Multi-source data assimilation for physically based hydrological modeling of an experimental hillslope

Abstract

Abstract. Data assimilation has recently been the focus of much attention for integrated surface–subsurface hydrological models, whereby joint assimilation of water table, soil moisture, and river discharge measurements with the ensemble Kalman filter (EnKF) has been extensively applied. Although the EnKF has been specifically developed to deal with nonlinear models, integrated hydrological models based on the Richards equation still represent a challenge, due to strong nonlinearities that may significantly affect the filter performance. Thus, more studies are needed to investigate the capabilities of the EnKF to correct the system state and identify parameters in cases where the unsaturated zone dynamics are dominant, as well as to quantify possible tradeoffs associated with assimilation of multi-source data. Here, the CATHY (CATchment HYdrology) model is applied to reproduce the hydrological dynamics observed in an experimental two-layered hillslope, equipped with tensiometers, water content reflectometer probes, and tipping bucket flow gages to monitor the hillslope response to a series of artificial rainfall events. Pressure head, soil moisture, and subsurface outflow are assimilated with the EnKF in a number of scenarios and the challenges and issues arising from the assimilation of multi-source data in this real-world test case are discussed. Our results demonstrate that the EnKF is able to effectively correct states and parameters even in a real application characterized by strong nonlinearities. However, multi-source data assimilation may lead to significant tradeoffs: the assimilation of additional variables can lead to degradation of model predictions for other variables that are otherwise well reproduced. Furthermore, we show that integrated observations such as outflow discharge cannot compensate for the lack of well-distributed data in heterogeneous hillslopes.

Covariance resampling for particle filter – state and parameter estimation for soil hydrology

“…Data assimilation methods, originally used for state estimation only, are adapted to also estimate parameters and other model components like the boundary condition. The ensemble Kalman filter (EnKF; Evensen, 1994;Burgers et al, 1998) is a popular data assimilation method in hydrology. It has the advantage of using the ensemble covariance to correlate dimensions with observations to unobserved dimensions.…”

“…Using a hydrological model based on the Richards equation, the EnKF is mostly applied in synthetic studies (e.g., Wu and Margulis, 2011;Song et al, 2014;Erdal et al, 2015;Shi et al, 2015;Man et al, 2016). However, some applications to real data exist (e.g., Li and Ren, 2011;Bauser et al, 2016;Botto et al, 2018).…”

Inflation method for ensemble Kalman filter in soil hydrology

“…In hydrology Vrugt et al (2005) combined an EnKF and the shuffled complex evolution Metropolis algorithm, while Moradkhani et al (2005) used a dual EnKF approach to estimate states and parameters for a rainfallrunoff model. The joint assimilation of states and parameters in an augmented state was successfully performed for example in groundwater research (e.g., Chen and Zhang, 2006;Hendricks Franssen and Kinzelbach, 2008;Kurtz et al, 2012Kurtz et al, , 2014Erdal and Cirpka, 2016), but also in soil hydrology for land surface models (e.g., Bateni and Entekhabi, 2012;Han et al, 2014;Zhang et al, 2017) and on smaller scales based on the Richards equation (e.g., Li and Ren, 2011;Margulis, 2011, 2013;Song et al, 2014;Erdal et al, 2014; H. H. Bauser et al: Inflation method for the ensemble Kalman filter in soil hydrology Erdal et al, 2015;Shi et al, 2015;Bauser et al, 2016;Brandhorst et al, 2017;Botto et al, 2018). Due to unrepresented model errors and due to a limited ensemble size, the EnKF underestimates model errors, which can lead to filter inbreeding.…”

Application of soil moisture information for operational water management

“…Although the application of multivariate data assimilation is promising, more research is required before implementation in operational applications. For example, multivariate data assimilation can lead to trade-o s. Botto et al (2018) shows that assimilating three related variables (pressure head, soil moisture, and subsurface out ow) in the CATHY hydrological model using an EnKF scheme leads to a decrease in accuracy of other hydrological variables which were otherwise simulated well.…”

“…Data assimilation aims to nd an optimal combination of merging hydrological model state estimates with observations. Several studies have shown the value of data assimilation schemes for integrated surface-subsurface modelling (Camporese et al, 2009a,b;Zhang et al, 2016;Botto et al, 2018;Zhao and Yang, 2018), some speci cally focusing on operational applications (Hendricks Franssen et al, 2011;De Rosnay et al, 2013;Kurtz et al, 2017;He et al, 2019).…”