The Genesis of Hurricane Guillermo: TEXMEX Analyses and a Modeling Study BISTER AND EMANUEL

Incipient disturbances are prevented from developing by convective downdrafts in their cores, which bring air of low into the boundary layer, suppressing further convection. OVERCOME?? 1) an increase of the in the middle troposphere 2) an increase of relative humidity so that evaporation of rain is suppressed 3) an increase of wind speed so that the sea surface fluxes keep replenishing the boundary layer

The main goal of the Tropical Experiment in Mexico (TEXMEX) was to test a hypothesis: The elevation of in the middle troposphere just above a near-surface vorticity maximum is a necessary and perhaps sufficient condition for tropical cyclogenesis. Assumption: the elevation of is accomplished by deep convection bringing high to the middle troposphere.

The pre-Guillermo MCS was the object of IOP 5 from 2 August 1991 to 5 August 1991, during which six flights were flown. The first, third, and fifth flights, labeled 1E, 3E, and 5E, respectively, were flown with the Electra NCAR while the second, fourth, and sixth flights, labeled 2P, 4P, and 6P, respectively were flown with the WP-3D,NOAA. Data were excluded from the analysis if the magnitude of the vertical velocity exceeded 1 m/s, in order to minimize the effect of active convective updrafts and downdrafts on the analyzed fields. The differences of the temperature and the dewpoint temperature measured by the two aircraft were less than 0.3 K during both flights, and they were accounted for in the data analysis.

large-scale flow : easterly wave

Flight 2P 0200 UTC 3 August

Flight 2P 0200 UTC 3 August in the stratiform region relative humidity peak the vortex has a cold core at 3km and a warm core in the upper troposphere (thermal wind balance)

Flight 4P 0600 UTC 4 August

Strengthening rain band still northwest Flight 4P 0600 UTC 4 August Strengthening rain band still northwest A small minimum inside the cold core Vertical shear generally cyclonic, but with a anti-cyclonic in the north (a warm core developing near the region of the developing strong convection, also where maximum surface wind observed).

Flight 5E 2000 UTC 4 August warm core at 3-km altitude is dominant with a reversal of the gradient of the virtual potential temperature about 100 km from the center (also found in boundary layer).

between the cold-core stage and the development of the central warm core, the system’s main thermodynamic change is moistening of the boundary layer. (result not sensitive to square area within 15km) The increase of and virtual potential temperature in both the boundary layer and at 3-km altitude while the tangential wind speed increases suggests that intensification owing to the feedback between the wind and the surface heat fluxes is underway.

An MCS with an extensive stratiform precipitation region forms An MCS with an extensive stratiform precipitation region forms. Diabatic heating in the upper troposphere and cooling at and below the melting level lead to a formation of a midlevel vortex. Evaporation of rain increases relative humidity in the lower troposphere and leads to a downdraft that advects the vortex downward. Convection redevelops, leading to a further increase of vorticity below the maximum heating and a formation of a warm core.

neglected the effects of deep convection in obs Simulation A steady mesoscale ‘‘rain shaft,’’ emanating from a prescribed altitude, is switched on at the beginning of the simulation. Any latent heating associated with the formation of this rain is neglected. Kessler microphysics (No ice physics, melting would deepen the layer of diabatic cooling by roughly a kilometer.) horizontal resolution is 7.5 km 10s time step 2-D turbulence No radiation or background wind A positive anomaly in the relative humidity in the upper troposphere, to reflect the high relative humidity of the anvil, is maintained for 36 h. rain shaft extends to 116-km radius, decreases linearly to zero. Coriolis parameter is at latitude 12N neglected the effects of deep convection in obs

Kessler microphysics

Kessler microphysics No ice physics, only supercooled water Kessler microphysics No ice physics, only supercooled water. Twice the specific heat than ice (Water: 4200J/kg C, ice:2100J/kg C) Decrease the same temperature, more latent heat

Control simulation Anticyclonic tangential wind

Cold core first forms above 2km and be advected down to the boundary layer. Extend to the surface within the innermost 90km. Explains why shallow convection can develop in the middle of the rain shaft (not shown). A decrease of temperature by 1K corresponds to 2.5K decrease of saturated . Evaporation efficiently reduces stability for shallow convection.

Negative RH anomaly to positive anomaly Moist level deepening to the surface

At the beginning, divergent flow near the surface, anticyclonic motion in the boundary layer. Downward advection of cyclonic wind which strengthens in both magnitude and horizontal extent. More sea surface flux into the atmosphere. The maximum wind speed isn’t at the lowest model level until 64h. There is still a cold poor but the core is not at the center.

Sensitivity Study

DI： more evaporation, larger downdraft and T anomaly. Maximum tangential wind not larger than control run, except in PBL. Larger downdraft leads to stronger anticyclonic wind which increases sea surface fluxes at the beginning stage. Outer convection occurs. HI: Warmer and drier PBL No outer convection. DA: Maximum tangential wind stronger than control run. Anomalies over a larger area. Wind anomaly closer to the sea surface than control. Outflow convers a larger area. Outer convection occurs. Outer convection weakens, storm starts intensification. HA: The same anomalies in , smaller wind.

Outer-region convection can reduce the rate of intensification Outer-region convection can reduce the rate of intensification. The outer convection seems to develop because of increased surface fluxes associated with the outflow and anticyclone at the model’s lowest level. Sensitivity study: Same as HDDI but no surface fluxes for r >340 km

Sensitivity study of midlevel humidity The initial middle-tropospheric relative humidity was increased by 30% in one experiment and it was decreased by 30% in another experiment. The results show that a dry middle troposphere favors the initial development of the cold-core vortex. However, a moist middle troposphere is slightly more favorable at later stages of the development. The overall development is not sensitive to the large-scale humidity in the middle troposphere.

Comparison between observation and model Agreement: Takes about 3~4 days to form a hurricane A cold-core vortex with high relative humidity at the beginning A warm core develops within the cold core Disagreement: It takes too long for model to form convection (due to lack of mean wind).

High humidity VS cold-core vortex (68km in radius, from 2.5 to 12.5 km) B2 contradicts simulation results of Emanuel (1995) showing that saturation of a mesoscale column is a sufficient condition for cyclogenesis. Extend to 150km a broad enough column of saturated air by itself can result in a hurricane.

B2 Low Cold PBL Sea surface PBL Why B1 develops?? 3K cold anomaly equals to 8K in , low level is more unstable inertial stability prevents inflow of low The downdraft in B3 brings smaller but the disturbance still develops.

For a depth of the layer of 4000 m, and vertical velocity of 0 For a depth of the layer of 4000 m, and vertical velocity of 0.1 m/s, the timescale is 10 h. W C W C C PV anomaly will be diffuse, vortex will not be strong W

A long lasting MCS forms Evaporative cooling (and anvil heating) results in a midlevel vortex Vortex extends to lower altitudes, and the cold anomaly expands downward Vortex wind enhances sea surface flux; Cold core reduces PBL and favors convection

convection Cold core vortex Warm core Downward motion Cold core downward Vortex downward Evaporation downward More sea flux convection Humid lower layer Warm core

Personal View How to define the stage of cyclone genesis? (Can a warm core vortex die?) Is it the only way to form a low level warm core vortex? Ignore the convection and heating and asymmetric motion in the model. How does a warm core develop inside a cold core (Or is shallow convection heating effective? Kessler?) Can a vortex form near the surface without downward motion? Stratification in stratification?