The ARM eddy correlation system for monitoring surface fluxes Page: 4 of 4
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to UNT Digital Library by the UNT Libraries Government Documents Department.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
3. DATA ANALYSIS
Data are analyzed and averaged over 30-min periods
starting precisely on the hour or half hour. If a single
velocity, sonic temperature, or water vapor density
value for each 0.1 a is reported as missing, that value Is
replaced by an interpolated value. If consecutive
values are missing or if 20% or more of the values have
been interpolated, the data for that 30-min period are not
analyzed. The number of missing measurements and
the number of sonic measurements taken from 1, 2, ...,
10 samples per 0.1 s are tabulated. The level angles for
the sonic anemometer, the computer system enclosure
temperature, and the battery voltage are also measured
and recorded. Means and variances from the means are
computed for the orthogonal winds, sonic temperature,
water vapor density, horizontal wind speed, air
temperature, and barometric pressure. Wind direction is
computed from the vector-averaged winds.
Covariances from the means are computed from the
orthogonal winds, sonic temperature, water vapor
density, and horizontal wind speed data. Variances and
covariances from the running means are computed from
the orthogonal winds, sonic temperature, water vapor
density, and horizontal wind speed data. A three-
dimensional coordinate rotation resulting in zero vertical
and transverse wind speeds is applied to the variances
and covariances from the running means. The mixing
ratio, air density, specific heat of dry air at constant
pressure, and heat of vaporization are adjusted for the
effects of air temperature, water vapor pressure, and
barometric pressure. These coefficients are used with
the coordinate-rotated covariances from the running
means to compute the friction velocity, sensible heat
flux, and latent heat flux. Postprocessing of the data is
necessary to adjust for the effects of separation
between the sensors on momentum and latent heat flux.
of density fluctuations on latent heat flux, and of water
vapor on sonic temperature and thus on the sensible
heat flux estimates, .
Power spectra and cospectra are routinely computed
for the orthogonal winds, sonic temperature, water
vapor density, and horizontal wind speed data by using
an in-place, direct fast Fourier transform (FFT). A two-
butterfly, Cooley-Tukey, radix-2 FFT and a lookup table
for the sine-cosine transfer functions are used. The
spectra are smoothed, and the number of estimates Is
reduced by averaging. Because the number of input
data to an FFT is preferably a power of two and the
fluxes are best computed from the same data as the
spectra, statistical and spectral analyses are performed
on the first 18,384 values (27 min and 18.4 s) of each
30-min averaging period. The data produced by the
ECOR system are checked during subsequent
processing against a set of minima and maxima and are
flagged If they are outside the expected range or
exceed the range for which calibrations are applicable.
Occasional spikes in the heat flux data that do not
exceed the limiting minimum or maximum normally occur
when the power spectra for the vertical velocity and
sonic temperature or their cospectra exhibit spectralE
4J
u
m100
-100
-200
-300
-400
-soo100
2 0
X -100
z-300
-400
-500.4... . .4
a
..
X + X
x x
X E10 X
. 4.4
4.
- - - - - - - - - t4. - - - - - -0 4
8 12 16 20 24
Hour (Centra S nndad 7im)Fig. 1. Fluxes measured at two locations in the
ARM SGP site on 24 July 1997. Negative values
indicate that the fluxes are directed upward.
response characteristic of white noise. The spectra can
be used to flag data with excess noise or abnormal
behavior.
The ECOR system was developed for continuous
measurements from an instrument system used
previously in short-term intensive experiments. A
period of time has been required to find and eliminate
problems. Some data collected recently are shown in
Fig. 1. Site E10 is located in Montgomery County,
Kansas, in the east central portion of the SGP site and
near the border between Kansas and Oklahoma. Site
E24, located in Caddo County in south central
Oklahoma (approximately 300 km to the southwest of
E10) exhibited considerably drier conditions than did
El 0. Such differences are critical to understanding the
thermal and moisture properties of the atmosphere
above the study region. Analysis of the flux estimates
made by both the ECOR and energy balance Bowen
ratio stations scattered across the area can yeild
estimates of averaged sensible and latent heat fluxes
suitable for use in study of single-column versions of
large-scale models.
Acknowledgment
This work was supported by the U.S. Department of
Energy, Office of Energy Research, Office of Biological
and Environmental Research, Environmental Sciences
Division, under contract W-31 -1 09-ENG-38.
Search Inside
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Hart, R.L.; Cook, D.R. & Wesely, M.L. The ARM eddy correlation system for monitoring surface fluxes, article, December 31, 1998; Illinois. (https://digital.library.unt.edu/ark:/67531/metadc693990/m1/4/: accessed March 28, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.