The upslope-downslope flow transition on a basin sidewall

1. (7) CONCLUSIONS The upslope-downslope flow transition on a basin sidewall D. Martínez(1,2), C. D. Whiteman(1), S. W. Hoch(1), M. Lehner(1) and J. Cuxart(2) (1) University of Utah, Salt Lake City, UT (2) University of the Balearic Islands, Palma (Illes Balears), Spain. 14thConference on Mountain Meteorology 30 August – 3 September 2010, Lake Tahoe Vicinity, CA *e-mail adress: dani.marinez@uib.cat (1) INTRODUCTION (4) THE UPSLOPE-DOWNSLOPE TRANSITION IN DETAIL Three smoke grenades (Fig. 4) were ignited at a site 50 m to the south of wu (Fig. 1) during the upslope-downslope transition period. A regular upslope-downslope transition occurred on the west inner sidewall of Arizona's Meteor Crater on the afternoons of the month-long METCRAX field campaign under quiescent synoptic conditions. This transition is seen in time-lapse photos taken during smoke releases on 20 October 2006. The present study analyzes METCRAX data to explain the phenomena recorded in the photographic images. 1510 MST 1530 MST 1550 MST (2) INSTRUMENTATION The primary data used for the analysis comes from three flux towers instrumented with 3-D sonic anemometers and temperature sensors at four levels from 0.5 to 8.5 m. These towers, located on the west inner sidewall, are West Upper (wu), West Lower (wl) and Floor (flr) (Fig. 1). Surface radiative and energy budget components were also measured near each tower (Whiteman et al. 2008). Fig. 4b: The slope flow reverses intermittently from downslope to upslope and vice versa. Smoke is confined to the ground in downslope flow, but is dispersed vertically due to enhanced turbulence when the flow turns upslope. Fig. 4c: The smoke plume travels downslope in a shallow layer whose depth increases along the slope. The flow is disturbed sporadically by turbulent episodes that disperse the smoke plume vertically, reaching a height of tens of meters. Fig. 4a: Upslope easterly winds are still present over the shadowed slope, with big eddies that are able to transport the smoke up to the rim. Fig. 1: Meteor Crater, 10 m contour intervals (top). Topographic cross-section, indicating the 3 flux towers (bottom). (5) WIND DIRECTION AND POTENTIAL TEMPERATURE EVOLUTION At wu temperature stratification becomes stable due to surface cooling, while the stability arrives at wl with the downslope current. At wu and wl wind directions turn downslope at about the same time, 1540 MST. At flr the wind direction turns downslope 10 minutes later, when the local stratification is still unstable. (3) OVERVIEW OF THE EVENING TRANSITION ON THE WESTERN SIDEWALL The evening transition within the crater basin is strongly affected by the propagation of shadows cast by the crater rim. Local sunset arrived 160, 130 and 75 minutes before astronomical sunset (1734 MST) at the wu, wl and flr sites, respectively. Fig. 5: Time evolution of wind direction at 5-m height (red line) and potential temperature for all levels at the (a) wu, (b) wl and (c) flr towers. Wind and temperature data have a 30-s and 1-min resolution, respectively. Vertical lines as in Fig 2. (6) ANALYSIS OF THE SCALES OF MOTION INVOLVED IN THE TRANSITION A wavelet analysis of kinetic energy using sonic anemometer data identifies the relevant scales of motion (or, equivalently, time scales). At all sites a reduction in kinetic energy and a decrease in the scale of motion occurs with the onset of the drainage flow. During the downslope period the relevant scales are the turbulent scale (T < 5 minutes) and the 10- and 15-minute periods, related to crater-scale dynamics. At flr there is no gap between the turbulent and crater scales, probably because of phenomena on a range of scales generated at other parts of the crater. Kinetic energy reaches a maximum during the 5 minutes before the downslope current onset, especially at wl. This energy enhancement occurs at periods up to 25 minutes, indicating that the event has a crater scale. Fig. 2: Meteograms from wu (left) and wl (right) for the evening transition period. (a)-(b) Radiation budget components and (c)-(d) surface energy budget components. Vertical lines indicate local sunset at wu (red), wl (black), flr (blue) and astronomical sunset (pink). 5-min resolution. The sharp decrease in net radiation at local sunset affects the times of reversal of sensible (H) and ground (Gsfc) heat fluxes (Fig. 2). At wu and wl, the Rnet drop (Fig. 2) leads to a decrease in surface and air temperatures (Fig. 3). 15 minutes after local sunset a temperature inversion has formed. At this time, the sensible heat flux at 3 m also reverses sign. Wind directions are variable from 1515 to 1545 MST and shift to a steady downslope direction by 1545 MST at all sites. At flr this shift occurs before local sunset! Upslope wind speeds are higher than downslope speeds at all sites. Downslope wind speeds are highest at wl (1545-1700 MST). Fig. 6: Time evolution of kinetic energy per unit period at 5-m height computed from the wavelet transform of the wind components at the (a) wu, (b) wl and (c) flr sites. Vertical lines as in Fig.2. Fig. 3: (a) Wind direction, (b) wind speed, and (c) potential temperature at 5 m and (d) 0.5-5 m temperature difference, for wu (red), wl (black) and flr (blue). Vertical lines as in Fig. 2. • The passage of the shadow between the two west sidewall towers takes 30 minutes, but downslope flow is initiated at about the same time at both towers. • The downslope flow starts when the two western towers are in shadow and immediately after a high energetic event at the crater scale. • At flr wind direction turns downslope ten minutes after the onset of downslope flow on the western sidewall, when the stratification at flr is still unstable. • The wind reversal at wl might be caused by a gravity current that formed upslope of wl, but existing evidence is not compelling. • Wavelet analysis shows that motions of turbulent and crater scale are both important during the downslope period. References Whiteman et al. (2008). METCRAX 2006 – Meteorological experiments in Arizona's Meteor Crater. Bull. Amer. Meteor. Soc., 89, 1665-1680. This work was partially funded by the grant BES-2007-16272 of the research project CGL2006-12474-C03-01 of the Spanish Government, including the European Regional Funds FEDER.