**Overview** This dataset includes large-eddy simulation (LES) output from a neutrally stratified atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on Aug. 17, 2012. The dataset was used to assess LES models for simulation of canonical neutral ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is the model output, and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** This dataset includes large-eddy simulation (LES) output from a convective atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on July 4, 2012. The dataset was used to assess the LES models for simulation of canonical convective ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is model output and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** This dataset includes large-eddy simulation (LES) output from a neutrally stratified atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on Aug. 17, 2012. The dataset was used to assess LES models for simulation of canonical neutral ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is the model output, and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** This dataset includes large-eddy simulation (LES) output from a neutrally stratified atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on Aug. 17, 2012. The dataset was used to assess LES models for simulation of canonical neutral ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is the model output, and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** This dataset includes large-eddy simulation (LES) output from a convective atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on July 4, 2012. The dataset was used to assess the LES models for simulation of canonical convective ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is model output and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** This dataset includes large-eddy simulation (LES) output from a convective atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on July 4, 2012. The dataset was used to assess the LES models for simulation of canonical convective ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is model output and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** This dataset includes large-eddy simulation (LES) output from a convective atmospheric boundary layer (ABL) simulation of observations at the SWIFT tower near Lubbock, Texas on July 4, 2012. The dataset was used to assess the LES models for simulation of canonical convective ABL. The dataset can be used for comparison with other LES and computational fluid dynamics model outputs. **Data Quality** This is model output and no specific data quality procedures were used. **Uncertainty** This is output from a deterministic model. **Constraints** There are no specific constraints.

**Overview** Scaled Wind Farm Technology (SWiFT) Facility meteorological tower (MET), turbine, and Technical University of Denmark (DTU) SpinnerLidar data acquired on 20161216 UTC during a neutral atmospheric boundary layer inflow at a single focus distance of 2.5 D (D=27 m). **Data Quality** Data information is provided in uploaded documentation: *SWiFT Wake Steering Instrumentation and Data Processing* (PDF).

**Overview** In support of the Wind Forecasting Improvement Project, Pacific Northwest National Laboratory (PNNL) deployed surface meteorological stations in Oregon. **Data Details** A PNNL computer is used as the base station to download the meteorological data acquired by the data logger at each site via a cellular modem. The data collected will be made available to the National Oceanic and Atmospheric Administration (NOAA) each hour and used to support the short-term forecasting project by providing an independent evaluation of the added value of new data to meteorological forecasts. Each meteorological station consists of a solar-powered data acquisition system and wind speed, wind direction, temperature, humidity, barometric pressure, and solar radiation sensors on a 3-meter tower. Specifically, the stations comprise the following instruments and equipment: * Campbell Scientific CM10 Tripod * Campbell Scientific CR10X Measurement and Control System * R.M. Young 05106 Wind Monitor * Vaisala HMP45C Temperature and Humidity Probe * Vaisala PTB101B Barometric Pressure Sensor * Li-Cor LI200X Pyranometer The data logger is used to sample, at 1-second intervals, the horizontal wind speed and direction at 3 meters above ground level (AGL); the air temperature, relative humidity, barometric pressure, and solar radiation at 2 meters AGL; and the logger temperature and power supply. The logger outputs the 1-minute averages of these measurements to final storage and power on the cellular modem, so the data can be retrieved and downloaded to a base-station computer. The data are archived as 1-hour comma-delimited ASCII files (see "Table 2. Format of the WFIP2 Comma-delimited ASCII Data Files" in **wfip2-met-data.pdf**). All dates and times in the file names and data records are in UTC (Coordinated Universal Time) and denote the end of the 1-minute average. Data known to be of bad quality have been replaced with "-9999." **Data Quality** Data for each primary measurement at every site are automatically plotted daily and reviewed about every three days. Instrument outages or events are reported with the Instrument and Model Data Problem Log at: . Data known to be of bad quality have been replaced with "-9999." **Constraints** None

**Overview** The dataset includes 15-minute average wind speed and direction records from 30 m to 330 m above ground level (AGL) in 10-m range gates. Data were collected by a Scintec MFAS wind profiler installed at the Decker Ranch in Oregon, about 4.4 km southeast of Kent, Ore., and are intended for validating WFIP2 model improvements. **Data Details** Instrument location: * N 45°09'54.42" (N 45.165117) * W120°39'20.87" (W 120.655799) Instrument clock and computer system time set to UTC. **Data Quality** The Scintec MFAS wind profiler instrument installed at the Decker Ranch is capable of measuring at heights up to 1000 m. For this study, the maximum height was set to 330 m. The instrument was oriented to true north, so no corrections to the wind direction should be made. Scintec wind profilers come with the APRun software package, which performs data collection and quality control (QC), among other functions. Version 1.46 of APRun was used in this study. The APRun manual states: *The primary results are checked against local signal quality criteria, combined signal quality criteria and two-dimensional spatial/temporal consistency tests. Any data that does not pass all quality control tests is devalidated and removed*. Devalidation means replacing the value with an error value, usually a series of ‘9’s, such as 99.99 or 999.99. Not all devalidated data are actually removed from the *.mnd files, so the user must filter them out. There are some error flags that indicate the type of error, but these are not included in the *.mnd files, and we have no access to them. Because QC already has been performed by APRun, our QC procedures consisted of removing samples with error values and performing a visual inspection of the data to see if larger patterns indicated any kind of problem. There are 623 gaps of two hours or less and 61 gaps of more than two hours. The longest gap is 15.31 days, from 2016-12-07 03:00Z to 2016-12-22 10:30Z. All gaps that exceed two hours are listed in file: Decker_Ranch_gaps.txt.