SEVERE DOWNSLOPE WIND STORMS

A Primer and Case Study of the February 3, 1998 Gabbs, Nevada Event

Kim Runk

 

 

1.                  INTRODUCTION

 

During the Winter and Spring months, windy days occur fairly regularly in Nevada, and indeed across the entire Great Basin.  Prolonged periods of strong, gusty wind in our area are typically caused by some combination of concentrated pressure gradient force and gap flow.  But occasionally conditions favor a more uncommon and very damaging windstorm that can be characterized as a mountain wave event.  Such was the case in the central Nevada mining town of Gabbs (Fig.1) on February 3, 1998. 

 

(Click on thumbnail to view full size image)

 

Fig. 1 – Topographic map of west central Nevada.  The mining town of Gabbs is on the northwest slopes of the

Paradise Range, and is located approximately 90 miles southeast of Reno.  

 

 

A downslope windstorm occurred early that morning along the western side of the Paradise Range, producing sustained winds estimated at 70-80 mph with gusts approaching 100 mph in Gabbs.  Several mobile homes were either overturned or blown off their moorings, numerous mature trees were completely uprooted, and there was widespread structural damage to small buildings around the mining facility (Fig. 2).  Most of the community was awakened from sleep by the storm as the strongest winds occurred between the hours of 3:00-6:00 a.m. PST.  Residents reported they had never encountered wind of that intensity before, describing the experience as frightening.   

 

(Click on thumbnail to view full size image)

 

Fig. 2 – Damage photos in Gabbs the morning after a 100-mph downslope wind event in February 1998.

 

 

This case study will document the windstorm as it appeared in analyses and forecast products centered around 1200 UTC, February 3, 1998, highlighting the most significant aspects of attempting to forecast such an event.  It is important to note that the available NCEP model guidance was unable to resolve several crucial details surrounding this event=s development, primarily owing to three deficiencies:

 

(1)        the highest horizontal resolution of any operational numerical guidance at the time was the 32-km Eta; a grid spacing that is generally too coarse to resolve the most useful and identifiable forecast parameters;

(2)        the operational Eta is a hydrostatic model, and some of the most important processes involved in generating downslope windstorms are non-hydrostatic.

(1)               the Eta step coordinate has been shown to be ineffective at generating mountain wave features, even when run experimentally at very high resolution.

 

These shortcomings notwithstanding, it is hoped that this discussion will provide sufficient background information that can be used to infer the presence or development of the key precursors when confronting similar situations in the future.  In order to gain insight into the processes surrounding the storm=s development, a high-resolution modeling simulation was conducted after the fact, using a non-hydrostatic, two-way interactive nest configuration of RAMS at 4-km horizontal grid spacing on the inner nest (Fig. 3).   Output from this simulation will be used to complement the Eta guidance in order to identify and explain the most important forecast parameters for severe mountain wave-type downslope windstorms.

 

(Click on thumbnail to view full size image)

 

Fig. 3 – Plan view depiction of double nest configuration for RAMS simulation of Gabbs wind event.

Horizontal grid spacing for innermost nest is 4 km. Outer nest is 12 km. Initialization and boundary

conditions were input from the 29-km MesoEta.

 

 

 

1.                  KEY CONCEPTS OF DOWNSLOPE WINDSTORMS

 

Downslope windstorms have been the topic of numerous studies, ranging from dynamical assessments (e.g., Klemp and Lilly, 1975; Smith, 1985) to detailed investigations of specific cases (e.g., Mass and Albright, 1985; Neiman et al., 1988).  From a forecaster=s perspective, one of the most practical references published in recent years was a two-part paper by Colle and Mass (1998), studying windstorms along the slopes of the Washington Cascades.  In this study, the authors identified four significant factors associated with the development of major downslope wind events:

 

(1)        Strength of the cross‑barrier flow

(2)        Magnitude of the cross‑barrier pressure gradient (measured as MSLP)

(3)        Presence of a critical level

(4)        A stable layer near ridge crest with lower stability above

 

All four factors were evident in the strongest downslope wind events, although the first two factors accounted for a significant amount of variance in the strength of observed winds.  For situations characterized by strong cross-barrier flow, the presence of a critical level was found to be crucial to the development of mountain wave-type storms.   However, it was noted that wave formation results in warming of the column and lowers the surface pressure in the lee of the mountains, which increases the cross‑barrier pressure gradient. So, a wave‑type event can actually induce stronger gap flow. 

 

Since the presence of a so-called Acritical level@ has been identified as a fundamental factor for mountain wave development, it=s important to understand what it is.  When strong flow encounters a topographic barrier, a vertically propagating gravity wave is generated.  Sometimes this energy is simply dispersed as turbulence.  But sometimes the gravity wave is prevented from continuing its ascent and expansion, and the energy is deflected and redirected toward the surface.  A critical level (or more accurately, a critical layer) is the region in the atmosphere that prevents the gravity wave energy from continuing upward.  This is key to mountain‑wave formation.  This process of redirecting packets of concentrated wave energy downward is why it is possible to produce surface wind speeds at the base of the mountains far exceeding wind speeds observed at any level in the free atmosphere.

 

If a critical level exists in the synoptic scale observed or forecast data, it is said to be a Amean-state@ critical level.  A mean‑state critical level is typically defined as a point in the atmosphere where the cross‑barrier flow goes to zero.  This can be a place where the winds become very light, or where the winds become parallel to the barrier.  In ideal situations, it occurs in a region where a wind direction reversal takes place in the vertical (e.g., going from easterly to westerly flow). 

 

Another important aspect of a critical level is that it is found within a layer of relatively unstable air aloft which is superimposed on a stable layer located near the ridge-crest.  A fluid oscillation can take place in the stable layer, but not in the unstable layer, so the energy from the oscillation (a gravity wave) cannot propagate upward.  An idealized diagram of this configuration is shown in Figure 4.

 

(Click on thumbnail to view full size image)

 

Fig. 4 – Idealized cross section diagram of a downslope wind event on the western slopes of a north-south oriented mountain

range (view is to the north). Vertical scale is pressure in mb.  Black lines are isentropes.  Shaded areas represent relative

wind speed (e.g., red > 75kts, blue > 55 kts, green > 35 kts).

 

 

Wave breaking seems to be enhanced in the presence of a cross‑barrier flow that decreases with height, also known as Areverse shear@.  It is thought that reverse shear can lead to wave breaking even in the absence of a mean‑state critical level.  Such a structure can be visualized as the turbulence generated by a breaking ocean wave.  As the wave crests and breaks, a region of turbulence is created where the flow may either reverse or go to zero.  This region acts to prevent energy from propagating through it.  This specialized type of critical level is referred to as a Aself-induced@ critical level.  It can be generated by the mountain wave circulation itself and occurs on a much smaller scale than the mean-state type.  In fact, a self-induced critical level can only be observed with a high-resolution network of equipment such as that which is available in a field experiment.

 

A closer look at the Gabbs example should help to clarify these concepts.

 

2.                  SYNOPTIC SETTING

 

Unfortunately it=s difficult to assess the actual local flow conditions at analysis time since the area of interest is well out of radar range, and a considerable distance from the nearest radiosonde site.  In addition, there are no automated observation platforms near Gabbs, nor were there any trained weather observers in the area.  Thus much of what occurred has to be inferred from the model forecast data, making the assumption that it is a reasonable representation of what really happened.  This is not the best method for evaluating an event such as this, but in this case there are some useful insights to be gained by starting with this premise.  

 

A deep trough was moving toward the Pacific coast with a vertically stacked closed low forecast to form about 150 miles west of Eureka by 1200 UTC, 03 February 1998 (Fig. 5).  Jet level energy was lifting out of the base of the trough with strong southwest flow aloft spreading across Nevada.  Pressure tendencies seen in the METAR observations, and captured by the RUC Surface Analysis, showed pressure rises in eastern Utah and northeast Arizona, while pressure falls were occurring over western Nevada and eastern California (Fig. 6).  A three-hour fall maximum was located over Pershing County in western Nevada.   The prevailing low level wind was directed from southeast to northwest, and the mean sea level pressure gradient was forecast to strengthen across Nevada to around 12-15mb from LAS-RNO by 1200 UTC.  A region of strong dynamic ascent was forecast in the 700-500mb layer over central California along and west of the Sierra (Fig. 7).  A weaker wave lifting out of the mean trough position was forecast to generate a modest up-down couplet of forcing for vertical motion in northwestern Nye County near the area of interest with an accompanying cross-barrier flow of 25-30 knots at 850mb and about 45 knots at 700mb.

 

(Click on thumbnail to view full size image)

 

 

Fig. 5 – Eta 12h 500mb height/vorticity forecast, valid 1200 UTC, February 3, 1998.

 

 

 

 

 

(Click on thumbnail to view full size image)

 

 

Fig. 6 – RUC surface pressure and 3h change analysis, valid 1100 UTC, February 3, 1998.

 

 

 

 

 

(Click on thumbnail to view full size image)

 

 

Fig. 7 – Eta 700-500mb omega forecast, valid 1200 UTC, February 3, 1998.

 

 

 

 

A northwest-to-southeast MesoEta cross-section between Fallon (NFL) and Tonopah (TPH) is depicted in Figure 8.  While there is a significant cross-barrier flow in the lower levels, and some suggestion of an unstable mid-tropospheric layer (as evidenced by a relatively large separation in the isentropes) above a marginally stable mountain-top layer, it does not reveal any compelling evidence that points to the presence of a mean-state critical level for mountain wave development.  The unusually strong sea‑level pressure gradient was sufficient to raise concerns for the likelihood of wind advisory criteria being exceeded, but the potential for a major mountain wave event was not recognized.

 

(Click on thumbnail to view full size image)

 

Fig. 8 – Eta cross section of isentropes and winds from Fallon NAS, NV (NFL) to Tonopah, NV (TPH), valid 1200 UTC, February 3, 1998. Gabbs is approximately half way between the two end points. 

 

 

 

 

3.                  MESOSCALE SIMULATION

 

A similar cross-section from the high-resolution RAMS simulation (Fig. 9) reveals a more detailed structure that might have provided needed clues for the forecaster to be wary of an impending mountain wave event.  The layer of moderately unstable air between about 550-300mb is comparable to the one shown by the MesoEta, but the stable layer beneath it (which tops out just above ridge-crest level) is slightly more pronounced in the RAMS output.  There is no flow reversal per se, but there is a well-defined windspeed minimum accompanied by a deep wave in the isentropes directly above Gabbs.  This region of relatively light mid-tropospheric winds coincides with the point where the mean wind became parallel to the central Nevada mountain ranges.  As a result, winds tangential to the plane of the cross-section diminished to less than 20 knots at 350mb.  This region of low-speed winds embedded in an unstable layer represents the critical level.  This feature along with the accompanying deep wave in the isentropes are important indicators of mountain wave formation.  Because RAMS was able to resolve these key features, it developed a much stronger low-level wind over the area of interest, indicating a sustained wind of 38 ms-1 (~80 knots) in the Gabbs area (Fig. 10).

 

(Click on thumbnail to view full size image)

 

Fig. 9 – RAMS cross section (inverted from Fig 8), depicting isentropes (black) and wind speeds (shaded with white contour lines). Note improved resolution of stable layer near mountain tops, presence of a critical layer centered near 360mb, and defined wave action in isentropes with accompanying descent of high speed winds.

 

 

 

 

(Click on thumbnail to view full size image)

 

Fig. 10 –  Plan view of RAMS 4 km wind forecast, valid 1200 UTC, February 3, 1998.  Grid point nearest Gabbs

suggests wind speeds in the 70-75 kt range. 

 

 

In eastern California and extreme southern Nevada, these conditions would manifest themselves in a strong upper tropospheric westerly flow against the Sierra or the Spring Mountains.  Such a case occurred in the Mt Charleston area on November 18, 1998 when an unusually strong wind was reported by both the Nevada Division of Forestry at Mt Charleston and by the U.S. Forest Service at Red Rock Canyon.  The cross-section from a 10km RAMS simulation (Fig. 11) indicates a stable layer from surface to mountain-top beneath a more unstable layer with up-down couplets and a wind reversal (east near the surface to west-southwest aloft).  In this case, the strong winds were localized to the lee of the Spring Range, with reports suggesting about 60-70 mph sustained, much higher than the gradient would support.  Localized high wind events are undoubtedly far more common along the eastern Sierra slopes, but lack of observational data makes it difficult to verify, or even build a climatology of key antecedent conditions based on actual cases. 

 

 

(Click on thumbnail to view full size image)

 

Fig. 11 – Cross section of Spring Mountains from operational 10 km RAMS simulation, valid 0000 UTC, November 12, 1998.  Winds reversed from easterly near the surface to westerly aloft through a mountain top stable layer.  Wave action was suggested in advance by isentropes and vertical velocity signatures.  Sustained winds along the lee slopes of Mt Charleston exceeded 60 kts.  

 

 

4.                  CONCLUSION

 

Despite these limitations, it is possible to identify and predict significant downslope wind events in remote areas, given an understanding of the physical mechanisms that generate such events.  The conditions outlined and illustrated here should provide forecasters the basic tools to assess similar events in the future. 

 

 

*********************

 

 

BIBLIOGRAPHY

 

Colle, B.A., and C.F. Mass, 1998: Windstorms along the western side of the Washington

Cascades.  Part I: A high resolution observational and modeling study of the 12 February 1995 event.  Mon. Wea Rev., 126, 28-52.

Colle, B.A., and C.F. Mass, 1998: Windstorms along the western side of the Washington Cascades.  Part II: Characteristics of past events and three-dimensional idealized simulations.  Mon. Wea Rev., 126, 53-71.

 

Klemp, J.B., and D.K. Lilly, 1975: The dynamics of wave induced downslope windstorms. 

J. Atmos. Sci., 32, 320-339.

 

Mass, C.F., and M.D. Albright, 1985: A severe windstorm in the lee of the Cascade mountains of Washington state.  Mon. Wea Rev., 113, 1261-1281.

 

Neiman, P.J., R.M. Hardesty, M.A. Shapiro, and R.E. Cupp, 1988: Doppler lidar observations of a downslope windstorm.  Mon. Wea. Rev., 116, 2265-2275.

 

Smith, R.B., 1985: On severe downslope winds.  J. Atmos. Sci., 42, 2597-2603.