GULF SURGE EVALUATION: August 19, 2003
Barry Pierce National Weather Service
Las Vegas, NV
During the North America Monsoon Season, moisture surges from the Gulf of California
add significant low-level moisture to the Desert Southwest which greatly increases the
convective instability. This moisture enhancement is also critical in the development
of thunderstorms that produce intense rainfall rates and amounts. Due to significant
urbanization and the geologic formation of the Mojave Desert and Southern Great Basin
the intense rainfall produces rapid runoff which leads to significant flash flooding.
This module will concentrate more on the gulf surge and increased convective instability
rather than the actual flash flooding event.
SYNOPTIC OVERVIEW
On 17-18 August 2003, a surge of moisture raced up the Gulf of California reaching
Southern Nevada on 19 August. The significant increase in low-level moisture aided
in both the development and intensity of thunderstorms over the Las Vegas county
warning area. One specific storm occurred over the Northwest portion of the Las Vegas
Valley on the afternoon of 19 August. The Las Vegas WSR-88D radar (KESX) estimated
over 3 inches of rain in one hour while Clark County Flood Control District gages
measured greater than 2 inches of rainfall in less than one hour.
One of the most common synoptic patterns for gulf surges is an easterly wave over
central Mexico and an intensifying thermal low over the Desert Southwest. But, gulf
surges are difficult to forecast and track due to the sparse data network over Northwest
Old Mexico and the Gulf of California region. Many disturbances in the sub-tropical
easterlies that impact the Southwest U.S. in the summer originate over the Gulf of Mexico
region. This was the case as Tropical Storm Erika was located in the western Gulf of
Mexico on 15 August 2003. Erika made landfall just south of Brownsville, Texas on the
morning of 16 August (Fig. 1) and weakened rapidly.
The remnants of Erika continued west across Mexico and on 17 August helped generate strong,
deep convection over the Mexican states from Sonora south to Jalisco in the afternoon and
evening hours (Fig. 2).
The southern convective complex in Figure 2 then moved out over the Sea of Cortez during the
early morning hours of 18 August (Fig. 3).
Because of the weak large-scale environment over the Southwest in the summer,
one of the best approaches to synoptic monitoring is to make use of 24-hour change
data (Maddox and Howard). In this case, the pressure fall/rise couplet was analyzed
as the remains of Erika moved west. Using data from 1200 UTC as the analysis time,
sea-level pressure data from 5 sites was used starting back to 15 August 2003. The
five sites were Yuma and Kingman, Arizona, Blythe and Needles, California, and Las Vegas,
Nevada. The sea-level pressure and 24-hour change is shown in Table 1. As Table 1 shows,
sea-level pressure was highest on the morning of 16 August as Erika made landfall.
Twenty-four hours later pressures fell significantly at locations in the lower Colorado
River valley as the remains of Erika moved west across mainland Mexico. Further sea-level
pressure falls were noted on the morning of 18 August as well.
The 1200 UTC surface observations from Guaymas were not available, but upper air data was
available. The 1200 UTC 17 August sounding shows a northwest wind in the lowest 5000 feet
with precipitable water at 1.41 inches. A day later, 1200 UTC 18 August shows that the gulf
surge had passed Guaymas. Note, the cooling of the temperature profile and increase in
moisture below 5000 feet brought by the switch to southerly winds. The precipitable water
value increased to 1.96 inches. The two soundings are overlaid for comparison in Fig. 4c.
Using the VAD wind profile data from the Yuma WSR-88D radar (KYUX), a significant gulf surge
can be detected as the depth of the southerly flow extending to greater than 6000 feet with
wind speeds sustained at 40-45 knots. The strongest winds occurred between 0700 UTC and
0900 UTC (midnight - 2 am PDT) and occurred from the surface to 4000 feet (Fig. 5).
Sea-level pressure rises behind the surge in the cooler, denser air are shown in Table 1 on
19 August. The greatest rises occurred in the lower Colorado River Valley from Yuma, north to
Needles, with the passage of the surge at Needles occurring between 1100 UTC and 1200 UTC.
The strong southerly winds continued to carry moisture northward up the Colorado River Valley
during the morning hours. The moisture surge became evident in the Las Vegas Valley by 1800 UTC
as the valley inversion broke and the moisture and southerly winds mixed to the surface.
The Las Vegas surface dew point at 1200 UTC was 51 degrees F and 6 hours later it rose to
67 degrees F. Due to the location of the Desert Rock Sounding (75 miles northwest of the
Las Vegas Valley), precipitable water data was obtained by an Integrated Precipitable Water
Vapor (IPW) sensor located on the grounds of McCarran Airport (Fig. 6). Figure 6 shows,
precipitable water values near 1.20 inches at 1200 UTC August 19, with values steadily rising
to near 1.50 inches by 1800 UTC August 19.
CONVECTIVE INSTABILITY AND WSR-88D SUMMARY
Based on the 1200 UTC August 19 Desert Rock sounding (closest sounding to Las Vegas)
the forecast CAPE was 602 J/kg, which is not representative of conditions in Las Vegas.
The LAPS analysis at 1800 UTC showed CAPE values over the Las Vegas valley was around
2400 J/kg (Fig. 7) with thunderstorms initiating over the Spring Mountains west of
Las Vegas.
In the early afternoon, the storms began to migrate away from the mountains
and into the desert valleys. By 2100 UTC, CAPE values had increased to around
3000 J/kg (Fig. 8) and the IPW sensor (Fig. 6) showed values near 1.60 inches.
Thunderstorms northwest of Las Vegas started to propagate southeast or toward the
unstable air. At 2230 UTC, the intense rainfall producing thunderstorms began entering
the extreme northwest part of the valley and the Clark County Regional Flood Control
District network of gages began reporting rainfall amounts. By 2300 UTC, the
Lone Mountain/Hualapai (#4054) gage recorded 1.50 inches in 30 minutes. By 2330, the
Lone Mountain Detention Basin (#4269) gage reported 2.36 inches in one hour. The
one-hour radar estimated rainfall from KESX between 2235 UTC and 2335 UTC indicated
greater than 3.00 inches of rainfall over a large area of the northwest valley bounded
by Summerlin Parkway on the south and U.S. Highway 95 to the east (Fig. 9). A ground
truth report was received by a National Weather Service forecaster who reported 2.30
inches in 20 to 30 minutes.
CONCLUSION
Although the flash flooding from this episode was not as widespread as the 8 July,
1999 flash flood, the Las Vegas Metropolitan Police Department made 9 swift water
rescues by air, with nearly 60 total rescues in a 3 hour period. The Small Business
Administration estimated total damage at 4.8 million dollars with 251 homes and 173
businesses sustaining damage. This event demonstrates the value of tracking 24-hour
change data and the usage of a variety of data sources like the WSR-88D VAD wind
profile, IPW and LAPS data. The utilization of these resources is needed to stay on
top of rapidly changing mesoscale atmospheric conditions. Furthermore, the synoptic
scale NCEP models do not properly represent these smaller scale features.
|