 |
| Titel |
Modeling of soil water content and soil temperature at selected U.S. and central European stations using SoilClim model |
| VerfasserIn |
P. Hlavinka, M. Trnka, J. Balek, Z. Zalud, M. Hayes, M. Svoboda, J. Eitzinger |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250029143
|
|
|
|
|
|
| Zusammenfassung |
| Within the presented study the SoilClim model was tested through various climatic and soil
conditions. SoilClim model enables to estimate reference and actual evapotranspiration from
defined vegetation cover and consequently the soil water content within two defined layers
(named as Moisture control section I and II) could be deduced. The soil temperature in 0.5 m
depth is also estimated (on the basis of simple empirical model). Mentioned outputs could be
additionally used for identification of soil climate regimes (both Hydric and Thermic) within
selected location. The SoilClim works in daily step and needs daily maximum and minimum
air temperature, global radiation, precipitation, air humidity and wind speed as input. The
brief information about soil layers (field capacity, wilting point, depth) and vegetation cover
is necessary. The algorithm for reference evapotranspiration is based on Penman-Monteith
method.
The main aim of the study was to assess accuracy and suitability of the SoilClim for
simulation of soil water content in the two defined layers and temperature in 0.50 m depth.
For this purpose the seven stations through central U.S. were selected (by twos from
Nebraska, Iowa and Kansas and one from South Dakota). Used measurements were observed
from 2004 to 2008. The central European region was represented by Austrian Lysimetric
station Gross-Enzersdorf. The data within three different soil profiles and for various crop
covers (spring barley, winter wheat, maize and potato) from 1999 to 2004 were
used.
During introduced reserch SoilClim provided reasonable results of soil moisture for both
layers against lysimetric measurements. Agreement between measured and estimated water
content (30 days averages) could be described by coefficient of determination (R2) which
varied from 0.45 to 0.75. The Mean Bias Error (MBE) for values in daily step was from
-12.87 % to 20.66 % and Root Mean Square Error (RMSE) varied from 14.49 % to 34.76 %.
The modeling efficiency index (MEI) was from 0 to 0.67 through the all lysimetric
experiments. The estimates for U.S. stations showed higher inaccuracy during the winter
months. It was caused by using automatic non heated rain gauge which distort information
about time of solid precipitation. Consequently SoilClim produced some deviation in
estimated snow cover occurence. On the other hand the variability and trends of
soil water content during vegetation seasons were sufficiently explained by the
SoilClim.
The satisfactory results were achieved within soil temperature (in 0.5 m depth) simulation
within U.S. stations while R2 varied from 0.69 to 0.93. Certain deviations were apparent
within winter months due to input data from non heated rain gauge.
It was concluded that tested model gives reasonable idea about soil water content and soil
temperature. The SoilClim could be successfully used for water balance analysis, drought
occurrence and soil climate assessment as well.
Acknowledgement: This study was conducted with support of The Czech Science
Foundation (GACR) project no. 521/09/P479 and project KONTAKT (AMVIS) ME 844. The
study was also supported by the Research plan No. MSM6215648905 “Biological and
technological aspects of sustainability of controlled ecosystems and their adaptability to
climate change“, which is financed by the Ministry of Education, Youth and Sports of the
Czech Republic. |
|
|
|
|
|
|