Abstract
Accurate crop status forecasting benefits from assimilating remote
sensing observations and crop model simulations. When conducting data
assimilation (DA) using an Ensemble Kalman filter (EnKF), arbitrary
inflation factors are normally adopted to account for unspecified
uncertainties, thus avoiding filter divergence. Here, we developed a
Bayesian methodology in which the uncertainties were systematically
quantified by combining disparate methods in one framework. Its
applicability and performance with crop model GECROS using the EnKF
framework were tested against data collected from two years of field
experiments, in which aboveground biomass
(W above), grain weight
(W grain), aboveground nitrogen (N) content
(N above), grain N content
(N grain) and leaf traits like leaf dry weight,
leaf N content and leaf area index were measured for rice. Using only
the observations from the first year, the uncertain parameters in GECROS
were calibrated by a Markov Chain Monte Carlo approach, while the
parameters in the assumed error model that describes the uncertainties
of crop model simulations were estimated simultaneously. The calibrated
model parameters performed well in the validation year, except for the
simulated leaf traits (Normalized Root Mean Squared Error (NRMSE )
> 0.38). Remotely sensed leaf traits predicted by a
Gaussian Process Regression (GPR) model were more accurate (NRMSE< 0.34), with uncertainties of the remote sensing observations
estimated from the GPR model itself. Assimilating simulated and
predicted leaf traits with their estimated uncertainties into EnKF
prevented filter divergence, and the forecast accuracy of crop model
improved in the validation year. Compared with simulation without
assimilating in-season remote sensing observations, the NRMSE of
updated whole-season W above andN above both decreased from 0.37 to 0.20; and
those of updated W grain andN grain at harvest decreased from 0.40 and 0.28 to
0.22 and 0.19, respectively. The developed method contributes to
systematic uncertainty analysis in DA and accurate forecasting of
in-season and end-of-season crop carbon and N status for smart farming.