Weighing lysimeter

An introduction is presented in Chapter 2.8.1. about the lysimeters. As an important part of my study, with providing data for calibration and validation, I give a detailed description about the two large weighing lysimeters in this subchapter, which located at Marchfeld (Groß-Enzersdorf). The scheme of the lysimeter facility is represented by Figure 4.3.

The lysimeters were installed in 1983 to study evapotranspiration at the surface, water content in the soil profile, and drainage water at the bottom outlet of the lysimeters (Neuwirth and Mottl, 1983). This facility was installed by the Swiss company “Compagnie Industrielle Radioelectrique” (Neuwirth and Mottl, 1983), and managed as well as maintained by the Institute of Hydraulics and Rural Water-Management at the University of Life Sciences (BOKU) in Vienna (Neuwirth and Mottl, 1983; Nolz et al., 2011a).

The cylindrical vessels have an inner diameter of 1.9 m, a resulting surface area of 2.85 m², and a hemispherical bottom with a maximum depth of 2.5 m. A typical soil profile was created by re-packing soil in layers as follows:

50

 sandy loam soil (0–140 cm) (30 % sand, 50 % silt, 20 % clay; porosity: 43 %),

 gravel (140–250 cm) (only macropores with low water holding capacity).

In the past years, one lysimeter and the surrounding area were permanently covered by grass and maintained in order to represent reference conditions for determination of reference evapotranspiration (Allen et al., 1998). Accordingly, the lysimeter and its surroundings were frequently cut (about twice a month during the vegetation period), irrigated (about twice a week during summer) and fertilized (twice a year) to guarantee uniform distribution and dynamic growth (Nolz et. al., 2016).

Figure 4.3. Scheme of the lysimeter facility in Groß-Enzersdorf (Nolz and Cepuder, 2008)

The soil water changes, which measured by the capacitance EnviroSCAN® measuring system, developed by the Australian company Sentek. The major components of the EnviroSCAN® sensors are the top cap, the access tube, the sensor electrodes, the sensors and the cable (Paltineanu and Starr, 1997; Sentek, 2003).

The access tube was installed directly in the lysimeter profile and that provide good contact between tube and soil. The tube is equipped with sensors in 10 cm-intervals from 10 to 160 cm with the aim of measure the changes of water content in different depth of the soil (Nolz and Cepuder, 2012).

EnviroSCAN® sensors were not able to evaluate soil water content (θ %) in the first centimeters of the soil profile and over the surface, since the zone of influence is about 10 cm in length along the axis of the probe. In addition, during the experiment in Groß-Enzersdorf, the last 90 cm (between 160 and 250 cm), created by gravel, were not detected (Nolz and Cepuder, 2012). Measurements of the sensor in 20 cm depth (θ₂₀) were assumed to represent the water content within the rooting zone (Nolz et al., 2016).

Figure 4.4. represents the schema how the lysimeter weighing facility is working.

51

Figure 4.4. Lysimeter weighing facility: a small fraction of the total mass is transmitted to an electronic load cell via a lever-arm mechanism with a counterbalance (measuring accuracy is

±0.18 kg.) (Nolz et al., 2013)

The determination of water balance can be achieved using the lysimeter with

 a tipping bucket to measure percolating water (Wdrain),

 a lever-arm-counterbalance weighing system to detect changes of mass (= water content) (Wlys) and determine fluxes at the soil-atmosphere interface. These fluxes were assumed to be positive due to precipitation and negative because of evapotranspiration. This technique requires short measuring intervals, accurate data, and a suitable data management as well.

To see how the actual ETLYS was determined from the lysimeter data, I have to introduce the water balance equation that demonstrates the correlation between measured (Wlys, Wdrain) and unknown (Plys, Ilys, ETlys) components. The equation represents a daily changes Δ of the analyzed components.

The water balance equation is the following according to Nolz et al. (2016):

(eq. 4.6.) ΔWlys: soil water [mm ∙ day⁻¹],

ΔWdrain: drainage water [mm ∙ day⁻¹], ΔP_lys: precipitation [mm ∙ day⁻¹], ΔIlys: irrigation [mm ∙ day⁻¹].

ΔETlys: evapotranspiration [mm ∙ day⁻¹],

The fluxes across the upper boundary of the lysimeter is represented by the right-hand side of the equation.

The weighing facility measures the mass changes, which is equal with changes of soil water (ΔW_lys): a mechanical lever arm counterweight system transmitted a fractional amount of lysimeter weight to an electronic load cell with a measuring accuracy of ±0.18 kg.

52

At the bottom of the lysimeter at a free drainage outlet there is a tipping bucket, which measures the drainage water quantity (ΔWdrain) flows through on this bucket.

Weighing data from lysimeters and data of the tipping bucket were measured every few seconds and stored every 10 minutes and collected on Excel sheets (Nolz et al., 2011b). The processed output signal of the load cell was registered every few seconds, averaged, and stored on a local server. Storage intervals were 15 min from 2005 to 2007 and 10 min from 2007 to 2010, respectively.

Weighing data and raw data of cumulated outflow were stored together (Nolz et al. 2011b).

Collected raw data were transformed into physical quantities using calibration factors and divided by the surface area and the density of water with the purpose of obtaining ΔWlys and ΔWdrain with a dimension of length (Nolz et al., 2013).

Noisy data as well as outliers were processed byway of smoothing operations using a natural cubic approximation spline with discontinuities for rainfall and irrigation and manually adjusted smoothing factors (Nolz et al., 2016).

ΔPlys as well as ΔIlys were calculated from increasing ΔW (=ΔWlys + ΔWdrain), but ΔETlys was recorded if ΔW was decreasing. ΔIlys was separated obviously, because the dates were known from record keeping. Ordinarily, this technique provides more credible values of ΔET_lys, than the ordinary method with rain gauge data (P), which often shows deviations to the increase of ΔW that resulting in unlikely, negative ΔETlys (Nolz et al., 2016).