MMM SIGNIFICANT ACCOMPLISHMENTS, FY 95

• The phenomenon of cyclogenesis in the lee of mountains such as the Rocky Mountains may often occur simultaneously with the decay of a synoptic-scale baroclinic wave has been demonstrated by Christopher Davis (joint appointment with RAP). The idea that an incident wave decays as it interacts with mountains has been shown in previous studies using simple, linear models. C. Davis's study concerns observed cases, and strongly links these to behavior found in an idealized, nonlinear, quasigeostrophic model. The model incorporates the interaction of a jet-like mean flow with a localized, large-scale mountain. In both the idealized model and observations, as a baroclinic wave approaches a mountain, the surface temperature perturbations induced as air flows over the terrain act to shift the lower part of the wave downstream. Because the wave is initially tilted upshear in a configuration favorable for growth, the increase in vertical tilt either reduces the rate of growth or causes the wave to decay. However, with the changing vertical structure, it is possible to produce a local spin-up of circulation even though the wave as a whole is decaying.

• Because fronts and frontogenesis are generally poorly resolved by both observational networks and numerical simulations, questions remain concerning the behavior and forecasting of very narrow and intense frontal zones. Christopher Snyder and Daniel Keyser (State University of New York, Albany) have investigated these questions at a fundamental level, by considering the "spin down" of a two-dimensional frontal zone, initially in thermal-wind balance, that is subjected to parameterized surface drag and turbulent mixing. They find that these physical processes suffice to drive intense frontogenesis, which is limited (even with turbulent mixing) only by the model resolution of a few hundreds of meters, and which results in frontal nose indistinguishable from a density current.

• Two-dimensional numerical simulations of tropical cloud systems under the influence of time-varying, large-scale forcing determined from the Global Atmospheric Research Program (GARP) Atlantic Tropical Experiment (GATE) and Tropical Ocean Global Atmosphere Coupled Ocean-Atmosphere Response Experiment (TOGA COARE) data sets have been performed by Wojciech Grabowski, Mitchell Moncrieff, and Xiaoqing Wu (visitor, University of California, Los Angeles). The model contains multi-phase microphysics, cloud-radiation and surface interactions, and has been integrated for up to 40 days -- comparable to intraseasonal time scales. The objective of formulating paradigms of the large-scale role of physically complex precipitating clouds is one step closer because the life cycle, structure, and bulk properties of tropical cloud systems can be reproduced provided the large-scale wind, temperature, and humidity profiles are specified in a simple (horizontally homogeneous) way. This demonstrates a link between large-scale forcing and convective response that is fundamental to cloud-related parameterizations for climate General Circulation Models (GCMs) and numerical weather prediction models. Moreover, preliminary integrations suggest these results hold in three spatial dimensions. A next step is to help improve the physical basis of cloud parameterizations in GCMs by fully verifying these explicit (cloud resolving) model results against observations and comparing them against "single-column" models (small regions of the globe represented by a single grid point in a GCM). This unique study is a mix of observations, cloud resolving modeling, and climate objectives in a reasonably realistic meteorological setting.

• A numerical study, which has been completed by Stanley Trier, William Skamarock, and Margaret LeMone, elucidates the behavior of several observed features examined in the 22 February 1993 TOGA COARE squall line, namely the mesoscale vortex at the north end of the line, the existence of a mid-tropospheric minimum vertical velocity midway between the leading-edge updraft and an upper-tropospheric updraft about 20 km to the rear, and a precipitation band that forms normal to and rearward of the leading edge of the squall line. The vortex, which in observations extends down to below 200 m, appears to receive its vorticity largely from the shear between the front-to-rear and rear-to-front current. Air parcels moving at the top of the rear-to-front current and then turn northward upon encountering the squall line acquire this shear, which is tilted vertically as the air moves upward in the rising current to the rear of the squall line. The vertical motion minimum between the two updraft peaks is associated with strong downward pressure forces between upward cold-pool forcing at low levels and buoyancy forcing aloft. The transverse band is an extreme case: it lies beneath air that rises only gradually after leaving the leading edge, held down by downward-directed pressure forces extending several tens of kilometers behind the leading edge. A companion paper on the observations, with David Jorgensen (NOAA, NSSL-Mesoscale Research and Applications Division) as lead author, is near completion.

• Is it a breeze or blocking? Hawaiian simulations and observations tell the story about circulations that force windward rainbands. Richard Carbone, William Cooper (joint appointment with ATD), and Wen-Chau Lee (ATD) have compared analyses of Hawaiian Rainband Project (HaRP) data to numerical simulations by Piotr Smolarkiewicz and collaborators to show that evaporation from diurnally-forced orographic rainfall is the principal initiation mechanism for flow reversal on the windward side of Hawaii. Immersed in a background condition of flow stagnation due to blocking, the evaporatively initiated reverse flow is strengthened at night by radiative cooling leading to the formation of a density current that slowly propagates toward the ocean. Rainbands often initiate or amplify along the convergence zone that separates this current from the incoming tradewinds. Farther off shore, the flow separation line possesses less thermal contrast and gains more characteristics of a pure, dynamically-induced flow as described by Smolarkiewicz and collaborators, thus leading to the conclusion that both breeze and blocking mechanisms cooperate to force Hawaiian rainbands.

• The existence of shallow supercell-type convective structures in landfalling hurricane environments that helps to explain the frequent occurrence of tornado outbreaks as hurricanes come ashore have been documented by Morris Weisman and Eugene McCaul (NASA, Huntsville). These storms develop strong, rotating updrafts at low levels despite the lack of convective available potential energy (CAPE), due to the dynamic interaction of the updraft with the very strong environmental low-level vertical wind shear. Such shallow supercells have also been documented for a variety of other nonclassic environments that display CAPE and vertical wind shear characteristics that are similar to the hurricane cases. Severe weather forecasting techniques are currently being modified to account for these kinds of storms.

• New aspects of critically nonhydrostatic critical-level dynamics have been uncovered by Terry Clark, Teddie Keller (joint appointment with RAP), Janice Coen, and Hsiao-Ming Hsu (visitor, Woods Hole, MA, joint appointment with RAP) in their studies of flow over complex terrain for the Hong Kong project. The width of Lantau Island was found to not be in the parameter space where strong resonant amplification of waves can occur, as is often the case in the Front Range of the Rocky Mountains, but rather where highly transient critical-level flows result in moderate turbulence. A second regime of flow identified was deep uniform flow due to tropical storms resulting in severe levels of mechanically produced turbulence. Simulations of these flows were validated against lidar, Automated Weather Systems (AWS), and aircraft observations from the Lantau Experiment in FY 95.

• The experimental identification of a new boundary layer in hard convective turbulence by Belmonte, Tilgner, and Libchaber (1994) using spectral cutoffs has been confirmed numerically by Robert Kerr, Axel Brandenberg (Nordita University, Denmark), and Jackson Herring. The work demonstrates how numerical simulations can now be used to test and understand laboratory and environmental turbulent flows.

• One of the first numerical investigations of the oceanic planetary boundary layer examining the interaction of turbulence and surface gravity waves has been carried out by James McWilliams (University of California, Los Angeles), Peter Sullivan, and Chin-Hoh Moeng. Their large-eddy simulation model of the oceanic planetary boundary layer includes additional terms proportional to the Lagrangian Stokes drift which arise from averaging over the high-frequency surface gravity waves. They found that the equilibrium solutions with steady, aligned wind and waves contain a dominant coherent structure, a turbulent Langmuir Cell. The long-lived turbulent Langmuir Cell has its strongest vorticity aligned with the wind and waves and is trapped near the surface on the scale of the prescribed Stokes drift profile. The presence of the turbulent Langmuir Cell significantly alters the mean profiles of the oceanic currents, enhances the turbulent fluxes of momentum, heat, and tracers, and increases the turbulent kinetic energy and dissipation near the water surface.

• It has been reported by Ilga Paluch, Charles Knight, and L. Jay Miller that there is an extremely tight correlation between liquid water content and radar reflectivity factor, at constant altitude in early cumulus clouds before coalescence starts. This correlation established from Forward Scattering Spectrometer Probe (FSSP) data in developing cumulus during the Convection and Precipitation/Electrification Experiment (CaPE) should prove useful in mapping liquid water content in nonprecipitating cumulus, using radar. The nature of this correlation sheds light upon the mixing of cloud with clear air. More complete data on this relationship were gathered in the summer of 1995 during the Small Cumulus Microphysics Study (SCMS), led by Knight.

• Using data from cold orographic wave clouds and cirrus from the First ISLSCP Field Experiment (FIRE) II, Andrew Heymsfield and Larry Miloshevich have developed a three-layer conceptual model of the structure of cirrus clouds. The production of ice crystals, in a thin zone of ice nucleation near cloud top, requires that the relative humidity with respect to water must be at least 100% at temperatures -40C and warmer and decreases to only 70% at -60C. The middle zone is a deep layer of ice supersaturation, where ice crystals grow and fall. The lowermost zone is a region of ice crystal sublimation, where ice crystals from zone 1 fallout and form the cirrus cloud base.

• Short-term quantitative precipitation forecasting has been a challenging problem for numerical weather prediction. The main difficulty lies in the lack of sufficient mesoscale details in the model initial conditions. The problem becomes even more difficult when there is an active convective system at the model initial time. Xiaolei Zou and Ying-Hwa Kuo have recently completed a set of experiments by directly assimilating the observed 3-h rainfall data into the NCAR/Pennsylvania State University (PSU) Mesoscale Model (Version 5) (MM5) using a four-dimensional variational data-assimilation system with moist physics for the Wichita Falls tornado case of the Severe Environmental Storms And Mesoscale Experiment (SESAME) 1979. The assimilation of rainfall data significantly improved the model rainfall in terms of position and intensity, and recovered realistic mesoscale wind, temperature, and moisture structure. The improved initial condition due to rainfall assimilation considerably improved the prediction of an observed mesoscale convective system. This work shows that significant improvement in quantitative precipitation forecasts is possible through continued development of a variational data-assimilation system with the use of mesoscale data sets (such as the rainfall observations).

Return to ASR Home Page