Implementation of Terrain Resolving Capability for The Variational Doppler Radar Analysis System (VDRAS) Tai, Sheng-Lun 1, Yu-Chieng Liou 1,3, Juanzhen Sun 2, Shao-Fan Chang 1 1 Department of Sciences, National Central University, Taiwan 1 Department of Atmospheric Sciences, National Central University, Taiwan 2 National Center for Atmospheric Research, USA 3 Taiwan Typhoon and Flood Research Institute, Taiwan Introduction A 4DVAR Doppler radar retrieval system for high-resolution and rapid updated analysis. Installed in more than 20 sites since (Sun and Crook, 1997) Cloud model inside the system has no ability to resolve terrain. Improve the accuracy of analysis and Quantitative Precipitation Nowcasting (QPN) by implementing terrain resolving capability. (IBM_VDRAS) Ghost-cell immersed boundary method (GCIBM) 2-D Linear Mountain Wave Simulation Fig. 1. Schematic of domain with an immersed boundary. Triangle points (▲) represent grid points in the flow region, Blue circle points (GC) locate the ghost-cells and red circle points are the image points (IP) of ghost cells. The line connected with IP and GC is perpendicular to the immersed boundary. Fig. 5. Assimilation strategy (left) and model domain with radar locations (right) for the OSSE. Two sets of radar data are assimilated in the assimilation cycle. References Sun, J. and N. A. Crook, 1997: Dynamic and microphysical retrieval from Doppler radar observations using a cloud model and its adjoint. Part I: Model development and simulated data experiments. J. Atmos. Sci., 54, 1642–1661. Tseng, Y., and J. Ferziger, 2003: A ghost-cell immersed boundary method for flow in complex geometry. J. Comput. Phys., 192, 593– 623. Chang, S.-F., Y.-C. Liou, J. Sun, and S.-L. Tai, 2016: The Implementation of the Ice-Phase Microphysical Process into a Four- Dimensional Variational Doppler Radar Analysis System (VDRAS) and Its Impact on Parameter Retrieval and Quantitative Precipitation Nowcasting. J. Atmos. Sci., 73, 1015–1038. This research is sponsored by Ministry of Science and Technology of Taiwan, under Grant MOST M Tseng and Ferziger (2003) Ghost-cell: first grid point under terrain surface (immersed boundary). Imposing terrain boundary condition implicitly by updating ghost-cell value. Update every time step in model integration. Terrain effect gradually obtained. Variables updated: U, V, W, θ, Qt and Qr. IP GC ▲▲ ▲ ▲ OSSE: data assimilation experiment Virtual radial wind and rainwater mixing ratio from WRF are assimilated. Verification for adjoint model and assimilation scheme Assimilation strategy and experiment design: Assimilation strategy and experiment design: Fig. 3. Both WRF and IBM_VDRAS simulations displayed along a vertical cross sections at Y=71 km at (a) 1 st hour, (b) 2 nd hour and (c) 3 rd hour. Grey shading represents rainwater mixing ratio in g kg -1 and dashed line marks the cold pool area where the temperature perturbation is less than -1.2 o C. Fig. 4. The Hovmoller diagram of forecast rainwater mixing ratio and vertical velocity along Y=71 at Z=10 level by (a) WRF model and (b) IBM_VDRAS model. The Y- axis ranges for 3 hours. WRF WRF (Truth)IBM_VDRAS Qr Fig. 6. Vertical cross sections (along Y=71) for WRF forecasts (a~e) and IBM_VDRAS retrievals (f~j). (a), (f): rainwater mixing ratio (gkg -1 ); (b), (g): U- wind (ms -1 ); (c), (h): vertical velocity (ms - 1 ); (d), (i):pressure perturbation (mb); (e), (j): temperature perturbation ( o C). Radar data successfully assimilated. All type of variables retrieved very well. For comparisons with the available analytic solution (Smith,1980). Isothermal atmosphere (dT/dZ = 0), T = 250 K. Bell-shaped mountain. Wind speed (U) = 20 ms -1, terrain half-width (a) = 10 km, max. height (h) = 1 m. Model resolutions: △ x = 2 km, △ z = 200 m. Fig. 2. The vertical velocity field of 2-D linear mountain wave. (a): Result by IBM_VDRAS simulation at seconds (Ut/a= 60). (b): Analytical solution ; both the magnitudes have been amplified by Parallel Simulations of IBM_VDRAS and WRF WRF IBM_VDRAS IBM_ VDRAS U Qr W P’ T’ WRF (Truth)IBM_VDRAS Preliminary Study of High-resolution Simulation for Snow Prediction Fig. 6. (Left) Diagram shows model domain and topography in shading, with location of GWRWO sounding station (▲) and DGRWO surface station ( ★ ). (Right) Skew T diagram of sounding released at GWRWO on 00UTC (blue) and 12UTC (red) of Jan. 29, DX=DY=300 m, DZ=200 m. Ice-phase scheme implementation (Chang, 2015). Fig. 7. Simulation results after 1200 seconds integration: (a), (b) show snow mixing ratio (gkg -1 ) and wind speed (ms -1 ) distribution at 900 meter height. (c), (d), (e) and (f) present vertical cross sections of ice mixing ratio (gkg -1 ), vertical velocity (ms -1 ), snow mixing ratio (gkg -1 ) and temperature perturbation ( o C) along the solid line shown in left panel of Fig st hr2 nd hr3 rd hr