Introduction

The NCAR ASTER (Atmospheric Surface Turbulent Exchange Research) field facility was operated during the STORMFEST (STORM - Fronts Experiment Systems Test) field program, February 1st to March 15th, 1992, at a site 2 miles west of Sabetha, Kansas, in the northeast quadrant of the 50 km square boundary layer array. The principal purpose for deployment of the ASTER facility was to support the boundary layer measurement program by providing micro meteorological measurements for the turbulent fluxes of momentum, sensible heat, and water vapor. The principal investigators were Robert Grossman from the University of Colorado, and Margaret LeMone from NCAR.

In addition to direct flux measurements, ASTER also provided near surface profiles of wind, temperature, and humidity to a height of 10 m and radiation and soil measurements to complete the surface thermal energy budget. Ancillary measurements were made to inter-compare the performance of three commercial sonic anemometers during winter conditions and to evaluate a solid state capacitance humidity sensor being used for the first time on both ASTER and PAM (Portable Automated Mesonet) surface meteorological stations during STORMFEST.

Generic information about the ASTER field facility may be found in Businger, et al. (1990), Semmer and Martin (1991), or Semmer, et al. (1992).

Site

The measurement site was located 2 miles west of Sabetha, Kansas, in Section 3 of R. 14E, T. 2S in Nemaha County, at Longitude 95°50'35'' E, Latitude 39°54'25'' N, altitude 410 m. The topography in the immediate vicinity of the measurement site was gently rolling farmland with a maximum local relief of approximately 70'/mile from west to east and 40'/mile from north to south (see the USGS 7.5 min Sabetha, Kansas, quadrangle for more detail). The surrounding countryside consisted of plowed fields characterized by bare dirt furrows and short stubble, although the ASTER tower array itself was located along an east fence line in the center of an 8.5 m wide strip of grass. The nearest structure was a two-story metal shed about 135 m west of the tower array, and the ASTER computer and laboratory trailers were parked on the north side of this building. The next closest structures were farm buildings located about 0.5 miles north of the array and the nearest trees were also at a distance of 0.5 miles.

Array

The main ASTER instrument array was mounted on six towers placed in a line from west to east along the fence line between the northern and southern halves of Section 3, with the western most tower about 155 m east of the western boundary of Section 3. The towers, from west to east, were a 10 m temperature/humidity profile tower, a 10 m wind profile tower, three 5 m sonic anemometer inter comparison towers, and a 10 m turbulent flux tower. The instrument booms on the temperature/humidity, wind, and flux towers pointed to the east and those on the sonic anemometer inter comparison towers pointed to the north. The radiation and soil heat flux sensors were located on the fallow cropland southwest of the temperature/humidity profile tower. The dimensions of the tower array are shown in the postscript file, "figure1.ps."

The ASTER towers have a triangular cross section 15 cm on a side. The ASTER Data Acquisition Modules (ADAMs) (marigold, cosmos, and ragwort) have dimensions of 50 cm (high) x 85 cm x 85 cm and were located immediately forth of the tower line as shown in "figure1.ps." The ASTER power transformer was located west of the array and has dimensions of 85 cm (high) x 60 cm x 85 cm; the adjacent shelter has dimensions of 110 cm (high) x 75 cm x 95 cm. The radiation sensors were mounted on a horizontal beam at a nominal height of 2 m. There were three groups of soil sensors, each measuring soil temperature at 4 depths (1, 3, 5, 7 cm) and soil heat flux at 8 cm, which were placed at three locations corresponding to the bottom, side, and top of a furrow (denoted in the data archive as fur, int, rid).

Sensors

The sensors used by ASTER in STORMFEST are shown schematically in "figure2.ps" and listed in Table 1. The variable names associated with the sensors in "figure2.ps" are the mnemonics or data id's (did) used to identify the archived data variables, e.g. "u.prop.4m" and "v.prop.4m." Each sensor in "figure2.ps" is associated with one or more system id's (sid) that are composed of an ADAM name and a channel number, e.g. cosmos 202. Channel numbers from 100 to 199 indicate an analog input channel that is usually associated with a single variable or did, and channel numbers from 200 to 299 indicate a serial input channel that is usually associated with a single sensor but may ingest multiple variables.

The propeller vane wind components are in a geographic coordinate system, where u is a wind from the west and v is a wind from the south, while the sonic anemometer wind components are in an instrument oriented coordinate system, where u is a wind along the instrument boom toward the tower and v is a wind from right to left as you face in the positive u direction. The sonic anemometer boom orientations are listed in Table 2 for conversion of the sonic wind components to geographic coordinates.

The sonic anemometers also measure the speed of sound, from which they calculate sonic (approximately virtual) temperature (tc) in °C. However the measurement of the speed of sound by the sonics is contaminated by the horizontal wind speed and only the ATI sonic temperature data is corrected (within the instrument) for this contamination.

The pyrgeometers also output some and case temperatures in Kelvin.

Table 1: STORMFEST Sensors and Variables

Table 2: Sonic Anemometer Boom Orientations

Diagrams of the layout and sensors of the array are available.

Operations

ASTER collected data continuously from February 1st to March 15th, 1992. Detailed notes on daily ASTER field operations as well as notes on aspects of the data reduction that are unique to STORMFEST are available in an electronic logbook at NCAR.

The major operational problem was caused by the sonic anemometers used for the turbulent flux measurements at 4 and 10 m. Two new Applied Technology (ATI) sonic anemometers were purchased for STORMFEST, but did not arrive until February 8th. Prior to that date, a similar sonic anemometer was borrowed from the manufacturer and mounted at 4 m on the flux tower (After the arrival of the new sonic anemometers, the borrowed sonic was moved to the inter comparision sub-array). Because of a faulty connector in the new anemometers, their data were quite noisy during the experiment. The problem was reduced, but not eliminated, by the manufacturer around February 15th (ATI replaced the connectors following the field program). The sonic anemometer data can be optionally filtered by software that compares each data point to a value predicted from the preceding data, notes those whose difference from the forecasted value exceeds a specific limit, and replaces them with the forecasted value. The placement is noted by setting a flag variable (e.g. uflag.ati.10m) to 1. This data quality software has been used during the processing of the data for the STORMFEST CDROM.

For various reasons, two of the inter comparison sonic anemometers were available for only a portion of the total operational period. The Gill or Solent sonic anemometer (loaned by the manufacturer) was operated for the entire period. The ATI inter comparison sonic (loaned by the manufacturer) was operated only after February 22nd and the Kaijo Denki (loaned by Prof. Shashi Verma of the University of Nebraska) was operated from February 4th to March 11th.

In order to evaluate a new solid state capacitance humidity sensor used on both ASTER and on the PAM (Portable Automated Mesonet) surface meteorological stations during STORMFEST, humidity measurement inter comparisons were made at the 1 m level on the temperature/humidity tower. These involved, at various times, an EG&G dew point sensor, a temperature/humidity sensor identical to those used at the other ASTER tower levels and on the PAM stations, and temperature/humidity sensors used as traveling inter comparison "standards" during service visits to the PAM stations. The latter two types of sensors were recorded as "(tdry,rh).psyc.1am."

Data

Daily plots have been made of a variety of parameters to provide an overview of the meteorological conditions for that day and to aid in the selection of time periods for detailed analysis. These are available on the CDROM as postscript files with the names "aster_wx.jjj.ps," where jjj is the Julian day. Each figure consists of four panels that contain (a) temperature (4 m), humidity (4 m), pressure (2 m), and rainfall; (b) wind speed and direction (10 m); (c) net radiation, sensible heat flux, latent flux, and the surface soil heat flux; and (d) z/L (10 m) and u_*. The  plotted data are 5 minute averages, with the exception of turbulent flux data (sensible and latent heat fluxes, z/L and u_*) which are 20 minute averages calculated from the ATI 4 m flux data. The Monin Obukhov stability parameter z/L has been evaluated with z set to 10 m; a line for neutral conditions (z/L = 0) is shown for reference.

The horizontal line at 90° on the wind direction plot denotes the east west orientation of the line of towers. Recall that the instrument booms on the temperature/humidity, wind, and flux towers pointed to the east and those on the sonic anemometer inter comparison towers pointed to the north. Wind and flux data collected with anemometers downwind of a tower should be used with caution. Thus the flux data is suspect for winds from the west, a direction already contaminated by the nearby metal shed and the ASTER trailers.

Soil samples were collected daily of every other day, and soil moisture was measured by a gravimetric technique. These data are available in the electronic logbook and are plotted in "figure3.ps," along with the rainfall rates and albedo. The percent soil moisture is defined as the difference in weight of the soil sample before and after drying, divided by the dry soil weight and multiplied by 100.

Means, variances, and covariances of the full time series data have been produced at five minute resolution and stored in Network Common Data Form (NetCDF) files. These NetCDF files contain the data in binary form, as well as supporting information about the data variables. These files are named "stormfest.jjj.cdf," where jjj is the three digit Julian day, numbered from one on January 1st; e.g. stormfest.033.cdf contains data for February 2nd, 1992.

Time (GMT) is represented in seconds since the beginning of the day. The start of day is stored in a "base_time" variable as the number of seconds since January 1st, 1970 00:00 GMT.

Each variable in the NetCDF file corresponds to a 5 minute statistic of the time series data. Associated with each variable is a NetCDF attribute indicating the ASTER dids, as show in "figure2.ps," which were the source of the value. The variable names correspond closely to the ASTER dids but, due to syntax restrictions, are not identical. Also associated with each NetCDF variable are a units field and the name of the NetCDF variable that contains the number of samples incorporated into the statistic.

The ZEB display software available on the STORMFEST CDROM provides a menu driven, interactive graphical display of the ASTER NetCDF files.

NetCDF files are not ASCII character files, and therefore special software is needed to access these files. NetCDF software is licensed by the University Corporation for Atmospheric Research and is available free to the university community. Call the UCAR Finance and Administration Office, (303) 397-8571, to get a copy of the license agreement. The software is available via anonymous FTP from unidata.ucar.edu.

The un-averaged time series data for each did are stored at NCAR. Contact Tom Horst or Gordon Maclean to request access to the full time series data or for more information about the five-minute averaged data provided on the CDROM.

Field Logbook

An electronic logbook was maintained during the field program. A total of 229 entries were made during the STORMFEST deployment in 16 different categories.

Data Download

Click here to download data.

Acknowledgments

ASTER STORMFEST field operations were supported by Jorgen Hojstrup and Niels Mortensen, RISO National Laboratory, and Tony Delany, Charlie Martin and Allan Abshire, NCAR. We also thank Applied Technologies Incorporated, Gill Instruments, and Prof. Shashi Verma for loans of their sonic anemometers for the inter comparison sub-array.

References

• Businger, J.A., W.F. Dabberdt, A.C. Delany, T.W. Horst, C.L. Martin, S.P. Oncley and S.R. Semmer. 1990. The NCAR Atmospheric Surface Turbulent Exchange Research Facility. Bulletin of the American Meteorological Society, 71, 1006-1011.
• Semmer, S.R., and C.L. Martin, 1991. NCAR Atmospheric Surface Turbulent Exchange Research Facility. Proceedings of the Seventh Symposium on Meteorological Observations and Instrumentation, January 13th, 18th, 1991. New Orleans, LA. American Meteorological Society, Boston, MA.
• Semmer, S.R., S.P. Oncley, C.L. Martin, T.W. Horst, A.C. Delany, W.F. Dabberdt, and J.A. Businger. 1992. A new research tool for surface exchange investigations; The NCAR Atmospheric Surface Turbulent Exchange Research (ASTER) Facility. Precipitation Scavenging and Atmosphere-Surface Exchange, Vol. 2, S.E. Schwartz and W.G.N. Slinn, Coords., Hemisphere Publishing, 619 626.