![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
Temporal downscaling of soil CO2 efflux survey measurements based on time-stable spatial patterns |
VerfasserIn |
A. Graf, N. Prolingheuer, M. Herbst, J. A. Huisman, L. Weihermüller, B. Scharnagl, C. Steenpass, R. Harms, H. Vereecken |
Konferenz |
EGU General Assembly 2009
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 11 (2009) |
Datensatznummer |
250027134
|
|
|
|
Zusammenfassung |
Chamber measurements of soil CO2 efflux are known to require repetitions at different points
in space in order to achieve a high accuracy time series of the area average. In the absence of
multiple gas analyzers, which are a limiting factor in most field studies, this is usually
achieved either by automatic multiplexing or by manual surveys. As a trade-off, if t1 is the
interval between two measurements and N the number of different measurement points used
to reduce the error in determining the area average, the new improved-accuracy time series of
the area average has a reduced temporal resolution characterised by the interval
t2 = N * t1.
However, if measurement points keep their (relative) deviation from the area average for a
time considerably longer than t2, this additional information can be used to either reduce
measurement effort or reconstruct an estimated unbiased time series of any resolution
between t1 and t2. The former has already been demonstrated for soil moisture and soil CO2
efflux.
Here, we give an overview of simple scaling methods that can be used to achieve the latter
objective, i.e. temporal downscaling. The raw time series consisting of different measurement
points is decomposed into a moving average over all points, a temporally stable deviation of
each point from this, and a residual term comprising both fast temporal variability
and random errors. By removing the second term, a time series of any resolution
t3 = t1 *n,n = 1...N can be regained, which is subject to an increasing random error with
decreasing n but not biased due to systematic deviations of single points from the area
average.
With respect to the time scale of stability and to the definition and removal of the stable
deviation of each point from the area average, several variations of this method can be
distinguished, e.g. constant offset, constant factor, constant relative offset or first order
regression (offset and factor). We compared these methods for a dataset of circular repeated
soil CO2 efflux measurements on transects of up to 30 points (t2 = 1.5 h). Rapid
meteorological changes in environmental conditions are used to qualitatively assess the
ability of the method to describe short-term changes in the area average of soil CO2 efflux. |
|
|
|
|
|