Transcript UTILIZATION OF THE WEATHER GENERATOR FOR MAIZE AND …

CALIBRATION AND EVALUATION OF THE WOFOST
MODEL FOR WINTER WHEAT
Eitzinger
1
J. ,
Zalud
2
Z .,
Diepen van
Semeradova
2
D. ,
3
C. ,
2
Trnka
Oberforster
M., Dubrovsky,
4
M. ,
5
M.
for Meteorology and Physics, University of Agricultural Sciences (BOKU),Vienna, Austria , 2Mendel University of Agriculture and Forestry Brno, Zemědělská 1, 613 00 Brno, Czech Republic
3 The Winand Staring Centre for Integrated Land, Soil and Water Research (SC-DLO) Wageningen, The Netherlands , 4Institute of Atmospheric Physics, AS CR, Hradec Králové, Czech Republic
5 Institute for Plant Production, Federal office and research centre for Agriculture, Vienna, Austria
1Institute
INTRODUCTION
The aim of this study was the calibration and evaluation of WOFOST (WOrld FOod Studies,
Supit, et al., 1994) model and exploring the possibility of its use in climate change impact
studies. The model was calibrated and evaluated for two experimental sites: Marchveld
(northeastern part of Austria, latitude 48°12´ N, longitude 16° 34´ E; elevation 153 m a.s.l.)
and Žabčice (southeastern part of the Czech Republic, latitude 49°53´ N, longitude 16° 05´
E; elevation 179 m a.s.l.). The impact of the climate change for the later locality is presented
as it was estimated by WOFOST model.
COMBINED EFFECT OF INCREASED CO2 ON WINTER WHEAT
YIELD
Present conditions = (the bottom red bar)
direct effect = present weather, ambient CO2 is increased
non-direct effect = changed weather, ambient CO2 is not increased
combined effect = combination of changed weather and ambient CO2 increased (red bars)
OBJECTIVES
1) Calibration and evaluation of the WOFOST model for localities Marchveld (Austria) and
Žabčice (Czech Republic) (Fig.1 and 2)
2) Assessment of impacts of elevated CO2 concentration and related changed climatic
conditions on grain yield (Fig.4) of winter wheat
•
Sensitivity analysis of modeled yields to the initial available soil water content (Fig.5)
EXPERIMENT
(a) model calibration and evaluation under conditions of the two selected localities
(b) present climate daily weather series: observed weather data from 1961-1990 were used to
parameterize a weather generator (wg) in order to create a 99-year weather series for
Žabčice and Marchveld (presented).
(c) changed climate daily weather series: 99-year weather series were generated using the
WG of which parameters were modified according to the climate change scenario (Fig. 3)
Fig. 4: Potential and water-limited grain yields (average and standard deviation) modeled by WOFOST in
99 years- simulation with synthetic weather series for Marchfeld. The direct (effect of atmospheric CO2)
and indirect effect (through changed climate; see Fig.3 for the scenarios) of increased CO2 is displayed.
(d) ”representative” year simulated by crop models using 99-year weather input series
(above): typical management, varieties, and soil type were used as inputs (Fig 4)
The climate change scenario defines changes of the means and variability of four daily weather characteristics (maximum and minimum
temperature,global radiation and precipitation). For each sensitivity scenario, a 99-year simulation with synthetic series was carried out for
potential and stressed (water limited) simulations and 1CO2 and 2CO2 concentrations in the atmosphere.
EVALUATION
11000
11000
10000
10000
9000
9000
8000
8000
7000
6000
5000
4000
5000
2000
1000
1000
0
0
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
8000
4000
2000
1977
9000
6000
3000
1976
10000
7000
3000
1975
(effect of changes in initial available soil water under current climatic conditions)
1985
1986
1987
1988
1989
1990
calculated -water limited
1992
1993
1994
1995
1996
1997
1998
1999
year
year
experimental data
1991
calculated - potential yield
Fig.1: Evaluation of the WOFOST model for winter
wheat in Žabčice- variety Mironovská- (1975-1987),
experimental data
calculated -water limited
calculated - potential yield
Fig 2: Evaluation of the WOFOST model for winter
wheat in Marchveld- variety Perlo (1985-1999),
grain yield (kg/ha)
yield (kg/ha)
yield (kg/ha)
SENSITIVITY ANALYSIS
7000
6000
5000
4000
3000
2000
1985
CLIMATE CHANGE SCENARIO
1986
1987
1988
1989
1990
WLY-iw 0-1 cm
WLY-iw 6-9 cm
WLY-iw 9-12 cm
1991
1992
1993
WLY-iw
year 1-2 cm
potential yield
1994
1995
1996
WLY-iw 2-3 cm
1997
1998
1999
WLY-iw 3-6 cm
experimental yield
Fig.5: Sensitivity analysis of the water limited yield (WLY) for different initial available water levels (iw).
Each bar represents grain yields in the range from 0 – 120 mm of the iw (for the climate change impact
simulations the iw was set to 90 mm). The sensitivity analysis enables to demonstrate the role of initial
available soil water on the simulated winter wheat grain yield.
CONCLUSIONS
Objective 1 (Calibration and Evaluation of the crop model for conditions of North Austria and South Moravia regions):
WOFOST: The length of the vegetation was successfully evaluated for locality Marchveld, yield predictions are generally higher. Water
limited yields are overestimated which might be caused by other factors (nutrients, pest and diseases) that were not included
in the simulations or by incorrectly estimated values of the initial available soil water (Fig. 5).
Objective 2 (Climate change impacts on yield):
Winter wheat: Combined effect of changed weather and increased CO2: slight increase of yields in 2CO2 climate; production potential
of winter wheat (ratio between stressed and potential yield) will slightly decrease
Fig. 3. Climate change scenario for South Moravia and North Austria and 2CO2 (year=2083; thick lines)
and 1.5CO2 year=2041; thin lines) conditions based on ECHAM4 transient simulation.
Objective 3 (Initial available soil water sensitivity analysis):
Generally: Errors in the estimate of initial available soil water alone might be held responsible for most of improper results in the
evaluation processes. The model seems to be too sensitive to this value under conditions of the North Austria – South
Moravia regions. The empirical value of 90 mm was used in climate change impact simulations as the best substitute for
unavailable data.