|
Titel |
In-situ comparison of evapotranspiration estimates at the Rietholzbach catchment |
VerfasserIn |
Irene Lehner, Sonia I. Seneviratne |
Konferenz |
EGU General Assembly 2011
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 13 (2011) |
Datensatznummer |
250055754
|
|
|
|
Zusammenfassung |
The eddy covariance (EC) method is widely used to measure water, energy, and carbon fluxes
and has been established as a reference for global networks (e.g. Baldocchi et al. 2001).
However, uncertainties in the energy flux estimates resulting in a non-closure of the energy
balance are known (e.g. Franssen et al. 2010) particularly in non-homogeneous and/or
complex terrain. Evapotranspiration (ET) which is a significant term both for the water and
the energy balance can be estimated with several methods (Seneviratne et al. 2010). In this
study we compare ET estimates derived with the micrometeorological EC method, the
hydrological lysimeter approach, and the catchment water balance.
The examined data set originates from the Rietholzbach catchment which
is a small, pre-alpine research basin in the north-eastern part of Switzerland
(http://www.iac.ethz.ch/url/rietholzbach). The area of 3.31Â km2 is only sparsely
populated and primarily used as pasture land (73Â %), steeper slopes are forested
(24Â %). In 1975 measurements were initiated to determine and understand the
water balance and related processes. Amongst others a weighing lysimeter was
built and catchment runoff as well as precipitation measurements were introduced.
Recently, in May 2009 a flux tower in a grassland next to the valley bottom was
established.
Every mentioned technique to estimate ET has its general and site-specific advantages
and limitations:
(i) Applying the EC method half hourly statistics based on 10Â Hz data from a ultrasonic
anemometer (CSAT3, Campbell Scientific, Logan, USA) and an open-path CO2/H2O
infrared gas analyser (Li7500, Li-Cor, Lincoln, USA) are calculated. The flux footprint is a
few hundred m2 under convective conditions. Given the tower location, the ET
estimates represent grassland evapotranspiration only. In addition to the mentioned ET
underestimation, the gas analyser measurements fail when water is on the sensor surface
which is mainly a problem during precipitation events when ET can be assumed to be
negligible.
(ii) Lysimeters are an accurate but costly technique to determine ET. The change in
weight of the grass-covered lysimeter with a surface area of 3.14Â m2 and the outflow at the
lysimeter bottom are measured on an hourly time step (balance resolution 100Â g). A general
limitation is given by the lack of connection of the soil column to lateral flow and to ground
water. In addition, snow bridges in winter distort the measurements and precipitation events
result always in an increase in weight but evapotranspiration rates are very small under these
conditions.
(iii) The water balance approach integrates over the hydrological year and the whole
catchment and determines ET as the difference of catchment precipitation and runoff. Thus, it
is a basic long-term constraint for the consistency of ET on the catchment level. It is assumed
that the change in catchment storage is negligible. This is often the case but year-to-year
carry-over effects are observed in extreme years and limits the applicability of this constraint.
Moreover precipitation is a point measurement while runoff is a catchment wide
integrator.
These various approaches are evaluated over the time period May 2009 to October 2010,
with a special focus on two drought events in autumn 2009 and summer 2010,
respectively.
References:
Baldocchi et al., Bulletin of the American Meteorological Society 82: 2415-2434,
2001
Franssen et al., Agricultural and Forest Meteorology 150: 1553-1567, 2010
Seneviratne et al., Earth-Science Reviews 99: 125-161, 2010 |
|
|
|
|
|