Airflows and turbulent flux measurements in mountainous terrain: Part 2: Mesoscale effects

The location of the Niwot Ridge Ameriflux site within the rocky mountains subjects it to airflows which are common in mountainous terrain. In this study, we examine the effects of some of these mesoscale features on local turbulent flux measurements; most notably, the formation of valley/mountain flows and mountain lee-side waves. The valley/mountain flows created local non-stationarities in the wind flow caused by the passage of a lee-side convergence zone (LCZ) in which upslope and downslope flows met in the vicinity of the measurement tower. During June-August, 2001, possible lee-side convergences were flagged for ∼26% of all half-hour daytime flux measurement periods. However, there was no apparent loss of flux during these periods. On some relatively stable, summer nights, turbulence (designated via σ_w), and scalar fluctuations (temperature and CO₂, for example) exhibited periodicities that appeared congruent with passage of low frequency gravity waves (τ ∼ 20 min). Spectral peaks at 0.0008 Hz (20 min) in both vertical velocity and scalar spectra were observed and indicated that 25-50% of the total scalar covariances were accounted for by the low frequency waves. During some periods of strong westerly winds (predominantly in winter), large mountain gravity waves were observed to form. Typically, the flux tower resided within a region of downslope "shooting flow", which created high turbulence, but had no detrimental effect on local flux measurements based on valid turbulence statistics and nearly complete energy budget closure. Periodically, we found evidence for re-circulating, rotor winds in the simultaneous time series of wind data from the Ameriflux tower site and a second meteorological site situated 8 km upslope and to the West. Only 14% of the half-hour time periods that we examined for a 4 month period in the winter of 2000-2001 indicated the possible existence of rotor winds. On average, energy budget closure was ∼20% less during periods with rotor occurrence compared to those without. Results from this study demonstrate that the potential exists for relatively rare, yet significant influences of mesoscale wind flow patterns on the local half-hour flux measurements at this site. Occurrence of these events could be detected through examination of normal turbulence statistical parameters.