The Atmospheric Radiation Measurement (ARM) program is an initiative by the U.S. Department of Energy to study the effects of clouds on global climate change. Early effort for this program is focused on establishing several Clouds and Radiation Testbed (CART) sites which will make detailed measurements of four-dimensional fields of physical parameters over a 100 by 100 km area. A four-dimensional data-assimilation model probably will be used to interpolate over the available measurements to obtain evenly-gridded fields.

The National Center for Atmospheric Research (NCAR) is supporting the ARM program by developing the  data-assimilation model and by developing a ground-based Integrated Sounding System (ISS) which will provide continuous wind, temperature, and humidity profiles for input to the model. Prototypes of these developments were tested during the Winter Icing and Storms Program (WISP) in February and March of 1991 since WISP already had extensive upper-air measurements (wind profilers, integrated water radiometers, balloon sounding systems, and aircraft).

The NCAR Atmosphere/Surface Turbulent Exchange Research (ASTER) facility was deployed for ARM during WISP to provide surface fluxes of sensible and latent heat. It is possible that these fluxes may be necessary for the accurate retrieval of humidity profiles from ground-based radiometers. ASTER was located at a site selected to be  representative of the fetch observed by the lowest profiler levels. The profiler site, near Platteville Colorado (see Road Map), had been deliberately located in a "bowl" (see Topographic Map) to reduce ground clutter for the first radar systems which were used at this site.

The ASTER masts were placed about 700 m NNE of the profiler where the topography and surface coverage was more uniform. The surface coverage within 2 km was mostly wild grasses with clumps up to 1 m high (see Photos). A wire fence and a line of power poles ran East/West 150 m North of the masts. Oil and natural gas wells and tanks also were scattered around the region. The nearest was about 600 m to the NNE, the next about 800 m to the NNW, and the last one visible about 800 m to the WSW. It was felt that none of these structures would have a measurable effect on the flux measurements.

The ASTER trailers were placed 300 m ENE of the masts, in line with "backbone" of the masts (see ARM Layout). Although the trailers also were in a "bowl," winds from this direction should be rejected due to possible contamination of the fluxes. Good wind directions were considered to be West, through North, to Northeast. This was selected since data from the Boulder Atmospheric Observatory in Erie, Colorado indicated that North and West winds occured somewhat more frequently than others during February and March.

The RISO Wind Atlas Program (WAsP) was run on the topography for this site for heights of 2 and 10 m assuming a surface roughness of 3 cm (see Wind Atlas). It predicts that the horizontal wind would be decreased by 3% for NE winds and increased by 3% for WNW winds at 10 m in relation to what the wind field would be over horizontally-uniform terrain. This magnitude increases slightly for the lower measurements. The variance of horizontal velocity might be expected to have been measured to ±6%, however extrapolation to other second-order moments (such as momentum flux) is questionable since the model contains no vertical velocity information. Nevertheless, it may be concluded that the 10 m flux measurement is more representative than the 4 m measurement of the large-scale flux, though in fact little difference is seen.

The masts were arranged in a nearly "generic" configuration (see physical tower layout). Three 10 m towers were used for a scalar (temperature and humidity from psychrometers) profile, a vector (wind speed and direction from propeller-vane anemometers) profile, and fluxes (three-dimensional wind vector from the uw sonic anemometers plus temperature and humidity fluctuations from a platinum wire and Krypton hygrometer) at two heights (4 and 10 m). A "sawhorse" held upward and downward-looking radiometers for visible and infrared radiation along with a net radiometer. Soil temperature sensors were buried at 3 and 10 cm, soil heat flux plates were buried at 3, 10, and 20 cm, and integrated soil moisture for the upper 10 cm was manually measured gravimetrically most days at 1200 MDT (local noon) (see Soil Moisture summary).

In addition, the performance of three other sonic anemometers was tested during the second half of the experiment. An NCAR-modified Applied Technologies, Inc. (ATI) BAO-style array, a Gill (now called Solent) asymmetric head, and a new ATI K-style sonic anemometer were each placed at a height of 4 m (see sonic intercomparison layout).

In addition to the above sensors, a pitch and roll electrolytic level was used to determine the vertical orientation of the uw sonic anemometers. Since only one readout unit was available, it was switched every day (while the soil sample was being taken) between the sensors at 4 m and 10 m. Different channels were assigned (see Channel Configuration) for these two heights and were left floating when no signal was attached.

Other sensors were; a barometer with three independent solid state pressure sensors used to measure total pressure at 2 m, two Licor radiometers, and vaisala and rotronics solid state humidity sensors (see Sonic Intercomparison Channel Configuration). A complete list of the standard ASTER sensors and these additional sensors is given in Table 1.

Precipation during ARM (from the nearest PAM station, #22) is shown on the bottom panel of "Summary of Conditions." Several precipitation events are seen during the period. The upper panel shows the albedo measured by the ASTER radiation sensors. It is seen that the precipation on days 54, 64, and 65 probably is snow. The snow had only a modest effect on the albedo since the ground cover was faily tall and was never completely covered by snow.

Relative humidity from a Vaisala solid-state humidity sensor which was tested during the last half of the experiment is shown (see Humidity plot) to provide mean humidity information since the psychrometers often were frozen (see below). Good mean humidity values are not available during the first half of the experiment due to unexpected drift of the Krypton hygrometers.

The last section of this report contains daily plots of basic variables. The top panel has the mean temperature and humidity measured by the psychrometer along with the pressure from the barometer at 2 m. Sharp drops in humidity (for example, at 1800 on day 49) are caused by the wet-bulb thawing out after having been frozen and reading saturated values. The next panel is wind speed and wind direction, usually measured from the propeller-vane anemometer at 10 m. The dotted line represents direction of winds coming perpendicular to the mast array.  Acceptable fetches should be for winds ±55 degrees from of this direction (270 - 020 degrees). The next lower panel shows the net radiation, and sensible, latent, and soil heat flux contributions to the energy balance. The bottom panel shows the Monin-Obukhov stability parameter z/L evalutated at 10 m from measurements from the UW tower along with u_* from these measurements. A line for neutral conditions (z/L = 0) is shown for reference. All variables are plotted as a smoothed fit through 5 minute averages.

Sensors

Table 1: Instruments Used on ASTER

Field Logbook

An electronic logbook was maintained during the field program.

More information about this event…