87th AMS Annual Meeting

Wednesday, 17 January 2007: 4:15 PM
An analysis of the Northern Gulf of Mexico sea breeze and associated thunderstorms
206A (Henry B. Gonzalez Convention Center)
Christopher M. Hill, Geosystems Research Institute, Stennis Space Center, MS; and P. J. Fitzpatrick, Y. Lau, J. Corbin, S. Bhate, and P. G. Dixon
Much of the summertime weather along the Northern Gulf of Mexico is associated, either directly or indirectly, with the land-sea interface. A regional composite study of the summertime wind patterns and rainfall patterns was conducted. Some of the factors influencing rainfall also were investigated.

Three years of summertime radar data were archived for the Slidell, LA region, which covers southeastern Louisiana, southern Mississippi, southern Alabama, and the immediate coastal waters. Surface observations were also archived from buoys, ASOS, and Remote Automated Weather Stations (RAWS). The goal was to examine rainfall and surface wind patterns in this region associated with typical summertime conditions. Days influenced by upper-level troughs, surface troughs, tropical cyclones, and fronts were removed, leaving 168 days of "typical" summertime weather.

Composite surface wind results The sea breeze plays a prominent role in summertime surface wind patterns. The primary wind direction is from the south or south-southwest in most cases. In most land locations, the wind reaches a minimum around 4AM local time, with occasional calm conditions (winds less than 3.35 mph). The wind speed increases gradually but steadily after 6AM. After 12PM local time, the wind speed in many land locations increases abruptly, peaking for several hours around 7-10 mph. The full effect of the sea breeze generally takes place during the early afternoon. Associated with this period of higher wind speed is a shift in wind direction to the south or the south-southeast. The wind begins to decrease after 4PM. This diurnal pattern of the local wind is seen in all three months, and is most prominent near the coastline.

A few deviations from this pattern are noted. The Slidell NWS office consistently observed an east wind after midnight until sunrise, either due to effects from local topography or from Lake Pontchartrain. Two inland observing stations south of Lake Pontchartrain (one in New Orleans) did not show a wind shift.

However, the most important departures from this pattern come from two buoys located approximately 20 miles from the Mississippi coast, in Breton Sound. During the noon-4PM period of increased wind over land, the wind decreased at these two buoys. This indicates that the pressure gradient in Breton Sound decreased in association with mass adjustments occurring over land. The wind in the Breton Sound is typically fastest in this region at night and in the early morning. Furthermore, the wind is faster at these times than over inland regions. As will be discussed, this cross-shore wind configuration may be affecting offshore rainfall patterns.

Composite rainfall pattern results Rainfall occurrences were tallied for each radar scan returning a base reflectivity value greater than 30 dBz. They rainfall occurrences were summed over a period of time (1 h, 3 h, and 12 h). This methodology provided graphical results of rain coverage evolution, allowing for the identification of regions experiencing a consistent clustering of thunderstorms. While scattered thunderstorms were possible in any location, a clear climatology was discerned.

Around 4AM, precipitation typically forms south of the Louisiana coast, south of the Alabama coast, and in Breton Sound east of the Louisiana marsh and south of Mississippi. Concentrations of precipitation also formed south of concave bay regions such as Atchafalaya Bay and Barataria Bay. Overt time, offshore rainfall increases in coverage, peaking at 7AM and 8AM. As remarked earlier, the winds are faster offshore than inland. The wind direction throughout the period is south or southwest, so a land breeze is not forming in the traditional sense. Speed convergence of the southerly wind is hypothesized to occurring offshore, providing lift for rainfall formation. Enhanced convergence, or outflow interactions, increase the thunderstorm coverage until the sea breeze is reinitiated during the daytime.

Precipitation begins to form around 10AM along the immediate coast. This activity spreads inland with time. After 12PM, offshore thunderstorm coverage and activity begins to decrease markedly. Inland activity peaks in the late afternoon, and begins to noticeably decrease after 7PM. By 10PM, and up until 3-4AM, rainfall is generally non-existent. Some local regions are preferred for more rainfall: north of Mobile Bay, and the northwest panhandle region of Mississippi. New Orleans also shows a local maximum due to a possible interaction of the Lake Pontchartrain sea breeze with the Gulf sea breeze, and/or due to urban heat island effects. Less precipitation was routinely observed in August compared to June and July. It was hypothesized that more stablilty or more subsidence existed in the atmosphere during August, but plots of CAPE, lapse rate, and mid-level dewpoint depression showed only a very weak trend to support these possibilities. The reason for the decrease in August precipitation is yet to be determined.

Factors affecting rainfall coverage Scatterplots and regressional calculations were performed of regional rainfall coverage versus a variety of potential influences. The 00Z and 12Z Slidell sounding data were used in these computations. No correlation with CAPE was found - a somewhat surprising finding. The synoptic wind direction was examined for influences on rainfall coverage. Rainfall coverage was noted to be greater with a 850-mb south wind, and the lowest rainfall coverage was associated with a 850-mb north wind. Inland precipitation coverage also increased (decreased) with a 850-mb south (north) wind, consistent with other studies about sea breeze convection. Rainfall coverage was also correlated with the K-Index. A detailed analysis showed that, in most situations, the 700-mb dewpoint depression term of the K-Index contributed the most to this relationship by a factor of 2 to 8, followed by the 850-mb dewpoint temperature term; the lapse rate term often contributed weakly (usually over land) or not at all.

Multiple regression analysis was completed using a modified 850-mb wind term (wind direction value minus 180 degrees) and the K-Index. The variance was computed for both the whole radar region, as well as for quadrants of the radar region, for 4-h rain coverage. The largest variance explained was 23%. In most cases, the K-Index was 2-3 times more important than wind direction, but in some cases the two terms contributed equally to the variance.

Supplementary URL: