Abstract:
Croplands are man-made ecosystems that have
high net primary productivity during the growing season of
crops, thus impacting carbon and other exchanges with the
atmosphere. These exchanges play a major role in nutrient
cycling and climate change related issues. An accurate
representation of crop phenology and physiology is important
in land-atmosphere carbon models being used to predict
these exchanges. To better estimate time-varying exchanges
of carbon, water, and energy of croplands using the Simple
Biosphere (SiB) model, we developed crop-specific phenology
models and coupled them to SiB. The coupled SiBphenology
model (SiBcrop) replaces remotely-sensed NDVI
information, on which SiB originally relied for deriving Leaf
Area Index (LAI) and the fraction of Photosynthetically Active
Radiation (fPAR) for estimating carbon dynamics. The
use of the new phenology scheme within SiB substantially
improved the prediction of LAI and carbon fluxes for maize,
soybean, and wheat crops, as compared with the observed
data at several AmeriFlux eddy covariance flux tower sites
in the US mid continent region. SiBcrop better predicted
the onset and end of the growing season, harvest, interannual
variability associated with crop rotation, day time carbon uptake (especially for maize) and day to day variability
in carbon exchange. Biomass predicted by SiBcrop had good
agreement with the observed biomass at field sites. In the future,
we will predict fine resolution regional scale carbon and
other exchanges by coupling SiBcrop with RAMS (the Regional
Atmospheric Modeling System).
