Collins, Clarence; Hathaway, Kent K; Birkemeier, William A; Forte, Michael F; Dickhudt, Patrick J; The FRF Data Team (2019): Wind Speed and Direction at USACE Field Research Facility 1980-2018 [dataset]. U.S. Army Corps of Engineers, PANGAEA, https://doi.org/10.1594/PANGAEA.902189,

---

In: Collins, C et al. (2019): Nearshore Topographic and Bathymetric Surveys, Wave, and Wind Records from the USACE Field Research Facility in Duck NC from 1980 to 2018 [dataset publication series]. U.S. Army Corps of Engineers, PANGAEA, https://doi.org/10.1594/PANGAEA.902954

Over time, wind data has been recorded at 10+ distinct wind stations often with multiple stations active simultaneously. The highest quality records were selected to be included into a conglomerate. From 1981-1987, station locations included a tower near the main building and the roof of the main building. In 1987, measurements were additionally made at the end of the pier and, in 1994, the pier became their exclusive location.

At each station, wind speed and direction were measured by either Weather Measure Skyvane anemometers or RM Young (RMY) marine anemometers. Both were mechanical anemometers whereby the action of the wind turns an impeller.

In the early era, there was a gauge preference based on location (e.g., the pier location was preferred over the roof). In the modern era of multiple wind stations at the end of the pier, an algorithm chose a preferred station. Wind stations, and their preference, are listed in the Scientific Data paper. Most recently, station 1 or 2 is given preference depending on wind direction (the upwind sensor is preferred). If data from the preferred station passes QC, it is included in the conglomerate. If not, the same process is applied down the chain of preference until a record passes (stations 5 and 8 were never used).

Wind speed and direction were recorded a 1 or 2 Hz with various record lengths. There are 3 main eras of sampling. Sampling schemes varied until near continuous sampling began in 1992. All previous sampling schemes were reanalyzed into 10 minute segments for calculating statistics. Record remainders are kept if they are at least 6 minutes but forced to 10 minute time steps.

Data is in a netCDF file with variables (units): time (seconds), latitude (deg. N), longitude (deg. W), sensorID, sensorElevation (m), windSpeed (m/s), windGust (m/s), windDirection (deg.), qcFlagSpeed, qcFlagDirection. time is unix epoch time, sensorID is digit code representing unique ID and sensor preference, sensorElevation is relative to NAVD88, windSpeed is 10 minute average, windGust is highest 5 second average, windDirection is coming from convention - clockwise from true North, qcFlagSpeed & qcFlagDirection are indications of data quality - 1 = passed, 3 = questionable, and 4 = fail.

Please see the paper for full documentation.