TY - JOUR
T1 - Parameter estimation of a physically-based land surface hydrologic model using an ensemble Kalman filter
T2 - A multivariate real-data experiment
AU - Shi, Yuning
AU - Davis, Kenneth J.
AU - Zhang, Fuqing
AU - Duffy, Christopher J.
AU - Yu, Xuan
N1 - Publisher Copyright:
© 2015 Elsevier Ltd.
PY - 2015/9/1
Y1 - 2015/9/1
N2 - The capability of an ensemble Kalman filter (EnKF) to simultaneously estimate multiple parameters in a physically-based land surface hydrologic model using multivariate field observations is tested at a small watershed (0.08 km2). Multivariate, high temporal resolution, in situ measurements of discharge, water table depth, soil moisture, and sensible and latent heat fluxes encompassing five months of 2009 are assimilated. It is found that, for five out of the six parameters, the EnKF estimated parameter values from different test cases converge strongly, and the estimates after convergence are close to the manually calibrated parameter values. The EnKF estimated parameters and manually calibrated parameters yield similar model performance, but the EnKF sequential method significantly decreases the time and labor required for calibration. The results demonstrate that, given a limited number of multi-state, site-specific observations, an automated sequential calibration method (EnKF) can be used to optimize physically-based land surface hydrologic models.
AB - The capability of an ensemble Kalman filter (EnKF) to simultaneously estimate multiple parameters in a physically-based land surface hydrologic model using multivariate field observations is tested at a small watershed (0.08 km2). Multivariate, high temporal resolution, in situ measurements of discharge, water table depth, soil moisture, and sensible and latent heat fluxes encompassing five months of 2009 are assimilated. It is found that, for five out of the six parameters, the EnKF estimated parameter values from different test cases converge strongly, and the estimates after convergence are close to the manually calibrated parameter values. The EnKF estimated parameters and manually calibrated parameters yield similar model performance, but the EnKF sequential method significantly decreases the time and labor required for calibration. The results demonstrate that, given a limited number of multi-state, site-specific observations, an automated sequential calibration method (EnKF) can be used to optimize physically-based land surface hydrologic models.
UR - http://www.scopus.com/inward/record.url?scp=84939562590&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84939562590&partnerID=8YFLogxK
U2 - 10.1016/j.advwatres.2015.06.009
DO - 10.1016/j.advwatres.2015.06.009
M3 - Article
AN - SCOPUS:84939562590
SN - 0309-1708
VL - 83
SP - 421
EP - 427
JO - Advances in Water Resources
JF - Advances in Water Resources
ER -