Long-term analysis of trends in actual evapotranspiration measured by lysimeters

Recent studies based on analysis of experimental data have shown that over the last decades the magnitude of evapotranspiration (both potential and actual) has been affected by global climate change although the sign and size of the change in ET differ strongly between regions around the globe.
Basically, there are two approaches that are available to measure actual evapotranspiration: the measurement of latent heat flux based on energy balance considerations and the determination of evapotranspiration by measuring the components of the soil water balance. In the energy based approach, typically Eddy Covariance methods have been used. Eddy Covariance measurements, however, suffer from closure problems in the energy balance. Evett et al. (2012) showed that Eddy Covariance measurements of actual evapotranspiration obtained in irrigated cotton fields was 31 to 45% lower than estimates obtained from soil water balance measurements using lysimeters.
Despite the fact that lysimeter systems, especially the weighing based systems, are ideal tools to determine actual evapotranspiration no global assessment has been made of available data at present that might be valuable to assess the impact of climate change on actual evapotranspiration. This is surprising as lysimeter systems have been operated in many countries sometimes for several decades. A screening of literature showed that many data are either not reported or made available through research reports rather than peer reviewed literature. One of the first reviews about available lysimeter systems was performed by the United States Department of Agriculture providing an overview of more than 50 lysimeter systems operated between 1688 and 1939. It provides an overview of some of the major findings obtained from these studies dealing with soil moisture relationships of soils including weather and climate and amount and constituents of leachate. A more recent review was given by Allen et al. (1991) based on a meeting in Hawai dealing with the use of lysimeters for evapotranspiration and environmental measurements. Lysimeters have played a key role in deriving crop coefficients parameters to obtain potential evapotranspiration of crops based on the reference potential evapotranspiration calculated using e.g. Penman Monteith (Doorenboss and Pruitt, 1977). Typically lysimeter studies have been used for well-designed experimental studies for the assessment of flow and transport processes in cropped systems that were limited in time. Still at present, we have lysimeter systems operational that have long term time series available on soil hydrological fluxes. Recently, a few studies were reported that analyzed long term series of actual evapotranspiration derived from lysimeter measurements at specific locations. Harsch et al. (2009) published results from a large scale lysimeter site at St. Arnold in Germany including precipitation, drainage and evapotranspiration for the period between 1966 and 2007 for three lysimeters covered with grassland, deciduous forest and a coniferous forest. In addition, important weather variables such as air temperature, relative humidity, sunshine duration, global radiation and wind speed were measured. They observed an increase in both potential evapotranspiration and actual evapotranspiration over the observed period but only for the forested lysimeter systems. At the grassland site, a decrease in actual evapotranspiration was observed which was attributed to an observed decrease in wind speed over the time period or due the sheltering effect of the growing forest. Seneviratne et al. (2012) analyzed monthly actual evapotranspiration data between 1976 and 2008 obtained from a lysimeter system at Rietholzbach to validate the hypothesis that the 2003 European drought event was due to spring precipitation deficits. They showed that evapotranspiration excess in June 2003 was the main driver for initiating the summer drought conditions in Rietholzbach rather than precipitation deficits.

Objectives of the project

• To establish a database of long term actual evapotranspiration, drainage and precipitation time series including weather variables obtained from lysimeter systems and sites with a minimum length of 20 years across the globe.
• To try and separate the effect of climate forcing from management practices (as some lysimeters showed different management over the time series) based on the available time series with a specific focus on evapotranspiration.
• To detect and analyze possible trends in the observed time series as well as cross-correlations between time series.
  To falsify some of the existing hypotheses related to factors leading to an observed or postulated change in global actual evapotranspiration.
• To put the observed time series of weather variables, calculated evapotranspiration and precipitation at lysimeter sites in a long term perspective by using 100 year time series of weather and precipitation values observed at weather stations near lysimeter sites.

Hypothesis
Observed water storage changes, and evaporative and drainage fluxes in lysimeter systems combined with mathematical modeling of the soil water balance may help to separate climate forcing from natural forcing and management effects on the actual evapotranspiration.

Project steps
Relevant groups having access or operating lysimeter systems have been contacted and asked to provide summary data on the available lysimeter systems and data. Long-term lysimeter data from about 10 European sites have been made available. More recently, also contacts with Chinese lysimeter owners have been established.
We started analyzing long-term time series of actual evapotranspiration. Also procedures for gap filling of lysimeter data have been compared and a quality control system has been set up. A comparison of gap filling methods showed that calculating the ratio of actual evapotranspiration (as measured by lysimeter) and potential evapotranspiration (as calculated by Penman-Monteith) before and after a gap, and interpolating this ratio over the gap period, to estimate missing actual evapotranspiration data was the best procedure. The actual evapotranspiration for the gap is calculated then from the interpolated ratio and potential evapotranspiration. In some cases, actual evapotranspiration estimated by HYDRUS mode calculations can give better results, but is also much more labour intensive to be obtained.
First analyses also started to calculate trends in long-term actual evapotranspiration recorded by lysimeters.

Contact persons:
Harrie-Jan Hendricks Franssen
Agrosphere (IBG-3)
Forschungszentrum Jülich GmbH
Leo Brandtstrasse
52425 Jülich
Germany
Tel. 02461 / 61-4462
E-mail: h.hendricks-franssen@fz-juelich.de

Yafei Huang
Agrosphere (IBG-3)
Forschungszentrum Jülich GmbH
Leo Brandtstrasse
52425 Jülich
Germany
Tel. 02461 / 61-2367
E-mail: y.huang@fz-juelich.de

Relevant publications:

Gebler, S., Hendricks Franssen, H.J., Pütz, T., Post, H., Schmidt, M. and H. Vereecken, 2015. Actual evapotranspiration and precipitation measured by lysimeters: a comparison with eddy covariance and tipping buckets. Hydrology and Earth System Sciences 19(5), 2145-2161.

Kessomkiat, W., Hendricks Franssen, H.J., Graf, A., Vereecken, H. 2013. Estimating random errors of eddy covariance data: an extended two-tower approach. Agricultural and Forest Meteorology, 171, 203-219.

Hendricks Franssen, H.J., R. Stöckli, I. Lehner, E. Rotenberg and S.I. Seneviratne. 2010. Energy balance closure of eddy covariance data: a multi-site analysis for European FLUXNET stations. Agricultural and Forest Meteorology 150, 1553-1567.

References
Allen, R., G., and T. Howell, A. 1991. Lysimeters for evapotranspiration and environmental measurements: Proceedings of the international symposium on lysimetry:Proceedings of the international symposium on lysimetry Hawaii, July 23-25.
Doorenboss J., and P. W.O. 1977. Crop water requirements. Irrigation and Drainage paper. FAO, Rome.
Evett, S.R., R.C. Schwartz, T.A. Howell, R.L. Baumhardt, and K.S. Copeland. 2012. Can weighing lysimeter et represent surrounding field et well enough to test flux station measurements of daily and sub-daily et? Adv. Water Resour. 50:79-90.
Harsch, N., M. Brandenburg, and O. Klemm. 2009. Large-scale lysimeter site st. Arnold, germany: Analysis of 40 years of precipitation, leachate and evapotranspiration. Hydrol. Earth Syst. Sci. 13:305-317.
Seneviratne, S. I., et al. (2012), Swiss prealpine Rietholzbach research catchment and lysimeter: 32 year time series and 2003 drought event, Water Resour. Res., 48, W06526, doi:10.1029/2011WR011749.