The Effect of Changed Climate Variability on Simulated Crops
A variant of the daily weather generator of Richardson (1981) was
modified for these studies. In FY96 Linda Mearns, with colleagues at
the Goddard Institute for Space Studies (GISS) and Larry McDaniel,
published a paper on detailed sensitivity analyses. The results of
this study, wherein both interannual and daily variability was changed,
showed larger changes in simulated wheat yields. In FY97 they
published a paper in Climatic Change on the extension of the
work to forming scenarios of climate change with only mean changes, and
with mean and variability changes, as predicted from runs of the
regional climate model (RegCM) over the western two-thirds of the US.
These scenarios were then applied to the CERES-wheat model for four
locations in the US central Plains.
Substantial differences in simulated yield resulted between the mean
and mean-plus variability scenarios at most locations. In FY98 they
extended this research to other locations in the Great Plains and
are currently applying variance changes to corn and soybean models.
As an extension of this work, a project studying the impact of variability changes on forest/ecosystems in the Northwestern US, funded by NSF and in collaboration with scientists at U Washington and Oregon State U, is continuing.
Back to Top
Development of Interactive Vegetation Package for Regional Climate Model
Elena Tsvetsinskaya (U Nebraska), with Mearns and Filippo Giorgi (CGD and Trieste Institute of Physics), has coupled the CERES-maize model into RegCM2. The growth functions of CERES-maize were incorporated into the biosphere-atmosphere transfer scheme (BATS), which is the surface scheme for the regional climate model, RegCM2. Off-line tests of coupled CERES-BATS indicated that strong responses (of plant height, growth of leaf area index, and surface radiative fluxes) to different temperature and precipitation conditions were found. Coupling of the interactive surface scheme with RegCM2 has been completed. RegCM2, with the coupled surface package, has been run for the domain of the Great Plains of the US to determine the effect of the growing vegetation on surface fluxes and local climate.
The model was run using European Centre for Medium Range Forecasting (ECMWF) boundary conditions for 1991, a normal year, and 1988, a dry year. Results indicate that for 1988 large differences occur between the non-interactive run and the interactive run. With the interactive growth and development module, the simulated climate is warmer and drier than in the default run, and closer to the observed climate. Differences in 1991 were less striking. These results indicate that including growth and development of vegetation in a climate model can have important effects on the simulated climate. This work formed part of the ESIG contribution to CMAP.
Back to Top