Close Window

Univ. of Charleston: "Forecasting the development and movement of the sea breeze and its effect on severe weather "

Final Report


1.1 Introduction
1.1.1 Objectives (reproduced from our proposal)
The goal of this project is to increase our understanding of the sea breeze and its influence on severe weather along the coast of South Carolina and Georgia, and to refine a simple, but skillful sea-breeze forecasting tool that effectively depicts the location of the sea-breeze front. The objectives of our study consist of: 1) an extensive collection of meteorological data during the late spring and early summer of 1998, 2) case studies of sea-breeze events and their relationship to severe weather, and 3) refinement of a simple sea-breeze forecasting tool using knowledge gained from the case studies.

1.1.2 Participants
The Principal Investigator is B. Lee Lindner, Ph.D., an Associate Professor of Atmospheric Physics at the University of Charleston. The NWS partner is Mr. Stephen L. Brueske, Science and Operations Officer, National Weather Service Forecast Office, Charleston, South Carolina. Mr. James R. Frysinger, a graduate student in the Master of Science in Environmental Studies Program, a joint program of the University of Charleston SC and the Medical University of South Carolina, did the bulk of the work on this project and used this project as the backbone of his thesis. Other students involved in the project include Ms. Jessica Chenault, an undergraduate student at the College of Charleston; Mr. J. Michael Lindler, another undergraduate student at the College of Charleston; Mr. Christopher Sonekan, another graduate student in the Master of Science in Environmental Studies Program; and Ms. Sonya VanOordt, a high school student, Class of 1999, at the Academic Magnet High School, Charleston, South Carolina.

1.1.3 Division of Responsibilities
The responsibility for the overall direction of the project was split between Dr. Lindner and Mr. Brueske, both of whom were consulted frequently by the students who did most of the work. Dr. Lindner and Mr. Brueske also divided the project into several sections, which individual students used to participate in the study. Dr. Lindner and Mr. Brueske also split the duties of preparing the required reports. Dr. Lindner was responsible for handling the budget, and was responsible for the management of coursework that students were enrolled in as a result of this project (independent study research courses). Mr. Brueske worked closely with students in the design and implementation of the sea breeze forecasting tool, and worked directly with Ms. Sonya VanOordt as part of Charleston's Academic Magnet High School Mentor Program. The University of Charleston provided the TWI remote weather stations used in this project, and the NWS provided the batteries used to power one of these stations.

All funds for the project were used for student support. Students installed the remote weather stations, periodically collected the data from the weather stations, analyzed the data, and developed a forecasting tool to simulate the sea-breeze phenomenon, which was tested with additional data. Students have presented results from this project at American Meteorological Society conferences, and two manuscripts describing this project are being prepared for submission in peer-reviewed journals.

1.2 Description of research/development accomplishments
Field stations were employed by the College of Charleston at two key locations: the fishing pier at Folly Beach, SC and a rural inland site near St. George, SC. Data was collected from these two sites for about four months, except for a hiatus of about two weeks for maintenance and avoidance of severe weather (hurricanes). These stations are automated data-logging stations with the Folly Beach site providing sea temperature, air temperature, relative humidity, and barometric pressure and with the St. George site providing air temperature, relative humidity, barometric pressure, wind speed and direction, and rainfall. The sea temperature channel at Folly Beach is especially crucial to an on-going study to characterize the factors that are predictive of sea-breeze formation and behavior. Additionally, an automated weather station was set up in a suburban environment in Summerville, SC. This site provides air temperature, relative humidity, barometric pressure, and wind speed and direction. Data was collected from this site for about three months.

It was found that a portable station at the Folly Beach Fishing Pier (FBFP), Folly Island, South Carolina provided sea temperature data that yielded improved results in determining a sea-breeze index (SBI). The only local source available prior to this was a sensor mounted at a fixed height in the Cooper River, just above the Charleston Harbor. This places it 11.6-km inland from the ocean and the portion of the ocean that affects the sea breeze. The float-suspended sensor at FBFP also provides a more consistent reading since it is maintained at 0.6 m below the surface, regardless of the tide. Also, it is not affected by anthropogenic temperature changes caused by shipping or by water release from an upstream dam. This improvement should improve NWS forecasting of the sea breeze.

Other data sources were studied and considered for use in this analysis. It had been hoped that the station at St. George, South Carolina would be useful as a data source for sensing the synoptic wind field outside the domain influenced by sea-breeze cells. This is of concern because the sea-breeze front frequently moves inland past the CHS ASOS station. However, the inland distance of the St. George station was apparently too great. Another problem was a lack of precision in the station's wind direction readouts. In this case, this problem was not a limiting factor, in all probability. A statistical study was not done, but visual inspection of the data at the latter two stations indicated that there was probably little agreement between the two even when the front was well to seaward of CHS or even non-existent. A better technique might be to provide a line of stations located at increasing distances inland and to change the synoptic reference station selection as needed to maintain validity. Overruns of the CHS station were incorporated in the data as sea-breeze indicators in order to build the nowcasting models.

Other stations were also considered as data sources. Existing Handar stations at four locations were sampled by dial-in modem access for periodic data downloads. The Handar data dump format is extremely awkward for immediate use. Each data channel completes downloading all of its data for the period selected before the next channel downloads its data. Each channel has its own initial time stamp and individual data values are not time stamped. The time of the dial-in affects the staggering of times for each index value in the data arrays. Some of the stations had large errors in time and even in the dates in their internal clocks. These could be calibrated out by noting the time of the dial-in, but manual intervention is needed for each data channel download to note the time. Should these problems be solved, the stations at the City of Charleston Customhouse, the USDA Experimental Farm southwest of Charleston, and the station at Pinopolis Dam, Moncks Corner, South Carolina might be useful in future studies. The Handar station at Springmaid Pier, Myrtle Beach, South Carolina proved to be a useful data source for the proxy study done to correct the FBFP station's data following a nearby storm. The Springmaid Pier station maintained correct time and date. A program was written in Visual Basic to convert the Handar download to columnar form and tests showed it to work satisfactorily except that manual time corrections were still needed. The effort did, however, demonstrate the relative ease of using Visual Basic (and probably Visual C++) as a means of writing user-friendly programs with a graphical user interface. A portable Davis station was installed at a golf course in Summerville, South Carolina by the NWS Forecast Office, Charleston, but it was not fully operational until about halfway through the model test period.

Based on characteristics of the sea-breeze circulation cell and the physics involved, a preliminary forecasting tool was designed using weather data (including that from the FBFP station) from 1998 June. This was tested using data from the period of 1998 July through October and found to have improved nowcasting skills over climatology. Recalculation of the statistical parameters and probabilities using this latter data set improved the nowcasting skills considerably and led to refinements in the forecasting tool. The refined forecasting tool can be used with forecast information from mesoscale models available to the National Weather Service to provide short term forecast local flavor for the coastal area of South Carolina in the vicinity of Charleston. Extension of this and inclusion of other data sources, such as ocean temperatures at Springmaid Pier in Myrtle Beach and at places down the coast (e.g., Edisto Island; Savannah, GA; etc.) would extend this capability throughout the coastal portion of the area of responsibility of the National Weather Service Office in Charleston, South Carolina.

Once sea breezes initiated, the data became more difficult to characterize. Dynamic effects were noted in the data and were attributed to changes in cloud cover, storms (often a result of a sea-breeze front), and temperature changes due to cooling by the sea breezes. These dynamic effects could not be predicted by the forecasting tools built in this study and so subsequent hourly nowcasts might have been affected. Further improvement could be obtained by use of a local-scale model such as MM5, available from the University of Pennsylvania. This model allows flexible design of gridding plans and orthographic databases tuned to the orography of the area. Considerable tuning would be necessary to allow for the percentage of coverage and location of marshland versus dry land, urban heat island effects, and so forth and would probably take about two years. The use of a dynamic model is necessary for forecasting due to the effects the sea-breeze cell itself. These include the formation of a front, the front's speed of advance, and cloud cover (resulting from convective lifting) which all can have an effect on the driving forces behind sea breezes. These provide a negative feedback on the system that must be analyzed in the temporal domain.

Results from the investigations into the correlation between the Charleston Harbor water temperatures and coastal ocean temperatures at both the Myrtle Beach Pier and the Folly Beach Pier showed that Charleston Harbor water temperatures were not truly representative of near shore ocean temperatures. This information was used by the Nation Ocean Service (NOS) Center for Operational Products and Services when they conducted and published "A Requirements Assessment for a Charleston Harbor Physical Oceanographic Real-Time System, 17 Feb 99". This assessment provides a plan for the NOS to instrument the Charleston Harbor with numerous meteorological and oceanographic sensors that would benefit the commercial maritime industry and could be implemented as early as next year. Funding for the project will be provided by commercial maritime users of the Charleston Harbor.


As a result of the COMET Outreach Program collaboration, the University has become a UNIDATA site, and has begun additional collaboration in connecting the Charleston NWS office and the University of Charleston Weather Center. NSF funds have recently been secured to purchase hardware to enhance connectivity to sources of raw weather data with UNIDATA software, including particularly the local NWS office. We are excited at the possibilities this will have for enhancing the learning environment for students, for enhancing the research environment for faculty and for enhancing forecasting techniques for NWS personnel.

Also, this COMET Outreach Program collaboration has resulted in enhanced student interest in working with NWS personnel on other forecasting problems, and has led and will almost certainly lead to further student involvement at the local NWS office. In addition, it has directly led to another mentor project between the NWS and another Charleston Academic High School student who will graduate in 2000.

As a result of this collaboration with the University of Charleston, the Charleston NWS office was able to provide the Charleston Air Force Base Weather Station with detailed meteorological information which they are using as part of a local sea-breeze study. Their study attempts to develop forecast techniques that will accurately predict wind shifts and sea-breeze-initiated convection over the airfield, which directly impact military operations. Results from their study will directly benefit the aviation forecasts issued by this office.

The Charleston NWS office used preliminary data from this study to conduct an in-house seminar for their new lead forecasters. The forecasters were familiarized with the strong impact the sea breeze has on coastal winds, temperatures and precipitation patterns. In addition, interactions occurred between NWS forecasters and students on approximately a weekly basis. These interactions allowed forecasters to share their expertise and questions with the students while providing the forecasters valuable new insights into the local sea breeze using results obtained by the student's research. These exchanges have proved highly beneficial to both NWS forecasters and students by enhancing their knowledge of local weather phenomena.


Results from this project describing the improved sea temperature modeling were presented at the Fifth Mini-technical Conference of the Palmetto Chapter of the American Meteorological Society, held in Columbia, South Carolina on March 19, 1999. The abstract for this presentation was published in a proceedings volume (see full citation below). The sea-breeze modeling results from this project will be presented at the Third Conference on Coastal Atmospheric and Oceanic Prediction Processes, to be held in New Orleans from Nov. 3, 1999 to Nov. 5, 1999. An abstract describing this presentation has been published in the conference program (see citation below) and a six page manuscript describing this work will be published in a proceedings volume (see full citation below). A discussion of sea-breeze timing was presented in a senior thesis at the Academic Magnet School in Charleston, S.C. in May 1999. The thesis was also published (see full citation below). Lastly, all of the results for this project will be presented in a thesis in the Masters of Environmental Studies Program in Charleston, S.C., probably in August 1999. This thesis is also planned to be published in full (see citation below), and an abbreviated version of this thesis will be submitted for publication in a peer-reviewed journal, such as Monthly Weather Review or Weather and Forecasting.

Frysinger, J.R., Sea Breeze Initiation in Coastal South Carolina, Masters in Environmental Studies, M.S. Thesis, in Preparation, University of Charleston, South Carolina and the Medical University of South Carolina, 1999.

Frysinger, J.R., J. Chenault, B.L. Lindner and S. Brueske, Sea Temperature Monitoring with a Portable Weather Station near Charleston, SC (abstract), Proceedings Volume, Fifth Mini-technical conference of the Palmetto Chapter of the American Meteorological Society, p. 13, 1999.

Frysinger, J.R., S.L. Brueske and B.L. Lindner, Prediction and Characterization of Sea Breezes in Coastal South Carolina (abstract),, 1999.

Frysinger, J.R., S.L. Brueske and B.L. Lindner, Prediction and Characterization of Sea Breezes in Coastal South Carolina, Proceedings Volume, Third Conference on Coastal Atmospheric and Oceanic Prediction Processes, in press, 1999.

VanOordt, S., A Study of the Thermal Forcing of the Sea Breeze In Charleston, Senior Thesis, 29 p., Charleston Academic Magnet High School, Charleston, South Carolina, 1999.


4.1 University's Perspective
The participants in this project from the University of Charleston consider the project to have been a resounding success and extremely beneficial. Four students (two graduate students and two undergraduate students) received financial support from this grant, and three students were enrolled in research independent study courses while doing research on this project (Mr. James R. Frysinger received 36 hours of credit in EVSS 691, EVSS692 and EVSS699 between Summer 1997 and Fall 1999; Ms. Jessica Chenault received three hours of credit in PHYS 399 in Fall 1998; and Mr. J. Michael Lindler received three hours of credit in PHYS 390 in Fall 1998 and Spring 1999). Such financial support helped these students attend the University of Charleston, and the course credit these students received is very helpful in future employment applications or graduate school applications.

Unanticipated delays were one problem encountered. This project was conceived in the winter of 1996. There followed several months of revisions and signature gathering from NWS administration and it was not until the summer of 1998 that funds were received for the project. This delay caused some problems in the support for students and in the purchase of equipment through university funds. However, this thorough review of the proposal and the valuable feedback received as a result was quite valuable in the design of the project and this feedback was deeply appreciated by the project directors.

The exposure of two of our graduate students and one of our undergraduate students to operational forecasting has directly aided their learning. The data, which has been acquired, has been very useful as classroom presentation material, again enhancing student learning. Several new courses are being developed which will now make use of presentations by NWS personnel; tours of NWS facilities and data prepared by NWS personnel.

This collaboration has also led to collaboration in other areas, notably in discussion with psychology faculty in the design of effective hurricane warning messages. The recent addition of more staff at the Charleston NWS office will also lead to enhanced access to new observing systems and improved courses.

4.2 NWS Perspective
This project has shown that one of only two water temperature sensors along the South Carolina coast is unrepresentative of near-shore ocean surface temperatures, and is therefore not particularly helpful in forecasting the sea breeze. The sensor in the Charleston Harbor is located at the Ports Authority passenger terminal more than 11 km from the ocean and is fixed to a piling. Its depth in the water column therefore changes with the tides, and at extremely low tides the sensor comes out of the water and measures air temperatures. The sensor is also in an area where cooling water discharge from large ships and periodic openings of the sluice gates at the dam upstream can bias the temperature readings. The Charleston NWS office and the University of Charleston have worked with the County Parks Commission to place a floating fixed-depth sensor at a more representative site on the Folly Beach pier. This is on a temporary basis, and data is unavailable in real-time. However, the NWS is attempting to make this a permanent site that will provide real-time, representative data to the local NWS office.

As mentioned previously, the meteorologists at Charleston Air Force Base are developing new forecasting techniques to forecast the impact of the sea breeze on operations at the Base.

The Charleston NWS office is analyzing this data and looking for new ways to forecast the onset and movement of the sea breeze and its impact on convection. Implementation of a Physical Oceanographic Real-Time System by the NOS and Charleston maritime community may occur during the next year and would provide the Charleston NWS office and maritime users additional high resolution meteorological and oceanographic data from seven different locations throughout the harbor, harbor entrance offshore jetties, and navigable rivers that penetrate inland from the harbor. High time and space resolution observations would be ideal to use with high-resolution non-hydrostatic meteorological models (such as RAMS, or MM5). The Charleston NWS office anticipates running such a model locally with a University of Charleston student during the next school year. The detailed dataset obtained as a result of this partners project will be used to verify the ability of the model to accurately depict the movement and intensity of the sea breeze along the South Carolina coast.

This project has served as a catalyst to bring scientists from the university, NWS, and DOD communities together to work on solving common problems. Continued collaboration between the various communities will surely prove equally beneficial to all.