High-Resolution Simulation of Hurricane Bonnie (1998). Part II: Water Budget

1. High-Resolution Simulation of Hurricane Bonnie (1998). Part II: Water Budget SCOTT A. BRAUN J. Atmos. Sci., 63, 43-64

2. Introduction • The total heat content of normal tropical air, if raised by undilute ascent within cumulus towers, is insufficient to generate a warm core capable of reducing the surface pressure below 1000 mb (Riehl 1954; Palmen and Riehl 1957; Malkus and Riehl 1960; Kurihara 1975). • Horizontal advection tended to transport drier air into the core in the boundary layer and moist air from the eye to the eyewall within the low-level outflow above the boundary layer (Zhang et al. 2002). • Few studies of the condensed water budget have been conducted for hurricanes (Marks 1985; Marks and Houze 1987; Gamache et al. 1993). • In this study, we compute budgets of both water vapor and total condensed water from a high-resolution simulation of Hurricane Bonnie (1998).

3. 8/23 8/24 • OBS • MM5 Simulation and analysis descriptiona. Simulation description Coarse-resolution: Started at 1200 UTC 22/08/1998 (36 hrs) 36 km: 91× 9712 km: 160×160 High-resolution:Started at 1800 UTC 22/08/1998 (30 hrs)6 km: 225×225 2 km: 226×226Vertical: 27 levels

4. Simulated dBZ at 2 km MSL valid 1200 UTC 23/08. TRMM dBZ at 2 km MSL at 1800 UTC 22/08 TRMM dBZ at 1800 UTC 22/08 MM5dBZ at 1200 UTC 23/08 >10 dBZ contoured frequency by altitude diagrams (CFADs; Yuter and Houze 1995) of reflectivity

5. 40 m 2.7km 12km 6.8km

6. dBZ + w (qcl+qci) + w dBZ + Vr

7. tangential velocity radial velocity 56ms-1 qv vertical velocity qcl + qci qra, qsn, qgr

8. Budget formulation qvis mixing ratio of water vapor;qcis the mixing ratio of cloud liquid water and ice;qpis the mixing ratio of rain, snow and graupel; V’is the storm-relative horizontal air motion;w is the vertical air motion;VTis the hydrometeor motion;+ is source; - is sink;Cis the condensation and deposition;Eis the evaporation and sublimation;Bis the contribution from theplanetary boundary layer;Dis the turbulent diffusion term;Zis the artificial source term associated with setting negative mixing ratios to zero. the azimuthally averaged horizontal advective flux is simply that associated with radial transport U and V are the Cartesian grid storm-relative horizontal velocities in the xand ydirections; uand v are the storm-relative radial and tangential winds,

9. the temporal and azimuthal mean: h-1·(kg/m3)·[(kg/kg) · h-1]·h =kg·m-3·h-1 the time-averaged and vertically integrated amount: h-1·(kg/m3)·[(kg/kg) · h-1]·m·h =kg·m-2·h-1 the time-averaged, volumetrically integratedamount: (kg·m-3 ·h-1)·m3 =kg·h-1 Zxis artificial source terms associated with setting negative mixing ratios (caused by errors associated with the finite differencing of the advective terms) to zero, that is, mass is added to eliminate negative mixing ratios.

10. Budget resultsa. Water vapor budget condensation horizontal flux divergence, (a), (e), (f) interval: 2 g m-3 h-1(b) and (d) interval: 20 g m-3 h-1(c), (g), (h) interval: 0.5 g m-3 h-1thin solid lines show the zero contour evaporation vertical flux divergence, (b) + (d) (a) + (c) boundary layer source term divergence term

11. eyewall region (30-70 km) outer region (70-200 km) updraft condensation occurring in updraft The smaller contribution of stronger updrafts is indicative of the larger role of stratiform precipitation processes outside of the eyewall. much of the eyewall condensation is associated with hot towers.

12. condensation (total source of cloud) b. Condensed water budget cloud sink horizontal flux divergence net source vertical flux divergence added water mass to offset negative mixing ratios boundary layer source (a) interval: 2 g m-3 h-1(b) to (e) interval: 0.5 g m-3 h-1(f) interval: 0.125 g m-3 h-1thin solid lines show the zero contour cloud budget

13. cloud sink source for rain sink for rain source for graupel sink for graupel cloud budget (a) to (f) interval: 2 g m-3 h-1thin solid lines show the zero contour source for snow sink for snow net microphysical source horizontal flux divergence precipitation budget precipitation fallout andvertical flux divergence added water mass to offset negative mixing ratios (a) to (c) interval: 2 g m-3 h-1(d) interval: 0.5 g m-3 h-1thin solid lines show the zero contour

14. (a) and (c) interval: 20 kg m-2 h-1(b) interval: 5 kg m-2 h-1 6.8km qv evaporation condensation total rain source warm rain source precipitation fallout cold rain source graupel source

15. c. Volume-integrated budgets Zero/C ~ 12 % Zero/C ~ 13 % P/C ~ 65 %

16. d. The artificial water source cloud liquid water rain snow cloud ice hydrometeors: (a) shaded interval: 0.1 g m-3(b) to (e) shaded interval: 0.5 g m-3 source terms: (a) to (e) line interval: 0.5 g m-3 h-1 graupel

17. cloud water rain graupel

18. Conclusion • A detailed water budget is performed using a high-resolution simulation of Hurricane Bonnie (1998). The simulation generally reproduces the track, intensity, and structure of the storm, but overpredicts the precipitation as inferred from comparison of model and TRMM radar reflectivities. • The water vapor budget confirms that the ocean source of vapor in the eyewall region is very small relative to the condensation and inward transport of vapor, with the ocean vapor source in the eyewall (0.7) being approximately 4% of the inward vapor transport into the eyewall (16.8) region. • Inthe eyewall, most of the condensation occurs within convective towers while in the outer regions condensation results from a mix of convective and stratiform precipitation processes, with the stratiform component tending to dominate. • Precipitation processes acting outside of the eyewall region are not very dependent on the condensate mass produced within and transported outward from the eyewall. Instead, the precipitation derives from convection in outer rainbands and the subsequent transition to stratiform precipitation processes.

19. Conclusion • Although the artificial water mass source is very small at any given grid point, its cumulative impact over large areas and over time is more substantial, contributing an amount of water that is equivalent to 15%–20% of the total surface precipitation. • This problem likely occurs in any MM5 simulation of convective systems, but is probably much less a concern for purely stratiform precipitation systems.