 |
| Titel |
Assessing Tillage Effects on Soil Hydraulic Properties via Inverse Parameter Estimation using Tension Infiltrometry |
| VerfasserIn |
Andreas Schwen, Gernot Bodner, Willibald Loiskandl |
| Konferenz |
EGU General Assembly 2010
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 12 (2010) |
| Datensatznummer |
250036510
|
|
|
|
|
|
| Zusammenfassung |
| Hydraulic properties are key factors controlling water and solute movement in soils. While
several recent studies have focused on the assessment of the spatial variability of hydraulic
properties, the temporal dynamics are commonly not taken into account, primarily because its
measurement is costly and time-consuming. However, there is extensive empirical evidence
that these properties are subject to temporal changes, particularly in the near-saturated range
where soil structure strongly influences water flow. One main source of temporal
variability is soil tillage. It can improve macroporosity by loosening the soil and
thereby changing the pore-size distribution. Since these modifications are quite
unstable over time, the pore space partially collapses after tillage. This effect should be
largest for conventional tillage (CT), where the soil is ploughed after harvest every
year.
Assessing the effect of different tillage treatments on the temporal variability of hydraulic
properties requires adequate measurement techniques. Tension infiltrometry has become a
popular and convenient method providing not only the hydraulic conductivity function but
also the soil rentention properties. The inverse estimation of parameters from infiltration
measurements remains challenging, despite some progress since the first approach of
Šimůnek et al. (1998). Measured data like the cumulative infiltration, the initial and final
volumetric water content, as well as independently measured retention data from soil core
analysis with laboratory methods, have to be considered to find an optimum solution
describing the soil’s pore space.
In the present study we analysed tension infiltration measurements obtained several
times between August 2008 and December 2009 on an arable field in the Moravian
Basin, Lower Austria. The tillage treatments were conventional tillage including
ploughing (CT), reduced tillage with chisel only (RT), and no-tillage treatment using a
direct seeding technique (NT). Infiltration measurements were supplemented by
retention data for dryer conditions (-10 to -300 kPa) as determined by pressure plate
extraction on steel core samples. The HYDRUS 2D/3D software package was used to
inversely fit the parameters of suitable soil retention models to the data. Beside the
most common model of Van Genuchten (1978), we also assessed the lognormal
distribution model proposed by Kosugi (1994) and the dual porosity approach of Durner
(1994).
We will show that a dual porosity model best fits the infiltration data. It allows the
retention curve to account for both the cumulative infiltration (structure-controlled flow) as
well as the measured retention data points (texture-controlled flow). The pressure plate
extraction data are used to determine one of the shape-determining parameter sets of the
model equation. As the tillage effect on the temporal variability is expected to be negligible
for texture-controlled water flow, these parameters are set constant with time for each tillage
treatment. All remaining model parameters were inversely determined by the infiltration
data.
The advantage of a bimodal retention model is greatest when macropores strongly
contribute to water movement. This can be observed especially under conventional tillage,
where a strong increase of macropores is caused by annual ploughing, but also for reduced
(oder minimum) tillage treatments, where biopores from earthworm burrows and dead plant
roots increase macropore flow. |
|
|
|
|
|
|