 |
| Titel |
Assessing factors that influence deviations between measured and calculated reference evapotranspiration |
| VerfasserIn |
Marek Rodny, Reinhard Nolz |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250143635
|
| Publikation (Nr.) |
 EGU/EGU2017-7377.pdf |
|
|
|
|
|
| Zusammenfassung |
| Evapotranspiration (ET) is a fundamental component of the hydrological cycle, but
challenging to be quantified. Lysimeter facilities, for example, can be installed and operated
to determine ET, but they are costly and represent only point measurements. Therefore,
lysimeter data are traditionally used to develop, calibrate, and validate models that allow
calculating reference evapotranspiration (ET0) based on meteorological data, which can be
measured more easily. The standardized form of the well-known FAO Penman-Monteith
equation (ASCE-EWRI) is recommended as a standard procedure for estimating ET0
and subsequently plant water requirements. Applied and validated under different
climatic conditions, the Penman-Monteith equation is generally known to deliver
proper results. On the other hand, several studies documented deviations between
measured and calculated ET0 depending on environmental conditions. Potential reasons
are, for example, differing or varying surface characteristics of the lysimeter and
the location where the weather instruments are placed. Advection of sensible heat
(transport of dry and hot air from surrounding areas) might be another reason for
deviating ET-values. However, elaborating causal processes is complex and requires
comprehensive data of high quality and specific analysis techniques. In order to assess
influencing factors, we correlated differences between measured and calculated ET0 with
pre-selected meteorological parameters and related system parameters. Basic data were
hourly ET0-values from a weighing lysimeter (ET0_lys) with a surface area of
2.85 m2 (reference crop: frequently irrigated grass), weather data (air and soil
temperature, relative humidity, air pressure, wind velocity, and solar radiation), and
soil water content in different depths. ET0_ref was calculated in hourly time steps
according to the standardized procedure after ASCE-EWRI (2005). Deviations between
both datasets were calculated as ET0_lys−ET0_ref and separated into positive
and negative values. For further interpretation, we calculated daily sums of these
values. The respective daily difference (positive or negative) served as independent
variable (x) in linear correlation with a selected parameter as dependent variable (y).
Quality of correlation was evaluated by means of coefficients of determination
(R2). When ET0_lys > ET0_ref, the differences were only weakly correlated with
the selected parameters. Hence, the evaluation of the causal processes leading to
underestimation of measured hourly ET0 seems to require a more rigorous approach. On
the other hand, when ET0_lys < ET0_ref, the differences correlated considerably
with the meteorological parameters and related system parameters. Interpreting the
particular correlations in detail indicated different (or varying) surface characteristics
between the irrigated lysimeter and the nearby (non-irrigated) meteorological station. |
|
|
|
|
|
|