Geoscience Reference
In-Depth Information
2030-2050—RegClim.
*
Agriculture is one of the sectors that
is most likely to be sensitive to the primary effects of climate
change, such as changes in growing season, temperature and
precipitation. We seek to establish a statistical relationship
between yield per decare for four crops, based on meteorologi-
cal data from 1958 until 2001, through regression analysis at
county level in Norway. In addition, we undertake analyses at
the national level. The four crops we investigated were pota-
toes, wheat (spring and winter), oats and barley. The meteoro-
logical data consist of growing degree days (GDD) and annual
precipitation. In addition, a time trend was included to account
for long-term technology and productivity changes in agricul-
ture. It accounted for, in part, the CO
2
fertilisation response due
to the steady rise in the CO
2
concentration level in the atmo-
sphere. Assuming that there were no major changes in agricul-
tural production technologies and practices during this period,
we made a prediction of yields per decare for 2040 (as a repre-
sentative year for the period 2030-2050) based on the RegClim
scenario. Through this analysis we tried to detect a climate sig-
nal in the annual weather variation and agricultural yield data
at a relatively aggregated level (county) in Norway. If such a
signal is found, the estimated impacts on agricultural produc-
tion across regions and four major crops in Norway should be
of interest for climate policy planners, agricultural authorities
and farmers in preparing for a warmer future.
The main methodological approaches studying impacts on
agriculture from climate change are presented in a handbook
by the UNEP and IVM [4]. There are two categories of tools,
biophysical and economic. Biophysical tools can be divided
into experimentation, agro-climatic indices, statistical mod-
els, process-based models and spatial or temporal analogues.
Economic tools can be divided into economic regression mod-
els, microeconomic models and macroeconomic models.
In this study, we have chosen a biophysical statistical
model, which links the primary climate change impacts on
temperature and precipitation to changes in yield per unit of
land. This choice gives priority to the secondary impacts of
climate change. A weakness of this approach is its limited
ability to predict the effect of future climate change that lies
outside the climate variability of the last decades (upon which
the estimates of the model parameters are based); another is
that there is an implied assumption of mixed technology [4].
*
See http://regclim.met.no.
