SECTION 1. PROJECT OBJECTIVES AND ACCOMPLISHMENTS
The sea breeze commonly occurs along the southern Maine coastal zone during the spring, summer, and fall months. Passage of the sea-breeze front is often accompanied by a wind shift, drop in temperature, and change in moisture content. Therefore, the sea breeze plays an important role in controlling the maximum temperature an inland location realizes during the day and its attendant wind shift is an important aviation hazard.
The objective of this research is to develop a forecast scheme that will predict the onset time of sea breezes observed along the southern Maine coast.
Previous research has shown that sea-breeze propagation rate is dependent upon the maximum temperature difference between the water and land realized during the day (DTmax) and the strength of the large-scale flow that opposes the sea-breeze front (Uopp). Uopp is therefore, the component of the large-scale flow directed perpendicular to the frontal orientation. If the maximum temperature difference is large or if the opposing flow is weak, one would expect the sea breeze to move inland quickly.
Motivated by these results, we set out to test the hypothesis that DTmax and Uopp could be used independently or together to also predict sea-breeze onset time. Hence, for 21 sea-breeze events observed during the summer of 1999, the following data were collected.
DTmax was calculated using surface data from Buoy 44007 and Fryeberg, ME. The
DTmax value was determined by computing the temperature difference between the
two stations at each hour and then selecting the maximum value.
The second predictor, Uopp, was a bit more difficult to accurately determine. 12 UTC sounding data from Gray, ME, averaged over the lowest 500 m, were used to compute Uopp. The value of 500 m was chosen to approximate the sea-breeze depth and is consistent with values previously documented in the literature. These values of Uopp, however, were not well correlated with the sea-breeze onset time. This is likely due to the large-scale flow changing with time. Indeed, Uopp may have changed by 14-16 UTC when sea-breeze onset is observed. This should not be surprising since the synoptic-scale forcing is active during the spring, summer and fall months over New England.
To circumvent this problem, Uopp was calculated with Velocity Azimuth Display
(VAD) derived winds from the clear-air data collected by the WSR-88D radar at
Gray, ME (KGYX). The VADs were performed at the same time that sea-breeze onset
time was observed. Again, the VAD winds were averaged over a depth of 500m when
Finally, sea-breeze onset time needed to be accurately determined. Due to the complexity of the southern Maine, coast, sea-breeze onset time can be different for a given day, depending upon your location. Our definition of sea-breeze onset time is the time that the sea breeze was observed to pass the Portland ASOS station located at the airport.
All available data were used to identify the sea breeze at the Portland surface site. These include looking for changes in temperature, moisture and winds in the Portland ASOS data and also making use of the KGYX clear-air data. It is well known that the sea-breeze front is often observed as a thin line in clear air radar data. The thin lines tend to be more pronounced when the flow is offshore, or opposing the inland penetration of the sea breeze. However, many of the sea-breeze events that occur along the southern Maine coast do so when the ambient flow is either parallel to the coast or onshore. During these days, identifying the sea breeze was difficult since a clearly identifiable signature was absent from the surface data and no radar-detected thin line was present.
To circumvent this problem, a rather unique analysis technique was developed
to detect sea breeze onset time. The technique uses the clear-air radial velocity
data within a circle with a 5 km radius, centered on the Portland ASOS site.
The clear-air velocities within this circle were then averaged for each 0.5
degree surveillance scan collect by the KGYX radar. A time series of these values
were then plotted and often showed clearly sea breeze passage.
Using the above data, a good correlation between sea-breeze onset time and Uopp (r = 0.853) and a weaker correlation with DTmax (r=-0.445) was observed. Hence, it appears from these preliminary analyses that sea breeze onset time can be forecast to within acceptable uncertainty with a good estimate of Uopp. Hence, our initial hypothesis appears to be valid. Future efforts will expand the number of cases included in our analysis and testing the technique at the Gray, ME NWS office. This testing will occur during the summer of 2002.
The data used for this study were collected by the NWS personnel. The above analyses and analyses techniques were created and performed by the LSC personnel. Testing of the technique will be the responsibility of the NWS participants.
SECTION 2: SUMMARY OF UNIVERSITY/NWS EXCHANGES
The LSC and NWS-Gray personnel met on a number of occasions to discuss the propose research and results to date. A list of these meetings follows:
October 99 - Meeting at NWS, Gray ME to discuss possible sea-breeze project
October 00 - Meeting at LSC to discuss results to date.
March 01 - Meeting in Saratoga Springs, NY at 26th Northeast Storm Conference to discuss future data analysis.
SECTION 3: PRESENTATIONS AND PUBLICATIONS:
Preliminary research results have been presented at the following meetings:
Vibert, S., N. Atkins, J. Hayes, and F. Ronco, 2001. Forecasting the Sea Breeze over Southern Maine. Presented at 26th Northeast Storms Conference. Saratoga Springs, NY.
Atkins, N., S. Vibert, J. Hayes, and F. Ronco, 2001. Forecasting the Sea-Breeze Onset Time over Southern Maine. Presented at 8th Weather at the Summit Conf., Sugarbush VT.
Hayes, J., F. Ronco, N. Atkins and S. Vibert, 2001. Forecasting the Sea Breeze in Southern Maine. Presented at the GYX Spring Weather Workshop, Gray ME.
SECTION 4. SUMMARY OF BENEFITS AND PROBLEMS ENCOUNTERED.
4.1 - ACADEMIC PARTNER
BENEFITS - There are numerous benefits of this research to the LSC personnel. First, a new and unique forecast tool is being developed. Eventually, the new knowledge that results from this research will be published in a refereed journal. Second, the research funded an LSC Meteorology student for one summer. The student was exposed to the research process and working with others towards a common goal. The student presented results of this study at the 26th Northeast Storm Conference in Saratoga Springs, NY. This student has since graduated and is now working for Raytheon, Inc. Third, the results of this research will be incorporated into Mesoscale Meteorology course material taught by N. Atkins. Finally, the project has enhanced the research infrastructure on the LSC campus.
PROBLEMS ENCOUNTERED - We encountered one major problem. After analyzing the 1999 sea breeze cases, it was clear to us that more cases needed to be analyzed to bolster the statistical significance of our conclusions. Thus, the NWS personnel identified a number of good sea-breeze cases from 2000. The data was ordered from NCDC and sent to LSC for analysis. Unfortunately, we could not read the data. After three months of trouble shooting, we learned that NCDC changed the format of the Level II data. It then took another two months to have the appropriate software modified such that it would read the data. As the LSC student had graduated by the time this issue was resolved, the year 2000 cases await analysis. It should be noted that NCDC made no note of the format change on their web site or in the documentation they send with the requested Level II data.
4.2 - FORECAST PARTNER
BENEFITS - There were two specific benefits of the research to the forecast office in Gray, ME. First, a new and useful forecast tool is being developed. Anecdotal forecast techniques for the arrival time of the sea breeze had been used in the past, but this research will provide a much better method for determining onset time.
Second, the research established a link to the academic community that did not exist before. Hopefully, this will lead to additional collaborative research with LSC and other members of the nearby academic community.
PROBLEMS ENCOUNTERED - There was a significant problem obtaining readable Level II radar data for LSC from NCDC. First, there was a delay in receiving the data after it was ordered (this occurred on a number of occasions; the last such delay was attributed to mechanical problems with the tape readers at NCDC). Second, once the data was received, it was discovered that the data could not be read by either LSC or the forecast office in Gray. It was eventually discovered that the format of the Level II data had changed, without any notification from NCDC. This resulted in a significant delay in processing the data at LSC.
There were also some local archive problems, due mainly to the sporadic interruption of the satellite feed from Eastern Region Headquarters (ERH). This resulted in some potential dates being excluded from the study, mainly because of the prohibitive replacement cost of the missing satellite data.