Abstract

We compare vertical wind fluctuations observed by VHF radar wind profilers in the tropical troposphere over a large, mountainous island (Pohnpei, at 7°N, 158°E) and a large, low-profile atoll (Christmas Island, at 2°N, 157°W). The major difference in the two datasets appears to be due to the frequent occurrence of mountain waves over Pohnpei and to their absence over Christmas Island. These waves are generated continuously at low levels over Pohnpei by the steady ENE trade winds acting on the 800 m high terrain, and can extend at least into the lower stratosphere. We find the occurrence of mountain waves at middle- and upper-tropospheric levels over Pohnpei to be governed primarily by a “critical-layer” relationship between the winds aloft and the near-surface winds: a reversal in the direction of the upper level winds relative to the lower-level trade winds precludes the upward propagation of mountain waves, whereas a nonreversed wind profile allows the waves to propagate freely into the lower stratosphere, and possibly to much higher heights, depending on the phase of the QBO.

Abstract

The authors report results of a study of finescale turbulence structure in the portion of the nocturnal boundary layer known as the residual layer (RL). The study covers two nights during the Cooperative Atmosphere–Surface Exchange Study 1999 (CASES-99) field experiment that exhibit significant differences in turbulence, as indicated by the observed turbulence dissipation rates in the RL. The RL turbulence sometimes reaches intensities comparable to those in the underlying stable boundary layer.

The commonly accepted concept of turbulence generation below critical values of the gradient Richardson number (Ri _g ) is scale dependent: Ri values typically decrease with decreasing vertical scale size, so that critical Ri _g values (≈0.25) occur at vertical scales of only a few tens of meters. The very small scale for the occurrence of subcritical Ri poses problems for incorporating experimentally determined Ri _g -based methods in model closures in models with poor resolution.

There appear to be two distinct turbulence “regimes” in the RL: a very weak but ever-present background turbulence level with minimal temporal and spatial structure and a more intense intermittent regime during which turbulent intensity can approach near-surface nighttime turbulent intensities. It is hypothesized that the locally produced RL turbulence can be related to upward-propagating atmospheric gravity waves generated by flow over the low-relief terrain. The presence of critical layers in the RL, caused by wind turning with height, results in the generation of intermittent turbulence.

Abstract

During the Cooperative Atmosphere–Surface Exchange Study-1999 (CASES-99) campaign in southeastern Kansas in the fall of 1999, the University of Colorado's Cooperative Institute for Research in Environmental Sciences (CIRES) made a series of vertical profiling measurements using the CIRES tethered lifting system (TLS). The results reported here began during a period when the nocturnal boundary layer (NBL) was characterized by a low-level jet (LLJ) peaking at 120 m and a temperature profile that increased smoothly with height to a point slightly above the height of the LLJ peak. Then, within a period of less than 30 min, the character of the NBL changed abruptly, with the breakdown of the well-defined LLJ and the appearance of a surprisingly steep temperature change of some 3.5 K around 180–190 m AGL. Part of this inversion was extremely sharp, with the steepest portion showing a temperature change of 1 K over an altitude range of only 5 cm, corresponding to a vertical temperature gradient in excess of 20 K m⁻¹. The general shape of this steep gradient was maintained—albeit with slightly reduced values—for at least 20 min. It is understood that the magnitude of the steepest portion of this gradient exceeds all previously observed atmospheric gradients by over an order of magnitude, although comparable gradients—albeit under very disparate conditions—have been observed in the ocean.

A second surprising feature apparent in these results was the steepness of the gradients in turbulence structure at the top of the NBL and within the residual layer (RL), the region above the NBL that is usually slightly stably stratified and extends upward to the height marked by the vestiges of the previous day's capping inversion. At the NBL top, energy dissipation rates and temperature structure parameters dropped sharply by more than one order of magnitude over a distance of only a few meters. At higher altitudes within the RL, a 60-m-thick region of very weak turbulence was observed. This low-turbulence region also exhibited sharp edges, where energy dissipation rates and temperature structure parameters changed by at least an order of magnitude over vertical distances of only a few meters.

Abstract

A high–Reynolds number direct numerical simulation (DNS) is employed to explore the instability and turbulence dynamics accompanying an idealized multiscale flow that approximates such environments observed throughout the atmosphere. The DNS describes the superposition of a stable gravity wave (GW) and a stable, oscillatory, finescale shear flow that together yield significant wave–wave interactions, GW breaking, Kelvin–Helmholtz instabilities (KHI), fluid intrusions, and turbulence. Larger-scale GW breaking and KHI events account for the strongest turbulence intensities, with intrusions competing with KHI and GW breaking at smaller spatial scales and later times. These dynamics drive a series of sheet-and-layer structures in the velocity, stability, and dissipation fields that persist for many buoyancy periods. Measures of local turbulence intensities include energy dissipation rates, Ozmidov and Thorpe scales (L _O and L _T , respectively), and a buoyancy Reynolds number sufficient to ensure sustained, strong turbulence events. These exhibit significant variability between and within instability events of different types. The Ozmidov and Thorpe scales for individual events are employed to assess variations of their ratio, C = L _O /L _T , with time. The value of C is highly variable with event type and time but typically increases with time because significant fluid overturning most often precedes turbulence. The value of C determined for the entire domain varies from 0 prior to instability to approximately 2 or larger at late times, with minima (maxima) prior to (following) significant instability and turbulence events. This appears to preclude an assumption that C is constant in stratified flows, except perhaps as an event average that may depend on event type.

Abstract

Data obtained with multiple instruments at the main site of the 1999 Cooperative Atmosphere–Surface Exchange Study (CASES-99) are employed to examine the character and variability of wave motions occurring in the stable nocturnal boundary layer during the night of 14 October 1999. The predominant motions are surprisingly similar in character throughout the night, exhibiting largely westward propagation, horizontal wavelengths of ∼1 to 10 km, phase speeds slightly greater than the mean wind in the direction of propagation, and highly coherent vertical motions with no apparent phase progression with altitude. Additionally, vertical and horizontal velocities are in approximate quadrature and the largest amplitudes occur at elevated altitudes of maximum stratification. These motions are interpreted as ducted gravity waves that propagate along maxima of stratification and mean wind and that are evanescent above, and occasionally below, the altitudes at which they are ducted. Modal structures for ducted waves are computed for mean wind and stratification profiles for three specific cases and are seen to provide a plausible explanation of the observed motions.

Abstract

The light-wind, clear-sky, very stable boundary layer (vSBL) is characterized by large values of bulk Richardson number. The light winds produce weak shear, turbulence, and mixing, and resulting strong temperature gradients near the surface. Here five nights with weak-wind, very stable boundary layers during the Cooperative Atmosphere–Surface Exchange Study (CASES-99) are investigated. Although the winds were light and variable near the surface, Doppler lidar profiles of wind speed often indicated persistent profile shapes and magnitudes for periods of an hour or more, sometimes exhibiting jetlike maxima. The near-surface structure of the boundary layer (BL) on the five nights all showed characteristics typical of the vSBL. These characteristics included a shallow traditional BL only 10–30 m deep with weak intermittent turbulence within the strong surface-based radiation inversion. Above this shallow BL sat a layer of very weak turbulence and negligible turbulent mixing. The focus of this paper is on the effects of this quiescent layer just above the shallow BL, and the impacts of this quiescent layer on turbulent transport and numerical modeling. High-frequency time series of temperature T on a 60-m tower showed that 1) the amplitudes of the T fluctuations were dramatically suppressed at levels above 30 m in contrast to the relatively larger intermittent T fluctuations in the shallow BL below, and 2) the temperature at 40- to 60-m height was nearly constant for several hours, indicating that the very cold air near the surface was not being mixed upward to those levels. The presence of this quiescent layer indicates that the atmosphere above the shallow BL was isolated and detached both from the surface and from the shallow BL.

Although some of the nights studied had modestly stronger winds and traveling disturbances (density currents, gravity waves, shear instabilities), these disturbances seemed to pass through the region without having much effect on either the SBL structure or on the atmosphere–surface decoupling. The decoupling suggests that under very stable conditions, the surface-layer lower boundary condition for numerical weather prediction models should act to decouple and isolate the surface from the atmosphere, for example, as a free-slip, thermally insulated layer.

A multiday time series of ozone from an air quality campaign in Tennessee, which exhibited nocturnal behavior typical of polluted air, showed the disappearance of ozone on weak low-level jets (LLJ) nights. This behavior is consistent with the two-stratum structure of the vSBL, and with the nearly complete isolation of the surface and the shallow BL from the rest of the atmosphere above, in contrast to cases with stronger LLJs, where such coupling was stronger.