State and parameter estimation of two land surface models using the ensemble Kalman filter and the particle filter

Abstract

Abstract. Land surface models (LSMs) use a large cohort of parameters and state variables to simulate the water and energy balance at the soil-atmosphere interface. Many of these model parameters cannot be measured directly in the field, and require calibration against measured fluxes of carbon dioxide, sensible and/or latent heat, and/or observations of the thermal and/or moisture state of the soil. Here, we evaluate the usefulness and applicability of four different data assimilation methods for joint parameter and state estimation of the Variable Infiltration Capacity Model (VIC-3L) and the Community Land Model (CLM) using a 5-month calibration (assimilation) period (March-July 2012) of arealaveraged SPADE soil moisture measurements at 5, 20, and 50 cm depths in the Rollesbroich experimental test site in the Eifel mountain range in western Germany. We used the EnKF with state augmentation or dual estimation, respectively, and the residual resampling PF with a simple, statistically deficient, or more sophisticated, MCMC-based parameter resampling method. The performance of the "calibrated" LSM models was investigated using SPADE water content measurements of a 5-month evaluation period (August-December 2012). As expected, all DA methods enhance the ability of the VIC and CLM models to describe spatiotemporal patterns of moisture storage within the vadose zone of the Rollesbroich site, particularly if the maximum baseflow velocity (VIC) or fractions of sand, clay, and organic matter of each layer (CLM) are estimated jointly with the model states of each soil layer. The differences between the soil moisture simulations of VIC-3L and CLM are much larger than the discrepancies among the four data assimilation methods. The EnKF with state augmentation or dual estimation yields the best performance of VIC-3L and CLM during the calibration and evaluation period, yet results are in close agreement with the PF using MCMC resampling. Overall, CLM demonstrated the best performance for the Rollesbroich site. The large systematic underestimation of water storage at 50 cm depth by VIC-3L during the first few months of the evaluation period questions, in part, the validity of its fixed water table depth at the bottom of the modeled soil domain.

A near-term iterative forecasting system successfully predicts reservoir hydrodynamics and partitions uncertainty in real time

“…The EnKF method of data assimilation is well-suited for non-linear mechanistic models like the GLM and enables ensemble-based forecasts of future states [ Dietze 2017a ]. Our implementation of the EnKF with state augmentation to calibrate parameters (Supporting Information A) follows Zhang et al (2017).…”

Immune Evolution Particle Filter for Soil Moisture Data Assimilation

“…Three different DA algorithms, including IEPFM, EnKF, and the differential evolution particle filter with MCMC (DEPFM), are applied in both experiments at the same time. It is noted that DEPFM follows the method of Vrugt et al to enhance sample diversity, because it outperforms the standard PF and demonstrates its effectiveness in many practical applications [15,24,25]. Comparing the performance of IEPFM with EnKF and DEPFM, it is found that IEPFM has the best data assimilation effect.…”

Assimilation of Cosmic‐Ray Neutron Counts for the Estimation of Soil Ice Content on the Eastern Tibetan Plateau

“…(i.e., by perturbing the model input, as well as the observations, to within their respective error ranges). For example, in Zhang et al (2017) where a joint stateparameter assimilation was undertaken using two land surface models, a value of zero for the model error was used since it was assumed that "uncertainty was captured by uncertain model parameters and model forcings." In this study, however, arbitrary model uncertainties, in the form of random global perturbations, were also assumed to represent the existing uncertainties in parameters, structure, and forcing data.…”

“…To account for these errors, data assimilation, which is defined as the integration of nearreal-time observations (likelihood) with the numerical model output data (prior) to give enhanced estimates (posterior) of the evolving system states (Swinbank & O'Neill, 1994), is carried out to update model state (and/or parameter) simulations. Majority of assimilation studies (e.g., Draper et al, 2011;Han et al, 2015;Montzka et al, 2011Montzka et al, 2012Moradkhani, 2008;Moradkhani, Hsu, et al, 2005;Zhang et al, 2017) have utilized measured and remotely sensed data within different assimilation frameworks. Among the assimilation algorithms, the particle filter has been found suitable for dealing with the highly nonlinear nature of soil water and heat flow models (Montzka et al, 2011;Moradkhani, Hsu, et al, 2005).…”