National Weather Service United States Department of Commerce


James A. Reynolds
WFO Medford, Oregon

Val J. MacBlain
WFO El Paso, Texas


The Franklin Mountains of the El Paso, Texas region (Fig. 1) are a north-south oriented mountain chain that is approximately 23.1 km (14.4 statute miles) long and 5.0 km (3.1 statute miles) wide (Harbour, 1972). This mountain chain rises more than 3000 feet out of the surrounding desert region (Fig. 2) with one peak reaching a maximum altitude of 7200 feet above mean sea level. They are asymmetrical with a steeper grade on the west side of the range than on the east. The Franklins create a divide between the western one-third of the city of El Paso and the central and eastern two-thirds of the city.

For years, meteorologists and lay-people alike, have marveled at and pondered the intriguing wind patterns that are generated by the Franklin Mountains and the effects that these winds have on the city of El Paso. Indeed, it wasn't until the relocation of the El Paso National Weather Service (NWS) office from the El Paso International Airport (east of the Franklins) to its current location in Santa Teresa, New Mexico (west of the Franklins) in the mid 1990's that an even greater appreciation for these winds was attained. Beginning with the relocation, it was often observed that, particularly between October and May, strong and potentially damaging winds could occur on the east side of the range with little wind on the west side during the same time. Also, temperature differences of upwards of 20 degrees were noted, at times, between both sides of the mountains in conjunction with this phenomenon. Obvious concerns for temperature forecasting and airport operations necessitated further research into this terrain-modified flow.


Data from several atmospheric soundings taken at the El Paso NWS office during significant east-side wind events, and also during times when no east-side winds were occurring, were inspected to determine clues to the existence of these terrain-modified winds.


The disparity between wind speeds and temperatures on either side of the Franklin Range during strong east-side wind events clearly indicated that downslope wind processes were at work in these situations. However, it was not always exactly clear why these downslope winds occurred. Review of El Paso atmospheric sounding data quickly yielded three equally important clues as to the occurrence of these downslope winds (Fig. 3). First, the direction of the winds between roughly 850 and 650 mb was usually quite uniform throughout this layer and generally blew from between 225 degrees and 315 degrees during any single event. Second, the average speed of the wind throughout the layer was generally 25 knots or greater and wind speeds in this layer slowly increased with height. Third, a strong inversion was found, usually between 600 and 700 mb. Thus, for northwest to southwest flow, normal to the range, the air in a layer between 850 mb and the inversion was being forced upward over the mountain chain, causing acceleration resulting in stronger winds and warmer temperatures as the layer descended the east slopes. If the inversion formed much higher or much lower than the previously defined range from 600 to 700 mb, the chances were much less likely that downslope winds would occur that evening.

These three sounding characteristics, however, only explained how strong winds blew on the east side of the range but did not help to account for the absence of significant wind on the west side of the mountains during these events. Through observation it appeared that the downslope wind events had the best chance of occurring from early evening on, when the boundary layer was first decoupled from stronger winds aloft. This decoupling appears to account for the lack of wind on the west side of the range.

It is important to note that the appearance of the three aforementioned elements found in the El Paso atmospheric soundings did not guarantee that a downslope event would occur. After further observation it was noted that a surface boundary of some sort must be in the general vicinity of the Franklin Range for these winds to occur. Presumably, a sufficient low level pressure gradient was necessary to help draw the winds from the west side of the range up and over the range to the east side.

Background and Theory

According to Whiteman (2000), the downslope flows experienced on the east side of the Franklin Mountains are a classic textbook example of a terrain-forced flow. By definition, terrain-forced flows are produced when large-scale winds are modified or channeled by the underlying complex terrain. Moderate to strong cross-barrier winds are necessary to produce terrain-forced flows, which occur most frequently in areas of cyclogenesis, or where low pressure systems, or jet streams are commonly found.

A flow approaching a mountain barrier will most likely go over the barrier rather than around it if the barrier is long, if the cross-barrier wind component is strong, and if the flow is unstable, near- neutral, or only weakly stable. These conditions are frequently met in the United States because the long, north-south oriented mountain ranges lie perpendicular to the prevailing westerly winds and the jet stream. Mountain barriers orientated perpendicular to the flow cause the highest accelerations across the barrier and frequently generate lee waves downwind of the obstacle as well as downslope windstorms.

Whiteman (2000) continues by saying that downslope windstorms occur on the lee side of high- relief mountain barriers when a stable air mass is carried across the mountains by strong cross- barrier winds that increase in strength with height. The strong winds are caused by intense surface pressure gradients, with a high pressure center on the upwind side of the barrier and a low pressure trough paralleling the lee foothills. The cross-barrier surface pressure gradient is intensified as the descending air on the lee side of the barrier produces local warming and, thus, a further decrease in pressure at the surface. The pressure gradient may be intensified further if the windstorm coincides with the arrival of a short-wave trough which causes the surface pressure to decrease on the lee side of the barrier relative to the windward side. The shortwave trough can also cause the winds at mountaintop level to shift direction and become more perpendicular to the barrier. Furthermore, elevated inversions have been noted in many windstorms where observations were available near mountaintop level. However, since elevated inversions and their precise height are usually difficult to observe and to forecast in real time, their presence is usually assumed when all other meteorological conditions for windstorms are met. Nevertheless, this assumption may result in the overforecasting of windstorms, depending on the exact height of the inversion.

Earlier findings by Durran (1990) agree with those of Whiteman. Durran notes that, where there is a deep cross-mountain flow and no mean-state critical layer, observational evidence suggests that conditions favorable for downslope winds occur when:

(i) The wind is directed across the mountain (roughly within 30� of perpendicular to the ridgeline) and the wind speed at mountaintop level exceeds a terrain dependent value of 7 to 15 m/s.

(ii) The upstream temperature profile exhibits an inversion or a layer of strong stability near mountaintop level (Colson 1954; Brinkmann 1974).

Flow speeds in downslope windstorms are highest in a narrow zone along the base of a mountain barrier. The highest wind speeds occur on elevated mesas or other terrain projections on the edge of mountains. Strong, gusty downslope windstorms can cause considerable damage along the mountain-plain interface. Damaging winds rarely extend more than 15 miles out onto the adjacent plain, although winds can still be strong at these distances. The strong winds pose a number of hazards. Open fires can be spread rapidly and smoke, windblown dust, and strong gusts can cause poor driving conditions. Aviation hazards include turbulent rotors that develop below the crest of lee waves and in the lee of the hydraulic flow cavity.

Durran cites three possible explanations for the production of severe downslope winds. The first of the three conceptual models was proposed by Long (1953). Long suggested that there is a fundamental similarity between downslope windstorms and hydraulic jumps. He states that if there is a sufficient acceleration in the stationary gravity wave; i.e., a sufficient increase in velocity and decrease in thickness as the fluid ascends toward the crest of a mountain ridge, a transition from subcritical (i.e. Froude number, FR<1)to supercritical (FR>1) flow occurs at the top of the ridge. Since the flow along the lee slope is supercritical, the fluid accelerates further as it descends the mountain. Finally, the readjustment of the fluid to ambient conditions, farther downstream, is accomplished through a turbulent hydraulic jump. Very high velocities are produced as potential energy is converted to kinetic energy, providing acceleration, on both the windward and leeward sides of the mountain. This hybrid case differs from the pure subcritical flow case, where the air accelerates on ascent and decelerates on descent, or the pure supercritical flow case, where the reverse occurs.

A second explanation, by Eliassen and Palm (1960), suggests that downslope windstorms are produced by large-amplitude vertically propagating mountain waves. They showed that when an upward propagating linear gravity wave encounters a region in which the Scorer parameter changes rapidly, part of its energy can be reflected back into a downward propagating wave. (The Scorer parameter measures the ratio of bouyancy to mean flow, adjusted by a term dependent on the vertical wind profile.) In later work, Klemp and Lilly (1975) examined the case of small-amplitude hydrostatic waves in a stratified atmosphere, each layer having constant stability and wind shear. They suggested downslope windstorms occur when the atmosphere is "tuned" so that the partial reflections at the boundaries between layers cause reinforcement between upward and downward waves. They found that a tropopause height equal to one-half vertical wavelength was key in optimizing this "tuning".

The third explanation of strong downslope winds was proposed by Clark and Peltier (1977, 1984), Peltier and Clark (1979, 1983), and Clark and Farley (1984). All found significant increases in the lee-slope winds subsequent to the vertically propagating waves becoming statically unstable and breaking. The strong mixing, due to this wave breaking, causes a local reversing of the cross-mountain flow and results in a critical layer (i.e., where Richardson number <0.25). This critical layer effectively serves as an upper boundary, reflecting upward-propagating waves back toward the surface. Drawing on the concept of tuning mentioned above, Peltier and Clark determined that an appropriate depth between the critical layer and the mountain would result in a resonant wave that would amplify with time, producing very strong surface winds.

Interestingly, research by Smith (1977, 1985) and Durran (1986) suggests that the factors listed above may be interrelated in causing downslope wind events, rather than contradictory theories.


The east side of the city of El Paso, Texas is quite often subjected to downslope wind events which are one type of a terrain-forced flow. These flows are produced locally by the Franklin Mountains and typically occur from late fall through spring. The flows are dependent on the presence of a line of mountains which are perpendicular or nearly perpendicular to the dominant westerlies over the area and which are long enough to prevent wind from simply going around the line, a strong pressure gradient resulting in a strong cross-barrier flow, and an elevated inversion.

These downslope flows often have a dramatic effect on ambient air temperatures and airport operations in El Paso. Downslope events in the area can easily be overforecasted due to the variable heights at which elevated inversions may form in the area.


Thanks to Todd Hall, WFO El Paso, Texas for his help in locating the graphics used in this paper.


Brinkmann, W.A.R., 1974: Strong Downslope Winds At Boulder, Colorado. Mon. Wea. Rev., 102, 592-602.

Clark, T.L., and R.D. Farley, 1984: Severe Downslope Windstorm Calculations In Two And Three Spatial Dimensions Using Anelastic Interactive Grid Nesting: A Possible Mechanism For Gustiness. J. Atmos. Sci., 41, 329-350.
------, and ------, 1977: On The Evolution And Stability Of Finite-Amplitude Mountain Waves. J. Atmos. Sci., 34, 1715-1730.
------, and ------, 1984: Critical Level Reflection And The Resonant Growth Of Nonlinear Mountain Waves. J. Atmos. Sci., 41, 3122-3134.

Colson, D., 1954: Meteorological Problems In Forecasting Mountain Waves. Bull. Amer. Meteor. Soc., 35, 363-371.

Durran, D.R., 1986: Another Look At Downslope Windstorms. Part 1: On The Development Of Analogs To Supercritical Flow In An Infinitely Deep, Continuously Stratified Fluid. J. Atmos. Sci., 43, 2527-2543.
------, D.R., 1990: Atmospheric Processes Over Complex Terrain, Meteorological Monographs, 23, 66-69, 75-76.

Eliassen, A. and E. Palm, 1960: On The Transfer Of Energy In Stationary Mountain Waves. Geofys. Publ., 22, 1-23.

Harbour, R. L. 1972. Franklin Mountains, Texas And New Mexico. USGS Bulletin 1298. U. S. Government Printing Office, Washington, D.C.

Klemp, J.B., D.K. Lilly, 1975: The Dynamics Of Wave Induced Downslope Winds. J. Atmos. Sci. 32, 320-339.

Long, R.R., 1953: A Laboratory Model Resembling The "Bishop-Wave" Phenomenon. Bull. Amer. Soc., 34, 205-211.

Peltier, W.R., and T.L. Clark, 1979: The Evolution And Stability Of Finite-Amplitude Mountain Waves. Part II: Surface Wave Drag And Severe Downslope Windstorms. J. Atmos. Sci., 36, 1498-1529.
------, and ------, 1983: Nonlinear Mountain Waves In Two And Three Spatial Dimensions. Quart. J. Rev. Meteor. Soc., 109, 527-548.

Smith, R.B., 1977: The Steepening Of Hydrostatic MountainWaves. J. Atmos. Sci., 34, 1634-1654.
------, 1985: On Severe Downslope Winds. J. Atmos Sci., 42, 2597-2603.

Whiteman, C. D. 2000. Mountain Meteorology: Fundamentals and Applications, 142-154