1208 MONTHLY WEATHER REVIEW VOLUME 135 Influence of Tropical Cyclones on Humidity Patterns over Southern Baja California, Mexico LUIS M. FARFÁN Centro de Investigación Científica y de Educación Superior de Ensenada B.C. (CICESE), La Paz, Baja California Sur, México IRA FOGEL Centro de Investigaciones Biológicas del Noroeste (CIBNOR), La Paz, Baja California Sur, México (Manuscript received 6 February 2006, in final form 10 July 2006) ABSTRACT The influence of tropical cyclone circulations in the distribution of humidity and convection over north- western Mexico is investigated by analyzing circulations that developed in the eastern Pacific Ocean from 1 July to 21 September 2004. Documented cases having some impact over the Baja California Peninsula include Tropical Storm Blas (13–15 July), Hurricane Frank (23–25 August), Hurricane Howard (2–6 Sep- tember), and Hurricane Javier (15–20 September). Datasets are derived from geostationary satellite imag- ery, upper-air and surface station observations, as well as an analysis from an operational model. Emphasis is given to circulations that moved within 800 km of the southern part of the peninsula. The distribution of precipitable water is used to identify distinct peaks during the approach of these circulations and deep convection that occurred for periods of several days over the southern peninsula and Gulf of California. Hurricane Howard is associated with a significant amount of precipitation, while Hurricane Javier made landfall across the central peninsula with a limited impact on the population in the area. An examination of the large-scale environment suggests that advection of humid air from the equatorial Pacific is an important element in sustaining tropical cyclones and convection off the coast of western Mexico. 1. Introduction cyclone activity serves as a source of humid air masses that provide conditions that support deep convection There were 12 named tropical cyclones in the eastern over northwestern Mexico, including the Baja Califor- Pacific Ocean from late May through October 2004 nia Peninsula. These conditions tend to occur through- (Avila et al. 2006), with 9 events occurring between July out the summer, with the most favorable locations over and September. Most of these systems remained at sea. the southern peninsula. This pattern is associated with Late in the summer Tropical Cyclone Javier made land- increased cloud cover followed by organized convective fall as a depression, moved across the Baja California systems that may result in heavy precipitation (Farfán Peninsula on 19 September, and reached mainland 2005). Mexico the next day. In contrast 15 systems developed Figure 1 shows the frequency of named storms from in the North Atlantic, but none made landfall along the 1990 to 2004, information derived from the best-track Gulf of Mexico or Caribbean Sea coasts of Mexico dataset available at the U.S. National Hurricane Center (Franklin et al. 2006). (NHC). The 219 storms during this 15-yr period yield a This study analyzes humidity and convection during mean of 14.6 storms per season. This is slightly lower the approach of tropical cyclones to Baja California in than the mean of 15.5 during the study period of 1971– the eastern Pacific basin. This basin encompasses the 2003 (Levinson 2005). Minimum activity occurred in area north of the equator from the west coasts of Cen- 1996 and 1999, with 9 tropical cyclones per season, con- tral and North America westward to 140°W. Tropical trasted with 24 cases in 1992. Figure 1 also indicates landfall events in western Mexico. The darker bars in- Corresponding author address: Dr. Luis M. Farfán, CICESE, dicate that from zero to four cases made landfall each Unidad La Paz, Miraflores 334, La Paz, BCS 23050, México. year, of which one or sometimes two cases, moved E-mail: [email protected] across the Baja California Peninsula. DOI: 10.1175/MWR3356.1 © 2007 American Meteorological Society Unauthenticated | Downloaded 09/29/21 05:35 PM UTC MWR3356 APRIL 2007 F A R F Á N AND FOGEL 1209 FIG. 1. Named tropical cyclones in the eastern Pacific basin (light bars) with the dotted line representing the seasonal mean (14.6). Tropical cyclones that made landfall along the west coast of Mexico from 1990 to 2004 are given by dark bars. Numbers in parentheses indicate cases making landfall in the Baja California Peninsula. Tracks of two tropical cyclones that developed dur- This study focused on tropical cyclones that had a ing the 15-yr period are shown in Fig. 2. The circles are significant impact on the Baja California Peninsula dur- used as a reference of distance of the storm circulation ing the 2004 storm season and documents the regional with respect to the southern peninsula (centered at and large-scale conditions present during the develop- 25.4°N, 111.6°W). Hurricane Juliette (2001), associated ment of these storms. This is an important season to with heavy precipitation and strong winds, is represen- examine, in part, because several systems moved close tative of landfall events with a strong impact on the to the southern part of the peninsula (22.9°–28.0°N) southern part of the peninsula. Juliette had a relatively and one of them, with a track similar to that from Juli- long track that started beyond the 1200-km radius, off ette (2001), made landfall there. The specific objectives the southeastern coast of Mexico, and made landfall on are as follows: the western side of the peninsula (Farfán 2004). As shown in Fig. 2, the system moved across the peninsula • based on the distance from the southern peninsula and northward over the Gulf of California, weakening and best-track positions, classify the development of over the following few days. tropical cyclones that occurred between July and In contrast to the impact from systems making direct September 2004, landfall, the passage of tropical cyclones with northwest • evaluate the vertical and horizontal structures of hu- tracks also provides rainfall. Accumulation of rainfall midity to provide documentation on the development seems to be directly proportional to the distance of the of convection and rainfall associated with the closest storm center (Englehart and Douglas 2001) from the approach of these tropical cyclones, and impacted area, making these tropical cyclones a • describe the impact caused by the landfall of Hurri- complementary and essential source of rainfall. For ex- cane Javier from observations recorded by the re- ample, while located 400–500 km southwest of Cabo gional network of surface stations. San Lucas, Tropical Cyclone Linda (2003; see Fig. 2) brought convective outbreaks on two consecutive days This paper is divided into six sections. Section 2 de- (Farfán 2005). Satellite and in situ data indicate that a scribes the data sources. Best-track information and the large-scale mass of moist air from the storm’s eastern evolution of humidity are described in section 3. Char- flank was advected into the southern peninsula and acteristics of the regional and large-scale environments gulf. for selected storms are discussed in sections 4 and 5, Unauthenticated | Downloaded 09/29/21 05:35 PM UTC 1210 MONTHLY WEATHER REVIEW VOLUME 135 FIG. 2. Best-track positions of Tropical Cyclones Juliette (2001) and Linda (2003). Positions are at 6-h intervals. Direct- and indirect-influence tracks are derived from Higgins and Shi (2005) and are explained in the text. Reference circles are drawn at 400-, 800-, and 1200-km radii. The MX label represents the position of maximum storm intensity, and specific geographic locations used through- out the paper are indicated. respectively. Section 6 provides the summary and con- dictions of warm season precipitation over North cluding remarks. America (Higgins et al. 2006). During the 2004 field season, soundings were taken twice daily (1200 and 2. Data sources 0000 UTC) for stations in northwestern Mexico and the The best-track dataset of tropical cyclones in the southwestern United States. During intensive observ- eastern Pacific, compiled by the NHC, provides a rec- ing periods (IOPs), soundings were taken at intervals of ord of the position of the center of circulation every 6 h 4–6 h over several consecutive days. (Avila et al. 2006). The dataset includes categories of in- Digital imagery from the Geostationary Operational tensity: tropical depressions (sustained wind Ͻ 17 m sϪ1), Environmental Satellite-10 (GOES-10) provides pat- tropical storms (sustained wind Ͼ 17 m sϪ1), and hur- terns of the mid- to upper-level humidity and the struc- ricanes (sustained wind Ͼ 33 m sϪ1). Our analysis cov- ture of the cloud cover. This includes images of the ers the storm events from 1 July to 21 September 2004. water vapor and infrared channels at resolutions of 8 Upper-air soundings were examined to identify and 4 km, respectively. Additional information on the large-scale characteristics of the atmospheric environ- three-dimensional structure of large-scale flow is sup- ment over northwestern Mexico, the southwestern plied by the gridded analyses of the Global Forecast United States, and the adjacent oceanic areas. These System (GFS) model. This model is operated by the data include regular (1200 UTC) releases and special U.S. National Centers for Environmental Prediction soundings taken for the North American Monsoon Ex- (NCEP), and analyses are available at a grid resolution periment (NAME), which is intended to improve pre- of 100 km ϫ 100 km. Unauthenticated | Downloaded 09/29/21 05:35 PM UTC APRIL 2007 F A R F Á N AND FOGEL 1211 FIG. 3. Tracks of tropical cyclones developing from 12 Jul to 26 Aug 2004 in the eastern Pacific Ocean. Initial positions are identified with name initial and two-digit day, subsequent positions are at 6-h intervals. Circles are drawn at 400-, 800-, and 1200-km radii.