significant.html

Figure 10: Sea level pressure (contour interval 1 mb) and 1-hour rainfall accumulation on domain 4 (3 km resolution), 54 hours into a simulation of Tropical Storm Diana initialized at 1200 UTC 7 September, 1984. Cumulus parameterization has been turned off on this domain. Click on image to view larger figure.

SIGNIFICANT ACCOMPLISHMENTS
*Surface Energy Bias in Climate Models (W. Pacific Warm Pool)*
The cause of bias in the surface energy budget, and the attendant climate drift problem in coupled atmosphere-ocean general circulation models (GCMs), is a key uncertainty in climate modeling. Xiaoqing Wu (joint appointment with CGD) and Mitchell Moncrieff investigated this issue using a CRM (cloud-resolving model wherein convection and clouds are explicit), a single-column model (SCM) of the NCAR Community Climate Model (CCM, convection and clouds parameterized), and observations from the Tropical Ocean and Global Atmosphere Program Coupled Ocean-Atmosphere Response Exper-ment (TOGA COARE). They showed that top-of-atmosphere radiative fluxes and sur-face energy budgets derived from the CRM simultaneously agree with observations, while the SCM (run with the same prescribed forcing) is strongly biased. An accurate horizontal and vertical distribution of condensate and its effect on radiative transfer is crucial [see Figure 1 at bottom left]. This result shows convectively driven cloud systems must be parameterized accurately before ocean-atmosphere coupling can be accurately realized, at least over the warm pool [see Figure 2 at bottom right].
Figure 1: Left panel: 30-day evolution of 6-hourly vertical profiles of (a) cloud ice water mixing ratio (g kg^-1) and (b) cloud liquid water mixing ratio (g kg^-1) from the CRM. Right panel: Same as left panel but from the SCM. Click on image to view larger figure.	Figure 2: The SST derived from the ocean model (dashed line) agreed very well with observations (solid line). On the other hand, a relatively poor prediction of SST was obtained when the ocean model was forced by surface conditions obtained from the single-column model (dotted line). Click on image to view larger figure.
*Convection Initiation over Tropical Islands*
Using a linear model and fully nonlinear numerical simulations, N. Andrew Crook (joint appointment with RAP) showed that radiative heating on the island scale is at least as important as sea breezes in initiating strong thunderstorms (known as Hectors) over the Tiwi Islands north of Darwin, Australia. In agreement with observations made during the Maritime Continent Thunderstorm Experiment (MCTEX), Figure 3 at the right shows the sea breezes from the north and south coastline converged to within 10 km of each other but did not collide, yet a Hector formed. This result raises the question of how convection over tropical islands should be initiated ('triggered') in parameterization schemes, and whether the sea-breeze collision mechanism is truly prevalent.	Figure 3: Surface rainwater and velocity vector field at 1400 Local Time from a simulation of flow over the Tiwi Islands. An intense thunderstorm, locally known as a `Hector' has developed over the center of Melville Island (island to the east). Click on image above to view a larger version.
*Turbulent Flux over a Wavy Surface*
Peter Sullivan, James McWilliams (University of California, Los Angeles), and Chin-Hoh Moeng showed that small-scale ocean waves significantly influence the mean flow, vertical momentum fluxes [see Figure 4 to the right], velocity variances, pressure, and form stress [see Figure 5 at below right]. The turbulent momentum flux can be altered by as much as 40 percent by the waves. A region of closed streamlines (cat's-eye pattern) centered about the critical layer height was found to be dynamically important at low to moderate values of wave age [see Figure 6 below]. Figure 6. Click on figure to view caption and larger image	Figure 4. Click on figure to view caption and larger image.
	Figure 5. Click on figure to view caption and larger image.
*Developing the Framework for Radar-Based Mesoscale Climatologies*
The groundwork for computational analysis of the climatological aspects of organized convection was established by L. Jay Miller,and Sherrie Fredrick, in collaboration with Richard Oye (ATD). Their extensive software development and adaptation efforts provide the capability to create spatial composites of level-II WSR-88D radar data [see Figure 7 below right]. These composites were adapted to view the coherence of convective episodes and identify the different mechanisms that govern their behavior. Analysis of convective events in a wide variety of environmental situations will lead to a dynamically based, mesoscale climatology of heavy precipitation.
Figure 7: A composite of gridded low-level (0.5 deg elevation scan) WSR-88D Level II reflectivities (top) and radial velocities (bottom) on May 27, 1998. Reflectivities (dBZ) and radial velocities (m/sec) follow the color scales on the right. Vector winds (scaled 20 m/s vector near the bottom of the color bar) were derived where radial velocity information existed from two or more radars. The radars locations are marked by stars and their call signs. Click on figure to view larger image
*Beyond Quasi-geostropy*
Richard Rotunno and Chris Snyder, in collaboration with David Muraki (Courant Insti-tute, New York University) and Gregory Hakim (University of Washington), extended quasi-geostrophic theory to include the next-order correction in Rossby number, and applied these equations to better understand the relation between meso- and synoptic-scale flows. The theory (dubbed 'QG+1') captures many key features of baroclinic wave evo-lution that are present in the primitive equations but ignored in quasi-geostrophy (Figure 8 below left). Using the QG+1 equations, the investigators developed the first mathematically complete and relatively simple explanation for the cyclonic bias in idealized baroclinic waves.
	Figure 8: Comparison of perturbation pressure (thin lines, c.i.= 0.1) and potential temperature (thick solid lines, c.i. = 0.553) from the QG+1 and PE models initialized with the base-state jet (shown at upper right) plus the most unstable normal-mode (MUNM). Also shown for reference is a QG integration initialized with a the QG MUNM. Base state jet (thin lines, c.i.= 0.1) and total potential temperature (thick lines, c.i. = .553) calculated under the condition of constant potential vorticity. Click on figure to view larger image
Weather Research and Forecast Model Development
Figure 9: Vertical velocity and cloudwater fields from 2D squall line simulations using flux-form height and mass coordinate models. There is moderate low-level vertical wind shear in the environmental winds, and the shear vector points to the right. Click on figure to view larger image	Joseph Klemp, William Skamarock, and Jimy Dudhia continued research for the development of a joint Weather Research and Forecast (WRF) model in collaboration with colleagues from NOAA/NCEP, NOAA/FSL, CAPS, and AFWA, and with university scientists. Working groups were established, detailed plans were formulated, and development began in each of the major aspects of the model system. There was significant progress in developing three prototypes for the nonhydrostatic model numerics that are being evaluated as candidates for the final numerical design. Two of these prototypes are split-explicit Eulerian models (see Figure 9 at left), while the third (designed by James Purser, visitor, NOAA/National Centers for Environmental Prediction) is a semi-implicit semi-Lagrangian formulation. Led by John Michalakes (visitor, Argonne National Research Laboratory), a demonstration prototype code for the WRF software architecture was designed and implemented, in which a single version of the code may be configured for efficient execution on platforms covering the range of current and foreseen high-performance computing hardware.
*NOx Production by Lightning in Thunderstorms*
From analysis of observations and a simulation of the 10 July 1996 Stratosphere-Troposphere Experiments: Radiation, Aerosols and Ozone (STERAO) storm in NE Colorado, James Dye (joint appointment with ATD), Skamarock, Mary Barth (joint appointment with ACD), and Eric Defer (visitor, Office National d'Etudes et de Recherches Aerospa-tiales, France), with the help of many STERAO collaborators, conclude that during a one-hour time period roughly 75 percent of the enhanced NOx (= NO + NO2) observed in the anvil by the University of North Dakota Citation aircraft was produced by lightning, and the remaining 25 percent was transport of anthropogenically produced NOx from the boundary layer. During the one-hour period just prior to and during the Citation observations the lightning flashes in the storm were almost exclusively intra-cloud. Often past model studies have assumed that intra-cloud lightning is not as important as cloud-to-ground flashes in producing NOx. This work clearly shows the importance of intra-cloud as well as cloud-to-ground lightning for the production of NOx.
*Simulation of Hurricane Diana (1984)*
Christopher Davis (joint appointment with RAP) and Lance Bosart (Affiliate Scientist, State University of New York at Albany) conducted and analyzed simulations of the genesis of hurricane Diana. The simulations are among the first known to predict the entire evolution of a tropical cyclone without the aid of vortex bogussing. Horizontal resolution as fine as three kilometers were used to resolve the eye, eye wall convection, and spiral bands around the center (see Figure 10 at right). Diagnostics reveal a two stage deepening process. In the first stage, convection triggered by an upper tropospheric disturbance produces anomalies of lower-tropospheric potential vorticity that are axisymmetrized as they spiral in toward the center. A quiescent stage ensues roughly 10 hours later during which the core approaches saturation below 500 mb but little deepening occurs. This stage is followed by rapid deepening as air-sea interaction dominates the dynamics.	Figure 10: Sea level pressure (contour interval 1 mb) and 1-hour rainfall accumulation on domain 4 (3 km resolution), 54 hours into a simulation of Tropical Storm Diana initialized at 1200 UTC 7 September, 1984. Cumulus parameterization has been turned off on this domain. Click on image to view larger figure.

SIGNIFICANT ACCOMPLISHMENTS