[an error occurred while processing this directive]


Table of Contents


This document is a standard product of NCAR/ATD/RTF which gives an overview of the measurements taken by ISFF and conditions during the OASIS98 field experiment. This document can be obtained either in hard copy from RTF or in electronic form from the NCAR/ATD WWW site.

Data Access

The NCAR data for OASIS are available in the following forms: Also available is a computer-readable logbook of comments noted by RTF personnel.

A complete list of variables is available. This list shows the NetCDF variable name, dimensions, descriptive variable name, units, and additional comments, if available. Each file contains all the statistics for one day, beginning at 00:00:00 UTC. Most variables are dimensioned (time, station), where time is in 5 minute increments (288 per day), station is 1 or 2 for variables common to NCAR (station=1) and Oklahoma (station=2). Variables without a station dimension were measured by either NCAR or OK, but not both. See the Sensors description below for guidance as to where the variables were measured.

Most descriptive variable names hopefully are self-explanatory. Those beginning with Capital letters came from slow response sensors and those with lower cases letters from fast response sensors. Higher order statistics are indicated by products of variables. For example, w'tc' is the covariance between vertical velocity and sonic virtual temperature and h2o'h2o'h2o'h2o' is the fourth order moment of water vapor mixing density. For some variables, the sensor name is appended, e.g. "v.nuw1" is the lateral velocity component from the one of the new sonic anemometers we built with a University of Washington-style array, to distinguish it from "v" which is the same variable measured by other sonic anemometers.


The NCAR sensors were deployed on the University of Oklahoma Mesonet site at Norman, OK. The towers were spaced about 7m apart along an east-west line beginning approximately 7m to the west of the prototype OASIS station near the center of the field at this site. From east to west, these towers were: hygrothermometer, prop, sonic, nuw sonic. The nuw sonic tower was only 5m high, all others were 11m. Our radiation stand was located approximately 15m north of the OASIS station.

A series of views forming a panorama from the 9m level of the sonic tower is available here:

  1. North
  2. Northeast
  3. East
  4. Southeast
  5. South
  6. Southwest
  7. West
  8. Northwest



Most of this section will be completed later.

The data files contain data from NCAR and the University of Oklahoma Mesonet. The NCAR sensors are mostly described in the ASTER Facility Description. A color-coded diagram of the variables from these sensors was created to help sort out the variables.

Table of Variables

[an error occurred while processing this directive]

Variable Name Description Units Dimension Sensor Notes
base_timeSeconds since
1970 Jan 1 00:00 GMT
timeSeconds since base_timesecondstime  
Gsoil.5cm.1Soil Heat FluxW/m^2time x station  
Gsoil.5cm.2Soil Heat FluxW/m^2time x station  
Gsoil.5cm.3Soil Heat FluxW/m^2time  
Mp.5cm.1 msectime  
Mp.5cm.2 msectime  
Mp.5cm.3 msectime  
Msoil.5cm.1Soil MoistureV frctntime x station  
Msoil.5cm.2Soil MoistureV frctntime  
Msoil.5cm.3Soil MoistureV frctntime  
PBarometric Pressurembtime x station  
RH.0.5mRel. Humidity%time  
RH.1.5m.2Rel. Humidity%time  
RH.1.5m.3Rel. Humidity%time  
RH.1.5mRel. Humidity%time x station  
RH.4.5mRel. Humidity%time  
RH.6.5mRel. Humidity%time  
RH.9mRel. Humidity%time x station  
Rlw.in.CNR1aIncoming LongwaveW/m^2time  
Rlw.out.CNR1aOutgoing LongwaveW/m^2time  
Rnet.CNR1aNet RadiationW/m^2time  
Rnet.NRLite.3Net RadiationW/m^2time  
RnetNet RadiationW/m^2time x station  
Rpile.in.CNR1aUp-looking ThermopileW/m^2time  
Rpile.inUp-looking ThermopileW/m^2time x station  
Rpile.out.CNR1aDown-looking ThermopileW/m^2time  
Rpile.outDown-looking ThermopileW/m^2time x station  
Rsw.in.CNR1aIncoming ShortwaveW/m^2time  
Rsw.inIncoming ShortwaveW/m^2time x station  
Rsw.out.CNR1aOutgoing ShortwaveW/m^2time  
Rsw.outOutgoing ShortwaveW/m^2time x station  
Spd.2mWind Speedm/stime x station  
Spd.9mWind Speedm/stime x station  
T.0.5mAir TempdegCtime  
T.1.5m.2Air TempdegCtime  
T.1.5m.3Air TempdegCtime  
T.1.5mAir TempdegCtime x station  
T.4.5mAir TempdegCtime  
T.6.5mAir TempdegCtime  
T.9mAir TempdegCtime x station  
Tadam.marigoldElectronics TempdegCtime  
Tadam.ragwortElectronics TempdegCtime  
Tcase.CNR1a W/m^2time  
Tcase.inUp-looking Case TempdegCtime x station  
Tcase.outDown-looking Case TempdegCtime x station  
Tdome.inUp-looking Dome TempdegCtime x station  
Tdome.outDown-looking Dome TempdegCtime x station  
TsfcSurface TempCtime  
Tsoil.5cm.1Soil TemperaturedegCtime x station  
Tsoil.5cm.2Soil TemperaturedegCtime x station  
Tsoil.5cm.3Soil TemperaturedegCtime  
U.10mEastward Windm/stime  
U.1mEastward Windm/stime  
U.2mEastward Windm/stime  
U.4.5mEastward Windm/stime  
U.6.5mEastward Windm/stime  
U.9mEastward Windm/stime  
V.10mNorthward Windm/stime  
V.1mNorthward Windm/stime  
V.2mNorthward Windm/stime  
V.4.5mNorthward Windm/stime  
V.6.5mNorthward Windm/stime  
V.9mNorthward Windm/stime  
cflag.nuw1  time  
cflag.nuw2  time  
diag.csat  time  
h2o'h2o'.4.5mh2o variance(g/m^3)^2time x station  
h2o'h2o'.9mh2o variance(g/m^3)^2time  
h2o.4.5mWater Vapor Densityg/m^3time x station  
h2o.9mWater Vapor Densityg/m^3time  
lev.rad.x degtime  
lev.rad.y degtime  
lev.u.4.5m degtime  
lev.u.9m degtime  
lev.v.4.5m degtime  
lev.v.9m degtime  
mr'mr'.bph.4.5m (gm/kg)^2time  
mr.bph.4.5m gm/kgtime  
rainaRain Accumulationmmtime x station  
rainrRain Ratemm/hrtime x station  
rh.bph.4.5m %time  
t't'.4.5mt variance(degC)^2time x station  
t't'.bph.4.5mt variance(degC)^2time  
t.4.5mAir TemperaturedegCtime x station  
t.bph.4.5mAir TemperaturedegCtime  
tc'tc'.4.5mtc variance(degC)^2time x station  
tc'tc'.9mtc variance(degC)^2time  
tc'tc'.csattc variance(degC)^2time  
tc'tc'.nuw1tc variance(degC)^2time  
tc'tc'.nuw2tc variance(degC)^2time  
tc.4.5mVirtual TemperaturedegCtime x station  
tc.9mVirtual TemperaturedegCtime  
tc.csatVirtual TemperaturedegCtime  
tc.nuw1Virtual TemperaturedegCtime  
tc.nuw2Virtual TemperaturedegCtime  
tcflag.4.5mSpike fract for tc time  
tcflag.9mSpike fract for tc time  
tcflag.csatSpike fract for tc time  
u'h2o'.4.5mu h2o covariancem/s g/m^3time x station  
u'h2o'.9mu h2o covariancem/s g/m^3time  
u'mr'.(,bph).4.5m m/s gm/kgtime  
u't'.(,bph).4.5mu t covariancem/s degCtime  
u't'.4.5mu t covariancem/s degCtime x station  
u'tc'.4.5mu tc covariancem/s degCtime x station  
u'tc'.9mu tc covariancem/s degCtime  
u'tc'.csatu tc covariancem/s degCtime  
u'tc'.nuw1u tc covariancem/s degCtime  
u'tc'.nuw2u tc covariancem/s degCtime  
u'u'.4.5mu variance(m/s)^2time x station  
u'u'.9mu variance(m/s)^2time  
u'u'.csatu variance(m/s)^2time  
u'u'.nuw1u variance(m/s)^2time  
u'u'.nuw2u variance(m/s)^2time  
u'v'.4.5mu v covariance(m/s)^2time x station  
u'v'.9mu v covariance(m/s)^2time  
u'v'.csatu v covariance(m/s)^2time  
u'v'.nuw1u v covariance(m/s)^2time  
u'v'.nuw2u v covariance(m/s)^2time  
u'w'.4.5mu w covariance(m/s)^2time x station  
u'w'.9mu w covariance(m/s)^2time  
u'w'.csatu w covariance(m/s)^2time  
u'w'.nuw1u w covariance(m/s)^2time  
u'w'.nuw2u w covariance(m/s)^2time  
u.4.5mSonic U Wind Compm/stime x station  
u.9mSonic U Wind Compm/stime  
u.csatSonic U Wind Compm/stime  
u.nuw1Sonic U Wind Compm/stime  
u.nuw2Sonic U Wind Compm/stime  
uaflag.nuw1 Gtime  
uaflag.nuw2 Gtime  
uasamples.nuw1  time  
uasamples.nuw2  time  
ubflag.nuw1  time  
ubflag.nuw2  time  
ubsamples.nuw1 Gtime  
ubsamples.nuw2 Gtime  
ucflag.nuw1  time  
ucflag.nuw2  time  
ucsamples.nuw1 Gtime  
ucsamples.nuw2 Gtime  
uflag.4.5mSpike fract for uGtime  
uflag.9mSpike fract for uGtime  
uflag.csatSpike fract for u time  
usamples.4.5m# of samples averagedGtime  
usamples.9m# of samples averagedGtime  
v'h2o'.4.5mv h2o covariancem/s g/m^3time x station  
v'h2o'.9mv h2o covariancem/s g/m^3time  
v'mr'.(,bph).4.5m m/s gm/kgtime  
v't'.(,bph).4.5mv t covariancem/s degCtime  
v't'.4.5mv t covariancem/s degCtime x station  
v'tc'.4.5mv tc covariancem/s degCtime x station  
v'tc'.9mv tc covariancem/s degCtime  
v'tc'.csatv tc covariancem/s degCtime  
v'tc'.nuw1v tc covariancem/s degCtime  
v'tc'.nuw2v tc covariancem/s degCtime  
v'v'.4.5mv variance(m/s)^2time x station  
v'v'.9mv variance(m/s)^2time  
v'v'.csatv variance(m/s)^2time  
v'v'.nuw1v variance(m/s)^2time  
v'v'.nuw2v variance(m/s)^2time  
v'w'.4.5mv w covariance(m/s)^2time x station  
v'w'.9mv w covariance(m/s)^2time  
v'w'.csatv w covariance(m/s)^2time  
v'w'.nuw1v w covariance(m/s)^2time  
v'w'.nuw2v w covariance(m/s)^2time  
v.4.5mSonic V Wind Compm/stime x station  
v.9mSonic V Wind Compm/stime  
v.csatSonic V Wind Compm/stime  
v.nuw1Sonic V Wind Compm/stime  
v.nuw2Sonic V Wind Compm/stime  
vflag.4.5mSpike fract for v time  
vflag.9mSpike fract for v time  
vflag.csatSpike fract for v time  
vsamples.4.5m# of samples averagedGtime  
vsamples.9m# of samples averagedGtime  
w'h2o'.4.5mw h2o covariancem/s g/m^3time x station  
w'h2o'.9mw h2o covariancem/s g/m^3time  
w'mr'.(,bph).4.5m m/s gm/kgtime  
w't'.(,bph).4.5mw t covariancem/s degCtime  
w't'.4.5mw t covariancem/s degCtime x station  
w'tc'.4.5mw tc covariancem/s degCtime x station  
w'tc'.9mw tc covariancem/s degCtime  
w'tc'.csatw tc covariancem/s degCtime  
w'tc'.nuw1w tc covariancem/s degCtime  
w'tc'.nuw2w tc covariancem/s degCtime  
w'w'.4.5mw variance(m/s)^2time x station  
w'w'.9mw variance(m/s)^2time  
w'w'.csatw variance(m/s)^2time  
w'w'.nuw1w variance(m/s)^2time  
w'w'.nuw2w variance(m/s)^2time  
w.4.5mSonic W Wind Compm/stime x station  
w.9mSonic W Wind Compm/stime  
w.csatSonic W Wind Compm/stime  
w.nuw1Sonic W Wind Compm/stime  
w.nuw2Sonic W Wind Compm/stime  
wflag.4.5mSpike fract for w time  
wflag.9mSpike fract for w time  
wflag.csatSpike fract for w time  
wsamples.4.5m# of samples averagedGtime  
wsamples.9m# of samples averagedGtime  


Most sensors performed as expected, so only a few changes were made to the data from the in-field configuration. The primary reason to reprocess the data were to ensure that the format was consistent throughout the project. (As usual, some changes were made during the program.)

Temperature-Humidity Sensors

Considerable effort was expended during OASIS98 to reduce radiation errors in the NCAR temperature measurements. No attempt to correct the data was made (since it would be very difficult to model this effect), so the NetCDF data are unchanged from the in-field results. See the logbook for details of how the configuration of the sensors changed. The final configuration, with a pump connected through a manifold, appeared to be the most effective. The sensors at 1.5 and 9m were operated in this mode beginning 23 July and the sensors at 0.5 and 4.5m were added on 29 July (when a bigger pump was delivered).

Propeller-vane Anemometers

The azimuth angles of the boom were entered into the props on June 21, so the wind directions for earlier data were not defined. Thus, the U and V components in the NetCDF files have been set to NA. Data from the sonic anemometers at 4.5 and 9m can be used to provide wind information before this date.

Note that small wind direction differences were observed during the field program which were tracked down to the vanes being slightly bent. Once this problem was identified, some of the vanes were flipped (upside-down) to make the directions more consistent. Again, no attempt to correct the data was made (since it would be very difficult to model this effect and the error is smaller than the accuracy needed for wind directions for this program). See the logbook for details of this problem.

Sonic Anemometers

The sonic anemometers performed reasonably well during OASIS. The ATIs had some spikes related to high temperatures, but these generally were under 1% of the time and should not affect the statistics. The major exception was that U2 on the 9m ATI developed an increased sensitivity to temperature and was replaced on 14 July.

The new UW sonic anemometers spiked much worse (5-10%; probably due to the longer pathlength). Furthermore, nuw1 developed a transducer problem on 7/26, and nuw2 stopped reporting data on 8/5. These problems are being investigated. The geometry for both arrays was measured on 10/6/98 and found to be within 1 degree of specifications. The new angles have been used in post-processing of the data.

Tilt corrections have been calculated for all of the sonic anemometers using our standard algorithm. These corrections have been applied to all of the statistics in the data files (including those from the Oklahoma Mesonet CSAT sonic anemometer). The maximum "lean" angle found was 1.9 degrees, and most were less than 1 degree, which is typical for a nearly flat site. The values used are:

# yr mon day hh:mm(GMT) lean    leanaz  woffset

98 6 23 00:00   0.642   176.3   0.018

98 6 17 00:00   1.239   25.3    0.006
# Releveled using bubble level
98 6 29 14:30   0.316   -146.9  -0.022

98 6 17 00:00   0.693   27.4    -0.017
# After v transducer replacement
98 7 10 00:00   1.908   20.5    -0.023

98 7 15 00:00	0.771   -26.6
# After rotating
98 7 24 15:00	0.882   -74.0
# Sensor goes bad (path a transducer?)
98 7 26 12:00   0       0       0

98 7  1 00:00   0.300   27.6
# after flipping and rotating 
# (179.3 = 180.0-0.6 and -177 = 2.7 - 180, to compensate for flip)
98 7 24 15:00	179.341 -177.3
The corresponding plots are:
4.5m.NCAR; first part
4.5m.NCAR; second part
9m.NCAR; first part
9m.NCAR; second part
nuw1; before flip & rotate
nuw1; after flip & rotate
nuw2; before flip & rotate
nuw2; after flip & rotate


Tony Delany has written a report on OASIS98 Radiometer Response. Several potential problems with some of the sensors have been identified, however the only correction which was applied to the data in post-processing was correction of the case temperature calibration from the Okalahoma CNR1 (and recalculation of its long-wave radiation). As noted in the last section of Tony's report,h this accounted for about half of the difference between the long-wave radiation from the Epplys and CNR1 observed in the field.

Soil Properties

Tony Delany has written a summary of Soil Measurements for OASIS98, including the gravimetric soil moisture samples, the manual bulk density determinations, and the resulting calibrations for the CSI soil moisture probes. The new second-order calibrations to the CSI probes have been applied to the data in the NetCDF files.

Daily Weather Plots

The following plots summarize conditions measured by the NCAR sensors for each day of the project. Each plot covers one days (00-23 CDT) and is labeled with time in GMT at the bottom and local time (CDT) at the top. The top panel displays temperature and specific humidity measured at both 1.5 and 9m, pressure, and precipitation rates (if present). Below that is a plot of wind speed and direction measured at 9m, with a dotted line showing the direction of the best fetch (South). The next panel shows net radiation, sensible and latent heat flux, and the heat into the surface (derived from the soil heat flux, temperature, and moisture sensors). The bottom panel shows the Monin-Obukhov stability parameter, z/L, the friction velocity, u*, and the Bowen ratio calculated from the flux data. Since these fluxes and derived parameters are based on smoothed, 5-minute average statistics, they should not be used quantitatively and are only shown for guidance in selecting periods to analyze further.

18 19 20
21 22 23 24 25 26 27
28 29 30 . . . .
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31 .
2 3 4 5 6 7 8

Other plots

ATI/CSAT Comparison

We observed differences in the statistics between the CSAT3 and ATI sonic anemometers which we were unable to explain due to spatial averaging, tilts, aliasing, or data sampling effects. (See logbook entries 44, 50, 89, and 105.) A plot of spectra for one case is given here in GIF (large) and PostScript (small) formats.

NUW Comparison

The first attempt at plotting the comparison of the new UW sonic anemometers is shown here. The left panels show the speed from each of the two sonics differenced from the prop (at 4.5m) and the right panels show the wind speed ratio from the two sonics. These data are 10-minute averages of the data which have been tilted and rotated into geographic coordinates. Only data with wind directions within 80 degrees of South are shown, since the wake from either the tower, boom, or other sonic would be present for other directions. Also, data with speeds less than 2 m/s have not been plotted.

If anything, the flipped and rotated comparision looks better than for the parallel configuration. This may be due to the arrangement of the supporting booms. The sonic arrays were closer together (and thus would have affected each other over a wider range of wind directions) for the parallel configuration than for flipped and rotated. However, in both cases directions from about 110 to 250 through 180 should have been okay. This still doesn't explain the apparent variation of the parallel data - perhaps the tilt correction was confused by the distortion from the sides and should be recomputed.

Also, the flipped and rotated data show a consistent speed difference (nuw2 higher by 2%). During this period, the Tc values also are higher from nuw2 by 4C (1.3%). Since u~Tc(t1-t2)/2d, about half of the error is explained by how the sensors were zeroed. I'm not sure why this 4C error occurred.

In any case, I would conclude that these arrays do not appear to have flow distortion (no flow distortion correction has been applied to these data) which is evident from this test.

[an error occurred while processing this directive]
This page was prepared by Steven Oncley, NCAR Research Technology Facility