hide
Free keywords:
-
Abstract:
Extreme weather conditions can strongly affect agricultural production, with
negative impacts that can at times be detected at regional scales. In France, crop yields were
greatly influenced by drought and heat stress in 2003 and by extremely wet conditions in
2007. Reported regional maize and wheat yields where historically low in 2003; in 2007
wheat yields were lower and maize yields higher than long-term averages. An analysis with
a spatial version (10×10 km) of the EPIC crop model was tested with regards to regional
crop yield anomalies of wheat and maize resulting from extreme weather events in France
in 2003 and 2007, by comparing simulated results against reported regional crops statistics,
as well as using remotely sensed soil moisture data. Causal relations between soil moisture
and crop yields were specifically analyzed. Remotely sensed (AMSR-E) JJA soil moisture
correlated significantly with reported regional crop yield for 2002–2007. The spatial
correlation between JJA soil moisture and wheat yield anomalies was positive in dry 2003
and negative in wet 2007. Biweekly soil moisture data correlated positively with wheat
yield anomalies from the first half of June until the second half of July in 2003. In 2007, the
relation was negative the first half of June until the second half of August. EPIC reproduced observed soil dynamics well, and it reproduced the negative wheat and maize yield anomalies of the 2003 heat wave and drought, as well as the positive maize yield anomalies
in wet 2007. However, it did not reproduce the negative wheat yield anomalies due to
excessive rains and wetness in 2007. Results indicated that EPIC, in line with other crop
models widely used at regional level in climate change studies, is capable of capturing the
negative impacts of droughts on crop yields, while it fails to reproduce negative impacts of
heavy rain and excessively wet conditions on wheat yield, due to poor representations of
critical factors affecting plant growth and management. Given that extreme weather events
are expected to increase in frequency and perhaps severity in coming decades, improved
model representation of crop damage due to extreme events is warranted in order to better quantify future climate change impacts and inform appropriate adaptation responses.
