency drops below 50% and during the afternoon regime wind consistency rebounds

to 80% as upslope winds strengthen to 4.5 m s-1. Several other stations near the desert including

LUF show the high wind consistency off the desert around 13-14 LST indicating that this is an

important feature of the diurnal cycle.

Stations in the southwest Central Arizona area, like the Southern Nevada region and unlike

the WFV and SRP regions, observe 80% wind consistency in the afternoon. The area is similar

to Southern Nevada, but with a desert to the southwest. The thermal wind cycle is heavily

influenced by up-valley winds from the Yuma desert. Stations within the northeastern part of the

Central Arizona Region experience the typical diurnal cycle of wind consistency seen in the

WFV and SRP regions.

5. Discussion

It is interesting to note that well-developed, coupled thermal wind systems have high

consistency in the middle of a broad valley or plain, as illustrated with POR in the Tooele Valley

8 km from the eastern sidewall and 14 km from the western side wall (Fig. 3e). At this site,

winds flow only in two directions with a 180 degree wind reversal. The consistency is high (~90

%) during the night and is moderately high (~60%) during the day.

Large valleys with a constricted outlet can produce strong drainage winds because of

convergence at the outlet. Several stations illustrate this, including UT5 in the WVF (Fig. 3h)

and HOW in the SRP (Fig. 4e).

Lack of data makes it difficult to determine whether the down-plain flow is continuous along

the central valley floor along the entire SRP. Horse Butte (HBT) never exhibits an easterly wind

component, while an easterly component is visible at Boise (BOI) on the west end of the plain

and at RHF on the east end of the plain. Several stations near HBT that did not have an

acceptably long period of record to be included in the figures exhibit results similar to HBT.

HBT, however, is on a hill 300 m above the Snake River. East of HBT, the river turns to the

northwest. Drainage flows from the mountains to the south of HBT and from the buttes to the

west flow to the east into the river gorge, converging with drainage winds from the eastern SRP

and flowing north around the river meander, then continuing westward towards BOI.

6. Conclusions

Diurnal wind patterns and day-to-day wind consistency were analyzed to provide an

improved understanding of the wind patterns and their evolution in the Salt Lake, Tooele and

Rush Valleys, the Snake River Plain, Southern Nevada, and Central Arizona area. Archived

MesoWest data were used for this analysis, and criteria were created to select thermally driven

wind days that could be composited into a mean diurnal pattern. Day-to-day wind consistency

patterns and time series were used to investigate the winds further.3 The wind fields were

animated in video loops to better visualize the diurnal turning of the winds and the interactions

between the various thermally driven wind systems.

The timing of reversals is consistent from area to area. During the morning transition period,

downslope winds changed to upslope winds shortly after sunrise and then the down-valley winds

reversed to up-valley winds. The evening transition period started within an hour after sunset in

all areas. Downslope winds were strongest during the first few hours, and were generally

stronger than the daytime upslope winds. In each of the regions investigated, except as

3 Animations are available at http://www.met.utah.edu/jimsteen/jstewart/mtnwind.html

mentioned in section 5, day-to-day wind consistency on thermally driven wind days was

generally high (~90%) at night, moderate (~60%) during the day, and low (~20%) during the

transition periods. The drop in consistency during transition periods reflects the variation in the

timing of wind reversals from day to day. The moderate consistency throughout the day is

attributed to wind speed and direction variations caused by convection. Diurnal wind rotations on

valley sidewalls were observed in all areas, supporting the conceptual model by Hawkes (1947).

Some stations in Southern Nevada and Central Arizona experienced regional daytime

southwesterly winds that increased their consistency to about 80%.

While there are marked similarities among the diurnal wind patterns and timing of wind

reversals in the areas investigated, each of the areas exhibited unique wind features. Within the

WFV, the land-lake breeze interacts with the other thermally driven wind systems. Within the

SRP there is a distinct channeling effect caused by the terrain. Both Southern Nevada and

Central Arizona thermal winds are strongly controlled in the afternoon by regional southwesterly

winds that blow into the interior of the Intermountain Basin.

Within the southwestern IW, another important thermal wind is observed. Up-valley flow

from the deserts to the southwest of the Southern Nevada and Central Phoenix regions are

observed to reach over 5 m s-1 during the afternoon hours (1300 - 1500 LST) with wind

consistency around 80%. The afternoon winds in these regions are about 20% more consistent

then the afternoon winds in the SRP or WFV regions. Stations such as A02 in the Southern

Nevada region maintain wind consistency over 60% for the entire diurnal cycle with downslope

winds during night and the southerly up-valley winds from the desert during the day (Fig. 6g).

In the Central Arizona region, stations fall into one of two different categories -- those affected

by up-valley winds from the Yuma desert and those that are not. Sites protected from the flow

off the Yuma desert observed thermal winds similar to those in the WFV and SRP regions,

whereas the exposed stations exhibit the higher wind consistencies mentioned above.

These and future analyses will be of value to boundary-layer and air pollution meteorologists,

fire weather forecasters, operational meteorologists and others who are interested in

understanding wind flow patterns in complex terrain.

6. Acknowledgements

MesoWest data were collected and processed by John Horel, Mike Splitt, and Bryan White of

the University of Utah and Larry Dunn and David Zaff of the National Weather Service's Salt

Lake City Forecast Office. We gratefully acknowledge the contribution of organizations

participating in MesoWest. One author (JQS) wishes to acknowledge the support of the Global

Change Education Program for supporting his summer appointment at Pacific Northwest

National Laboratory as part of their Summer Undergraduate Research Education program. This

research was supported by the U.S. Department of Energy, Office of Biological and

Environmental Research, Environmental Sciences Division as part of their Environmental

Meteorology Program under Contract DE-AC06-76RLO 1830 at PNNL. The U. S. Department

of Energy's PNNL is operated by Battelle Memorial Institute.

Figures

Figure 1. Topography of the western United States with study regions annotated. Elevation (m) based on scale at upper right.

Figure 2. Places and names for each study region for reference. Terrain shading as in Fig. 1. (a) Wasatch Front Valleys. (b) Snake River Plain. (c) Southern Nevada basin and range province. (d) Central Arizona.

Figure 3. Mean summer season vector winds in the Wasatch Front Valleys. Full and half barb denote 1 and 0.5 m s-1, respectively. Terrain shading as in Fig. 1. (a) 0400 LST. (b) 1000 LST. (c) 1600 LST. (d) 2200 LST (e) RES time series. (f) TOO time series. (g) POR time series. (h) UT5 time series.

Figure 4. Mean summer season vector winds in the Snake River Plain. Full and half barb denote 1 and 0.5 m s-1, respectively. Terrain shading as in Fig. 1. (a) 0300 LST. (b) 0900 LST. (c) 1500 LST. (d) 2100 LST (e) IDF time series. (f) HOW time series. (g) RHF time series.

Figure 5. Mean summer season vector winds in the Southern Nevada basin and Range Province. Full and half barb denote 1 and 0.5 m s-1, respectively. Terrain shading as in Fig. 1. (a) 0300 LST. (b) 0900 LST. (c) 1500 LST. (d) 2100 LST (e) DRA time series. (f) A36 time series. (g) A02 time series.

Figure 6. Mean summer season vector winds in Central Arizona. Full and half barb denote 1 and 0.5 m s-1, respectively. Terrain shading as in Fig. 1. (a) 0300 LST. (b) 0900 LST. (c) 1500 LST. (d) 2100 LST (e) LUF time series. (f) EEK time series. (g) STT time series.

REFERENCES

Astling, E. G., 1986: A study of mesoscale wind fields in mountainous terrain. US Army, Dugway Proving Ground, Dugway, Utah 84022, TECOM Project No. 7-CO-R87-DP0-009.

Carter R. G. and R. E. Keislar, 2000: Emergency response transprt forecasting using historical wind field pattern matching. J. Appl. Metor., 39, 446-462.

Cate, J. H., 1977: Complex airflow over the upper Snake River Plain. Joint Conf. On Appl. Of Air Poll. Meteor., Nov 29- Dec 2. Salt Lake City, UT. Amer. Meteor. Soc., Boston, Massachusetts, xxx-xxx.

Clawson, K. L., G. E. Start, and N. R. Ricks, 1989: Climatography of the Idaho National Engineering Laboratory. 2d ed. NOAA/Air Resources Laboratory., 155 pp.

Clements, W. E., 1979. Experimental design and data of the April, 1977 multitracer atmospheric experiment at the Idaho National Engineering Laboratory. LA-7795-M5. Los Alamos Sci. Lab., Los Alamos, NM.

Gudiksen, P. H., and J. J. Walton, 1981: Categorization of nocturnal drainage flows in the Anderson Creek Valley. Preprints, Second Conference on Mountain Meteorology, 9-12 November 1981, Steamboat Springs, Colorado. Amer. Meteor. Soc., Boston, Massachusetts, 218-221.

Gudiksen, P. H.,1989: Categorization of nocturnal drainage flows within the Brush Creek Valley and the variability of sigma theta in complex terrain. J. Appl. Meteor., 28, 489-495.

Hawkes, H. B., 1947: "Mountain and valley winds-with special reference to the diurnal mountain winds of the Great Salt Lake region," Ph.D. dissertation, Ohio State University, 312 pp.

Horel et al. 2000

Panofsky, H. A., and G. W. Brier, 1965: Some Applications of Statistics to Meteorology. Pennsylvania State University, University Park, Pennsylvania, 224pp.

Smidy, K. I., 1972: Diurnal and seasonal variation of the surface wind in the Salt Lake Valley," M.S. thesis, University of Utah.

Splitt, M. E., and J. Horel, 1998: Use of multivariate linear regression for meteorological data analysis and quality assessment in complex terrain. Tenth Symposium on Meteorological Observations and Instrumentation. Phoenix, AZ, Amer. Meteor. Soc., Boston, Massachusetts, xxx-xxx.

Steenburgh, W. J., S. F. Halvorson, and D. J. Onton, 2000: Climatology of lake-effect snowstorms of the Great Salt Lake. Mon. Wea. Rev., 128, 709-727.

Stone, G. L., and D. E. Hoard, 1990a: Daytime wind in valleys adjacent to the Great Salt Lake. Preprints, Fifth Conf. on Mountain Meteorology, Boulder, CO. Amer. Meteor. Soc., Boston, MA, 209-215.

Stone, G. L., and D. E. Hoard, 1990b: An anomalous wind between valleys - Its characteristics and a proposed explanation. Preprints, Fifth Conf. on Mountain Meteorology, 25-29 June,1990. Boulder, CO. Amer. Meteor. Soc., Boston, MA, 209-215.

Tucker, D. F., 1997: Surface mesonets of the Western United States. Bull. Amer. Meteor. Soc., 78, 1485-1495.

Wasatch Front Regional Council, 1980. Great Salt Lake Air Basin Wind Study. Wasatch Front Regional Coucil, Bountiful, UT 84010.

Wendell, L. L., 1972: Mesoscale wind fields and transport estimates determined from a network of wind towers. Mon. Wea. Rev., 100, 565-578.

Whiteman, C. D., and K. J. Allwine, 1986: Extraterrestrial solar radiation on inclined surfaces. Environmental Software, 1, 164-169

Whiteman, C. D., 1990: Observations of thermally developed wind systems in mountainous terrain. Atmospheric Processes over Complex Terrain, Meteor. Monograph., No. 45, Amer. Meteor. Soc., 5-42.

Whiteman, C. D., X. Bian, and J.L. Sutherland, 1999: Wintertime surface wind patterns in the Colorado River valley. J. Appl. Meteor., 38 1118-1130.

Whiteman, C. D., 2000: Mountain Meteorology: Fundamentals and Applications. Oxford University Press, New York, 355pp.

Wilkens, E. M., 1955: A discontinuity surface produced by topographic winds over the upper Snake River Plain, Idaho. Bull. Amer. Meteor. Soc., 36, 397-408.

Yanskey, G. R., E. H. Markee, Jr., and L. P. Richter, 1966: Climatography of the National Reactor Testing Station. USAEC Rep. IDO-12048. U.S. Dept. of Comm., Environ. Sci. Serv. Admin., Inst. For Atmos. Sci, Air Resour. Field Res. Off., Idaho Falls, ID.

TABLES

1. Solar radiation and 700 hPa wind observation sites.Region Solar Observation Site(lon/lat) 700 hPa Wind Observation(lon/lat)

Wasatch Front Valleys WBB (-111.85, 40.77) SLC ( -111.97, 40.78)Snake River Plain RID (-112.94, 43.59) BOI ( -116.22, 43.57)Southern Nevada DRA (-116.02, 36.62)Central Arizona TUS (-110.93, 32.12)