The impact of land and sea surface variations on the Delaware sea breeze at local scales
Date
2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The summertime climate of coastal Delaware is greatly influenced by the intensity, frequency, and location of the local sea breeze circulation. Sea breeze induced changes in temperature, humidity, wind speed, and precipitation influence many aspects of Delaware’s economy by affecting tourism, farming, air pollution density, energy usage, and the strength, and persistence of Delaware’s wind resource. The sea breeze front can develop offshore or along the coastline and often creates a near surface thermal gradient in excess of 5°C. The purpose of this dissertation is to investigate the dynamics of the Delaware sea breeze with a focus on the immediate coastline using observed and modeled components, both at high resolutions (~200m).
The Weather Research and Forecasting model (version 3.5) was employed over southern Delaware with 5 domains (4 levels of nesting), with resolutions ranging from 18km to 222m, for June 2013 to investigate the sensitivity of the sea breeze to land and sea surface variations. The land surface was modified in the model to improve the resolution, which led to the addition of land surface along the coastline and accounted for recent urban development. Nine-day composites of satellite sea surface temperatures were ingested into the model and an in-house SST forcing dataset was developed to account for spatial SST variation within the inland bays. Simulations, which include the modified land surface, introduce a distinct secondary atmospheric circulation across the coastline of Rehoboth Bay when synoptic offshore wind flow is weak. Model runs using high spatial- and temporal-resolution satellite sea surface temperatures over the ocean indicate that the sea breeze landfall time is sensitive to the SST when the circulation develops offshore.
During the summer of 2013 a field campaign was conducted in the coastal locations of Rehoboth Beach, DE and Cape Henlopen, DE. At each location, a series of eleven small, autonomous thermo-sensors (i-buttons) were placed along 1-km transects oriented perpendicular to the coastline where each sensor recorded temperatures at five-minute intervals. This novel approach allows for detailed characterization of the sea breeze front development over the immediate coastline not seen in previous studies. These observations provide evidence of significant variability in frontal propagation (advancing, stalling, and retrograding) within the first kilometer of the coast. Results from this observational study indicate that the land surface has the largest effect on the frontal location when the synoptic winds have a strong offshore component, which forces the sea breeze front to move slowly through the region. When this happens, the frequency of occurrence and sea breeze frontal speed decreases consistently across the first 500 m of Rehoboth Beach, after which, the differences become insignificant. At Cape Henlopen the decrease in intensity across the transect is much less evident and the reduction in frequency does not occur until after the front is 500 m from the coast. Under these conditions at Rehoboth Beach, the near surface air behind the front warms due to the land surface which, along with the large surface friction component of the urbanized land surface, causes the front to slow as it traverses the region.
Observation and modeling results suggest that the influence of variations in the land and sea surface on the sea breeze circulation is complex and highly dependent on the regional synoptic wind regime. This result inspired the development of a sea breeze prediction algorithm using a generalized linear regression model which, incorporated real-time synoptic conditions to forecast the likelihood of a sea breeze front passing through a coastal station. The forecast skill increases through the morning hours after sunrise. The inland synoptic wind direction is the most influential variable utilized by the algorithm. Such a model could be enhanced to forecast local temperature with coonfidence, which could be useful in an economic or energy usage model.