On 1 October 1986, two NOAA WP-3D aircraft were flown in Hurricane Paine as it passed just south of the Baja California Peninsula, providing an opportunity to observe how the category-1 hurricane responded to the presence of a 1200-m-high ridge. Ten fixes of the circulation center and 25 passes through the eyewall show no apparent impact by the peninsula on the track and intensity of Paine. Reflectivity fields reveal a subtle response with a distortion of the eyewall shape and the development of a weak reflectivity zone downwind of the ridge. The minimal response of the hurricane is hypothesized to be due to the large-scale trough in which the hurricane is embedded. Southwest winds inhibit the development of downslope flow from the Baja ridge and the subsequent introduction of this drier air into Paine's inner core.
On 1 October 1986, Hurricane Paine grazed the mountainous Baja California Peninsula on its way to landfall 300 km northwest of Mazatlan, Mexico, on the following day. The two National Oceanic and Atmospheric Administration (NOAA) WP-3D aircraft flew repeatedly through the western eyewall during 1 October and provided a rare opportunity to observe how the inner circulation of a tropical cyclone (TC) responds to the presence of a 1200-m-elevation ridgeline (Fig. 1). During the experiment, the circulation center came within 100 km of the Baja ridge, and the outer portions of the eyewall impinged on the high terrain shortly after the aircraft retired from the TC.
a. Prior work
Alterations in TC track caused by the presence of land have been discussed by Brand and Blelloch (1973, 1974), Chang (1982), Bender et al. (1985, 1987), Yeh and Elsberry (1993), and Lin et al. (1999). Of particular interest is the simulated response of a TC as it approaches a mountainous island (Bender et al. 1987). In a background easterly flow, islands with a north–south ridge (e.g., Luzon, Taiwan) generally produce a course deflection northward and an increase of TC speed of a few meters per second. Bender et al. (1987) emphasize that the deviations triggered by the presence of a mountain range can be equal to the typical forecast errors at 12 and 24 h. Observational results are mixed, with some TCs apparently not responding to the presence of mountains (Brand and Blelloch 1973) and others reacting hundreds of kilometers from the terrain supposedly triggering the response (Yeh and Elsberry 1993).
Changes in intensity; the kinematic, thermodynamic, and precipitation fields; and the overall symmetry of the TC are believed to occur because of the nearness of land and prior to the circulation center making landfall (Brand and Blelloch 1973, 1974; Tuleya and Kurihara 1978; Bender et al. 1985, 1987; Tuleya 1994; Li et al. 1997). Processes that trigger these alterations in the TC include subsidence from mountain ranges, the sharp and substantial decrease in evaporation over land, changes in the sensible heat flux, increases in surface friction, changes in the stability, and the advection of continental air into the TC circulation (Wu and Kuo 1999). Intensity diminishes chiefly in response to the drier air that subsides and enters the inner circulation of the TC (e.g., Bender et al. 1987). Occasionally an actual TC does react in a similar fashion, though the magnitude of the change is usually less (Brand and Blelloch 1973, 1974). Many of the aforementioned studies involved Taiwan. The Baja ridge near Cabo San Lucas is similar in orientation and latitude to the southern portion of the Central Mountain Range of Taiwan, but its mean height is about 500 m lower. We expect the response of a TC to either barrier to be broadly similar.
We seek to determine if the Baja ridge has any influence on the track, intensity, and other traits of Paine. A number of scenarios are envisioned. The TC may first deflect away and then toward the Baja ridge as it passes to the south of Cabo San Lucas. This deflection would be in response to enhanced environmental flow parallel to the Peninsula from the north-northwest, typical of October (Sadler et al. 1987). If the TC approaches Cabo San Lucas from the southwest, there may be upslope flow along the Gulf of California coast and accompanying deep convection. Subsidence of midlevel dry air on the Pacific Ocean side of the ridge may alter the precipitation field to the left of track, and if this dry air is mixed into the eyewall, it might reduce intensity. Low-level flow impinging on the Baja Peninsula mountains from the east may be deflected and enhance the radial flow to the eyewall, leading to stronger convection and, perhaps, intensification or changes in the radius of the maximum winds (RMW). In general, the rearrangement of where latent heat is released with respect to the circulation center has the potential to alter the organization of the TC.
The advantage of this study over prior work focused on landfall–TC interactions is that the research flights provide superior temporal and spatial resolution of the eyewall region. A disadvantage is that we cannot recognize any mutation induced by the Baja Peninsula that may have occurred prior to aircraft sampling, when Paine was much farther away.
2. Data and sampling strategy
NOAA aircraft No. 42 flew 16 radial legs between 270- and 770-m altitude, and NOAA aircraft No. 43 completed 9 radial passes at 1500- and 3000-m altitude in the western eyewall of the TC. Ten fixes of the circulation center were obtained; Fig. 1 shows a subset of these passes.
The specifications for the in situ instruments are presented by Jorgensen (1984) and the details of the lower-fuselage and tail radars are described by Marks (1985). Aircraft positions are determined using the inertial navigation system, which is subject to long-term drift (Fankhauser et al. 1985). NOAA-43's position is adjusted by assuming an increasing linear rate of error based on differences in the reported aircraft position between takeoff and landing. NOAA-42's drift error is negligible for this flight. No wind errors are detected for the two aircraft, based on the calibration maneuvers conducted on the ferry to the TC. The temperature and dewpoint sensors are compromised in saturated conditions, and so we use the carbon dioxide radiometer correction scheme discussed by Jorgensen and LeMone (1989). The 1-Hz data are smoothed using a 2-km running mean. More details concerning data treatment are discussed by Daida (2002).
Analyses for variations in the minimum sea level pressure include adjustments for the semidiurnal pressure oscillation. The circulation center positions are based on the technique developed by Willoughby and Chelmow (1982). In this technique, the flow is assumed to be in gradient balance. The intersection of several lines drawn normal to and to the left of the winds identifies the circulation center. No corrections for a possible tilted eye are made, because the aircraft were usually separated by only 1000 m of altitude. Even a 60° tilt from the vertical would shift the upper aircraft fixes relative to the lower fixes by less than 2 km. There is also no guarantee that the vertical axis would not wobble, compounding any tilt correction.
The National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis, with a horizontal resolution of 2.5° latitude, is used to identify the environment for the TC. We examine the evolution in the environmental conditions because they might mask any effect the topography of Baja California Peninsula may have an Paine. Environmental fields include deep-layer vertical shear of the horizontal wind (200–850 hPa), sea surface temperature (SST), and moisture content at midlevels. These variables are collected for grid points that fall in an annulus from 250 to 750 km from the TC center; means for this annulus are determined at 0000 and 1200 UTC. Prior work on developing versus nondeveloping typhoons (McBride and Zehr 1981) and TC motion (Elsberry 1995) led us to choose this size of annulus. Tropical cyclones show less sensitivity to environmental factors such as shear at ranges much beyond 750 km. We eliminate data within 250 km because it is most likely dominated by the TC circulation.
a. Negligible environmental evolution
Trends in the mean conditions for the 250–750-km annulus around Paine show only subtle evolution for 36 h before to 12 h after the research flights. SST is nearly constant at 28.3°C (Fig. 2a). Deep-layer vertical shear varies from 10 to 15 m s−1 after an initial high value at 0000 UTC 29 September (Fig. 2b). Midlevel thermodynamic conditions reveal inconsequential variations prior to and during the experiment (Figs. 2c–e). Note that shear and midlevel moisture, though steady, are not ideal for intensification and may be partially responsible for Paine achieving only category-1 status. Maximum potetial intensity for Paine, assuming outflow temperature to be the 200-hPa level, is 55 m s−1, based on the work by Emanuel (1986). The steadiness of the fields improves the likelihood of detecting any effects caused by the presence of the Baja ridge.
b. Track is steady
Gunther and Cross (1987) state that Paine came under the influence of a deep trough. The best track (Fig. 3) does not supply any evidence that the TC moved erratically or contrary to the deep-layer mean flow associated with this trough. With 10 circulation center fixes, we have the opportunity to identify any smaller-scale deviations that may have been masked by the coarse resolution of the best track. The meridional and zonal positions of the circulation center as a function of time (Fig. 4a) reveal that Paine moved with a constant velocity when near the peninsula. The linear regressions of the meridional and zonal positions as a function of time have correlation coefficients of 0.99 and 0.67, respectively. The zonal correlation coefficient is lower because the u component of Paine is very small and therefore is sensitive to position errors, typically 1–3 km, for any given fix. Motion is steady (u = 0.3 m s−1, υ = 6.1 m s−1) over the course of the experiment. The track of Paine (Fig. 4b) is well represented by a constant heading, based on the 10 circulation fixes obtained with both aircraft.
c. Intensity is steady
Minimum sea level pressure (MSLP), deduced from aircraft static pressure, decreases less than 1 hPa as the circulation center moves to within 100 km of the peninsula (Fig. 5a). This time is also when the outer portion of the eyewall crosses over Cabo San Lucas. The maximum sustained tangential wind relative to the circulation center is also steady (Fig. 5b). This estimate is a mean over 5 km centered on the maximum tangential wind found within a broad annulus ranging from 42 to 76 km from the circulation center (Fig. 5c). We have chosen this measure because the tangential wind as a function of radius is not the typical profile one associates with a TC, because there is no sharply defined radius of maximum winds (Fig. 6). We believe that an indistinct RMW is indicative of a TC that has little convective activity within the eyewall, is not undergoing eyewall contraction, or is in the decay phase, based on our interpretation of the work by Willoughby et al. (1982), Weatherford (1989), Stossmeister and Barnes (1992), and Willoughby (1990).
d. Reflectivity fields altered
We interpret that the Baja California Peninsula has an effect, though subtle, on Paine, based on the evolution of the reflectivity fields. First there is an apparent distortion in the stratiform bands in the northern portion of the eyewall and adjoining bands. There is a flattening of the reflectivity (Fig. 7a) that appears early in the experiment, and later, when the TC is much nearer to the peninsula, a kink is apparent (Fig. 7b). As the TC slides by the peninsula, these distortions fade. Partial blocking of the low-level flow from the northeast is a possible cause.
The second feature of interest is the development of a weak reflectivity zone downwind of Baja California. Near the peninsula, the region is about 50 km wide (Figs. 8a,b). It narrows as it moves downstream (Fig. 8c) and eventually disappears in the southwest quadrant (Fig. 8d). The weak reflectivity region may be initially due to either subsidence in the lee of the high terrain or more low-level divergence as some of the air flows around the ridge. One effect would tend to develop the other.
The third aspect of the reflectivity field is the absence in either the tail radar scans (Figs. 9a,b) or the lower-fuselage scans of any deep cumulonimbi induced on the eastern, upslope side of the peninsula. The reflectivity patterns of the ridgeline itself are expanded in the vertical direction because of incomplete beam filling. The returns in the lower-fuselage radar scans (e.g., Fig. 7b) are stationary and invariant in shape throughout the course of the research flight and, thus, are interpreted as terrain. The lack of any cumulonimbi on the windward side of the ridge lends credence to the argument that flow is not being forced upward for any substantial depth along the Gulf of California side of the peninsula.
The fourth feature of interest is the cluster of high-reflectivity cells initially observed in the west-northwest portion of the eyewall (Fig. 10a). These cells move cyclonically around the eye (Figs. 10b–d). Our first hypothesis was that these cells formed in response to enhanced radial flow caused by the barrier effect of the peninsula. However, the feature is persistent and could be tracked for nearly 3 h before the eyewall was near the peninsula (Fig. 11). The mean speed is 20 m s−1, about one-half of the tangential speed in the eyewall. Tail radar reveals that the persistent feature has a multicell structure, with the newer cells downwind, or on the leading-edge side (Fig. 12a). The multicell often contains reflectivity tops that are 5–7 km taller (Fig. 12b) than what is observed in the rest of the eyewall or anywhere else in the TC. Most of the reflectivity fields contain a well-developed bright band (Figs. 12c,d), evidence of stratiform conditions.
The multicell feature is located along the inner edge of the eyewall where there is a steep gradient of vorticity (Daida 2002). Montgomery and Kallenbach (1997), Schubert et al. (1999), and Reasor et al. (2000) have associated features such as this with a manifestation of a vortex Rossby wave. Reasor et al. (2000) argue that a vortex Rossby wave would move with about one-half of the speed of the maximum wind and that enhanced convection would result from the asymmetric flow associated with the feature.
Kossin and Schubert (2001) simulated mesovortices in the eyewall that form in strong or intensifying TCs. These mesovortices distort the flow, resulting in polygonal shapes in the eyewall. Several of the lower-fuselage scans of Paine (e.g., Figs. 7a,b and 10b,d) do have distortions of a polygonal nature. Though Paine is steady and weak, the multicell feature might be responsible for the minor distortions in the reflectivity field.
4. Discussion and conclusions
Prior researchers have identified a potentially strong interaction between high terrain and an approaching TC, but for Hurricane Paine the interaction resulted in only subtle variations in the reflectivity field. Track, intensity, and the outer wind field are essentially steady as the TC passes by the Baja California Peninsula, and the asymmetries in the convective activity in the eyewall are found to exist long before the eyewall grazed the ridge and are more likely a manifestation of a vortex Rossby wave.
Why does the peninsula have so little impact on the track or intensity of Paine? We believe that there are two factors. First, and foremost, the TC is embedded in the southernmost portion of a synoptic-scale trough (Fig. 13). The shape of this trough inhibits the cyclonic circulation to the north of the TC track, where the environmental southwest flow opposes the east-northeast wind associated with the TC. This interaction results in a more elliptical wind pattern, with much more flow parallel to the ridge than one might have expected. This flow regime inhibits subsidence and dry air from entering the inner core of the TC, which is argued by Bender et al. (1987) to be the major factor that weakens a simulated TC approaching high terrain.
The second factor is the high thermodynamic stability of the inner core. We infer this from the stratiform reflectivity structure of the eyewall and almost all of the features radially beyond. The only deep convection is the multicell feature swirling around the inner edge of the eyewall, which is the sole source of latent heating to maintain the upper-level warm core. This stable condition both limits convective development on the eastern, windward side of the peninsula and lessens the extent of subsidence on the western side.
The key issue for forecasters is to not treat the oncoming TC in isolation. The large-scale flow strongly modulates how a ridge affects TC precipitation fields, track, and intensity. We interpret these observations to be a verification of the modeling studies by Bender et al. (1985, 1987) that emphasized the importance of the mean flow in which the TC is embedded and not just the TC circulation itself.
We thank the seasoned aircrews from NOAA/OAO and NOAA/AOML/HRD for their efforts during the research flights into Paine. We also thank the reviewers for their suggestions. This work was supported by National Science Foundation Grants ATM-9714400 and ATM-0239648. Garpee Barleszi said, “if you fly, you must write about it,” and so we have.
Corresponding author address: G. M. Barnes, Dept. of Meteorology, University of Hawaii at Manoa, 2525 Correa Rd., Honolulu, HI 96822. Email: firstname.lastname@example.org