Case Study: Hurricane Bill (2009)

A precise calculation of ventilation requires high resolution spatial and temporal data.

An observational dataset does not exist to the author’s knowledge that would allow such a calculation. However, there are some case studies that indicate the possibility that ventilation may be taking place. Once such example is Hurricane Claudette, which weakened from a hurricane to a weak tropical storm in about six hours (Shel-ton and Molinari, 2009). Inflow of dry air on the northwest side of the storm is hypothesized to have ventilated the eyewall leading to the hurricane’s quick demise.

Another more recent case study is Hurricane Bill (2009). Hurricane Bill was stud-ied as part of the Saharan Air Layer Experiment. The Saharan air layer is character-ized as an elevated mixed layer with an associated midlevel jet and is hypothescharacter-ized to inhibit tropical cyclogenesis and TC intensification, as it periodically sweeps across large portions of the tropical Atlantic (Dunion and Velden, 2004). The combination of dry air (low entropy) at midlevels along with increased shear makes TCs affected by the Saharan air layer prime candidates for collecting data to study the ventilation hypothesis.

Multiple missions were flown into Hurricane Bill on 8/20/09. From 8/20/09 at 03Z to 8/20/09 at 15Z, Bill weakened from 115 knots to 105 knots as the storm began to take on a more asymmetric appearance, and the eye began to look somewhat ragged. Thereafter, Bill held a quasi-steady intensity until more rapid weakening began later the following day. Total precipitable water analysis indicates that Bill

was surrounding by two Saharan air layer masses to the east and west. A tongue of drier air can be seen wrapping in toward the center of the storm on 8/20/09 right around the time weakening begins (Jason Dunion, pers. communication). The deep layer vertical wind shear on 8/20/09, as estimated from the Cooperative Institute for Meteorological Satellite Studies, was from the SE at 6-7 m s⁻¹ and increased to 8 m s⁻¹as the day progressed. Is it possible that the combination of vertical shear and dry air at midlevels halted the strengthening of Hurricane Bill and subsequently led to weakening thereafter?

Preliminary analysis from the Gulfstream-IV dropsonde data from 8/20/09 09Z to 14Z provides possible evidence for ventilation. Figure 5-1 shows the θeand radial wind in a frame of reference moving with the TC. The TC center and motion is deduced from a combination of vortex center fixes from a concurrent reconnaissance mission supplemented with best-track positions. Only data at and below 500 mb is shown in order to search for possible ventilation signals, namely, inflow coupled with negative entropy anomalies relative to the azimuthal mean. The azimuthal mean is a bit hard to deduce given the sparse data, but there does appear to be an area to the northeast of the center of anomalously low θe along with 7.5-10 m s⁻¹ of relative inflow between 700-800 mb. At 700mb, the θe is especially low with values between 325-330 K. Total precipitable water analysis at 12Z indicates that the dropsondes likely sampled a piece of the Saharan air layer getting wrapped into the storm. Granted, these observations are about 200-300 km from the storm center and closer to an outer rainband than the inner core. It would interesting to see if this inflow of low-θeair extends closer into the inner core, possibly by examining the flight-level data from the other reconnaissance mission.

Shortly after the flight, a large band of arc clouds was observed to emanate from a rainband on the west and north sides of the circulation. The arc clouds are the visual manifestation of a large cold pool that had been generated by downdrafts along the rainband. The outflow from these cold pools appears to have been captured in Fig. 5-1 at 800 mb and below to the northwest of Bill’s center. At 800 mb, there is one observation of 15 m s⁻¹ radial outflow with a θe of 330-335 K. Lower down,

−500 0 500

Figure 5-1: Dropsonde observations at 500, 600, 700, 800, 900, and 1000 mb from Hurricane Bill taken on 8/20/09 from 09-14Z. Colored dots indicate the value of θe

(K), and wind barbs are the relative radial flow (m s⁻¹). The axes are the zonal and meridional distances (km) from the TC center.

there is weaker outflow coinciding with anomalously low θ_e values in the same region associated with a low-level cold pool propagating outward.

A vertical profile of the relative radial and tangential wind, the vertical wind, and the θe are given in Fig. 5-2 for the dropsonde that is about 170 km north of the TC’s center. There are clearly strong downdrafts below 800 mb, which would flux the low-θe air observed above 950 mb downward. The profile of relative radial flow consists of inflow below 820 mb and weak outflow above. Hence, this dropsonde snapshot indicates at least the possibility of low-θeair being transported down into inflow layer of the storm, where it can then be advected toward the radius of maximum wind.

However, the large downdraft in the northwest region of the storm appears to be dominated by outflow at low-levels, so it is unclear how much low-θe air is getting drawn in toward the center.

Obviously, there are many gaps in the data that would allow one to get a better sense of what’s going on, but there are pieces of interesting evidence that appear to point to the possibility that midlevel ventilation and downdraft modification of the boundary layer affected the intensity of Hurricane Bill. It may be that the current observational platform is too limited to yield a more thorough dataset, but perhaps that will change in the near future with unmanned aerial vehicles and advances in dropsonde technology that will allow data to be collected in regions that are too risky to sample with conventional aircraft. Until then, there are probably a number of datasets that can be examined from past reconnaissance missions that may yield better evidence for ventilation.