TYPE Mini Review

PUBLISHED 11 August 2022

DOI 10.3389/frwa.2022.948832

Recent advances and

OPEN ACCESS opportunities in data
assimilation for physics-based
EDITED BY
Harrie-Jan Hendricks Franssen,
Helmholtz Association of German
Research Centres (HZ), Germany

REVIEWED BY
hydrological modeling
Alain Pascal Francés,
Laboratório Nacional de Energia e
Geologia, Portugal Matteo Camporese1*† and Manuela Girotto2†
Benjamin Fersch,
1
Karlsruhe Institute of Technology Department of Civil, Environmental and Architectural Engineering, University of Padova, Padua,
(KIT), Germany Italy, 2 Department of Environmental Science and Policy Management, University of California,
Damiano Pasetto, Berkeley, Berkeley, CA, United States
Ca’ Foscari University of Venice, Italy

*CORRESPONDENCE
Matteo Camporese Data assimilation applications in integrated surface-subsurface hydrological
matteo.camporese@unipd.it
models (ISSHMs) are generally limited to scales ranging from the hillslope to
†
These authors have contributed local or meso-scale catchments. This is because ISSHMs resolve hydrological
equally to this work processes in detail and in a physics-based fashion and therefore typically
SPECIALTY SECTION require intensive computational efforts and rely on ground-based observations
This article was submitted to
with a small spatial support. At the other end of the spectrum, there is a vast
Water and Hydrocomplexity,
a section of the journal body of literature on remote sensing data assimilation for land surface models
Frontiers in Water (LSMs) at the continental or even global scale. In LSMs, some hydrological
RECEIVED 20 May 2022 processes are usually represented with a coarse resolution and in empirical
ACCEPTED 21 July 2022
ways, especially groundwater lateral flows, which may be very important and
PUBLISHED 11 August 2022
yet often neglected. Starting from the review of some recent progress in data
CITATION
Camporese M and Girotto M (2022) assimilation for physics-based hydrological models at multiple scales, we stress
Recent advances and opportunities in the need to find a common ground between ISSHMs and LSMs and suggest
data assimilation for physics-based
hydrological modeling.
possible ways forward to advance the use of data assimilation in integrated
Front. Water 4:948832. hydrological models.
doi: 10.3389/frwa.2022.948832

COPYRIGHT KEYWORDS
© 2022 Camporese and Girotto. This is
an open-access article distributed integrated surface-subsurface hydrological models, land surface models, remote
under the terms of the Creative sensing, data assimilation, groundwater, soil moisture
Commons Attribution License (CC BY).
The use, distribution or reproduction
in other forums is permitted, provided
the original author(s) and the copyright
Introduction
owner(s) are credited and that the
original publication in this journal is Physics-based hydrological modeling
cited, in accordance with accepted
academic practice. No use, distribution
or reproduction is permitted which Physics-based hydrological models are important in earth and environmental
does not comply with these terms. sciences thanks to their inherent capability of being applicable to the widest possible
range of scales and environmental conditions (Fatichi et al., 2016). These models are
increasingly being used to predict future water resources quantity and quality in response
to climate and land use change, to monitor and assess hydroclimatic hazards (such as
floods and droughts), and in general to understand the intertwined dynamics between the
hydrological, atmospheric, and carbon cycles. They are playing an ever-increasing role,
also thanks to their capabilities of considering multiple compartments of the terrestrial
water cycle (Clark et al., 2015). Depending on the scale of interest, we refer here to two
different categories of physics-based hydrological models: local to meso-scale integrated

Frontiers in Water 01 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

surface-subsurface hydrological models (ISSHMs) and 2000s (Paniconi et al., 2003). Since then, much progress has been
continental to global-scale land surface and hydrological made, also thanks to the widespread diffusion of ensemble-based
models (LSMs/GHMs). assimilation methods such as the ensemble Kalman filter and
Integrated surface-subsurface hydrological models (Furman, its variants (Evensen, 2003). Data assimilation for LSM/GHMs
2008; Maxwell et al., 2014; Paniconi and Putti, 2015; Kollet et al., has a similar history, with significant advancements over the
2018) typically resolve the Richardson-Richards equation (or past decade for estimating hydrological land surface variables
some approximation of it) in the subsurface, coupled with some (Reichle et al., 2002; de Rosnay et al., 2014).
form of the shallow water equation for surface flow. Although In general, LSM/GHM DA previous developments involve
their spatial resolution varies widely (from a few meter to several univariate data assimilation, i.e., assimilation of one observation
hundred meters), due to their high computational requirements to adjust model states (e.g., soil moisture) or parameters such
and need for detailed parameterization, ISSHMs are usually as precipitation scaling factors (e.g., Liu and Margulis, 2019;
applied to relatively small spatial domains (i.e., from hillslope to Girotto et al., 2021), albedo (Navari et al., 2018), and vegetation
regional or meso-scale catchments), with a handful of notable properties (Smith et al., 2020). More recently, the scientific
recent exceptions at the continental scale (Keune et al., 2016, community started to target multi-observational approaches.
2018; Maxwell and Condon, 2016; Kollet et al., 2018; Condon The simultaneous assimilation of multiple observation data
and Maxwell, 2019; Condon et al., 2020a,b). strongly improves model predictions compared with single
Global hydrological and land surface models (LSM/GHM), observation and/or state estimation alone (e.g., Girotto et al.,
on the other hand, were first developed in the context of earth 2019; Kumar et al., 2019; Khaki et al., 2020). Data assimilation
system modeling to resolve energy and mass balance processes applications in ISSHMs typically involve the assimilation of
at the interface between the land surface and the atmosphere. multiple measurements in both the surface (soil moisture,
Their spatial resolution typically ranges from 10 to 50 km, which streamflow) and subsurface compartments (pressure head, water
are often inadequate to fully capture the fine-scale variability of table depth) with or without the update of model parameters
surface and subsurface hydrological processes. As an example, (e.g., Camporese et al., 2009a,b; Pasetto et al., 2012; Rasmussen
they typically lack the representation of processes in highly et al., 2015, 2016; Zhang et al., 2015, 2016; Ridler et al., 2018;
heterogeneous areas such as mountains and coasts, which are Gebler et al., 2019; He et al., 2019).
also amongst the most vulnerable regions to hydro-climatic While promising in overcoming models and observations
hazards (Balica et al., 2012; Elalem and Pal, 2015). uncertainties, current literature in data assimilation for
Regardless of the scale of interest, both ISSHMs and physically based hydrological studies agree that: (i) it is very
LSM/GHMs rely on accurate representation of the hydrological challenging to improve model estimates in one compartment
physical processes as well as input meteorological forcing assimilating measurements from another (Camporese et al.,
data (e.g., precipitation, air temperature, wind speed, etc.), 2009b; Zhang et al., 2016; Botto et al., 2018); (ii) the joint
and parameters (e.g., land cover, soil type, soil properties, update of system states and model parameters usually leads
etc.), all of which can be difficult to obtain and can lead to better results (Botto et al., 2018; Gebler et al., 2019); (iii)
to large uncertainties in modeled hydrological states and data assimilation applications in integrated surface-subsurface
fluxes (Collier et al., 2018), especially when generalizing a hydrological models are generally limited to scales that go from
model to larger spatial domain. These uncertainties (and the hillslope (Botto et al., 2018; Gebler et al., 2019) to local-
errors) inherently affect all estimates of hydrological states scale (He et al., 2019) or meso-scale catchments (Rasmussen
(e.g., water table or soil moisture) obtained through modeling. et al., 2016; Ridler et al., 2018), mainly due to computational
However, they can be quantified and limited through data constraints; (iv) the robustness of the assimilation application
assimilation approaches. depends upon the accurate characterization of the combined
modeling and observation uncertainties, a task that can be
particularly challenging especially for large spatial domains
Data assimilation (Kumar et al., 2017).

Data Assimilation (DA) refers to a suite of statistical

techniques that incorporate observation data into mathematical Statement of the problem: The scale gap
models, with the goal of optimizing estimates of the system
state (and possibly parameters). The results of a DA framework From the analysis of recent literature in hydrological data
should be a statistically optimal estimate, superior to that from assimilation, a distinct scale gap emerges. Ground observations
either the model or observations alone (Evensen et al., 2022). can have a high time resolution but are typically characterized by
DA was pioneered by meteorologists and has been used to a very small spatial support, making them ideal for assimilation
improve operational weather forecasts for decades. The first in local catchment-scale or hillslope models, where the size
application of data assimilation in ISSHMs dates to the early of the computational cell/element can be of the same order

Frontiers in Water 02 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

of magnitude of the measurement volume. However, they of fine scale model processes and parameters knowledge (Beven
are expensive with respect to the (limited) spatial coverage and Cloke, 2012). Some of these challenges are listed as follows.
they offer. At the other end of the spectrum, satellite-based First, while meso-scale natural processes, such as
observations are under continuous development to reduce their precipitation-induced runoff or evaporation, are included
uncertainties and provide additional information, thanks to in most LSM/GHMs, more local anthropogenic processes, such
their large spatial coverage; for this reason, they are commonly as irrigation, are rarely modeled, especially in LSMs (Bierkens
assimilated in LSMs/GHMs. However, their use with ISSHMs et al., 2015). Human driven processes can directly control the
have been hindered by their typically coarse resolution, which is groundwater (GW) table, for instance, by lowering the GW
not consistent with scale requirements of ISSHMs (Samaniego table through pumping and by increasing recharge via flood
et al., 2010; Or et al., 2015; Or, 2020). In addition, hyper- irrigation. To date, representing human-induced impacts in
resolution modeling at continental to global scales with ISSHMs continental to global scale hydrological models remains a major
still requires prohibitive computational efforts for them to be challenge, mostly due to the lack of accurate global irrigation
used in ensemble-based data assimilation frameworks. The main information (Wada et al., 2017).
goal of this mini-review is to discuss possible new directions for Second, despite advancements in land surface model
hydrological modeling, observations, and DA methodologies, complexity, most LSM/GHMs route moisture in the vertical
which could lead to advances in data assimilation for larger dimension only, and allow no lateral communication between
spatial domains or to unlock the potential of remote sensing data vertical soil columns, i.e., they do not represent GW dynamics.
for assimilation in physics-based hydrological models. Such an assumption implies that controls of GW on land-surface
interactions, and consequently hydro-climatic hazards, may be
overlooked (Keune et al., 2016; Maxwell and Condon, 2016).
Bridging the gaps between spatial Subsurface flow can also be crucial for a series of applications
scales including (and not limited to) drought monitoring, flood
predictions, and water resources management. Only recently,
In the following, we discuss recent developments, the land surface community started to recognize that lateral
outstanding challenges, and possible research directions in groundwater flow significantly interacts with surface processes
physics-based hydrological modeling (Section Recent advances such as vegetation dynamics (Gochis et al., 2018; Zeng et al.,
and outstanding challenges in physically based hydrological 2018; Forrester and Maxwell, 2020; Rummler et al., 2022),
modeling), relevant observations (Section Recent advances and hydro-climatic hazards (Felsberg et al., 2021), and atmospheric
outstanding challenges in observation data), and assimilation forecasting (Getirana et al., 2020) and consequently to put efforts
systems (Section Data assimilation for non-gaussian systems). in representing this process (e.g., Batelis et al., 2020).
Third, although the number of high-resolution datasets
such as topography, land use, geology, and soil properties
Recent advances and outstanding are increasing, these data are also the products of remote
challenges in physically based sensing, land surface models, statistical downscaling techniques,
hydrological modeling or combinations thereof and thus affected by inaccuracies
and uncertainties.
In many cases, the breadth of differences in the data Finally, executing physically based models at high spatial
generated from in-situ and remote sensing observations raises resolution, while resolving surface and subsurface water
significant questions on how to best use these data in modeling dynamics in a coupled way and over large regions is
development and testing. Land surface and groundwater have computationally demanding, both in terms of CPU time and
intertwined processes, yet their scientific modeling fields are storage requirements (Kollet et al., 2010; Maxwell, 2013).
still disconnected. These challenges suggest the need to develop new modeling
Over the past decade, the LSM/GHM modeling community approaches at continental and global scales that can properly
has been embracing hyper-resolution (∼1-km or finer) simulate hydrological processes (especially lateral subsurface
integrated surface-subsurface hydrological modeling (Wood flows) at resolutions compatible with remote sensing data
et al., 2011; Bierkens et al., 2015; Condon et al., 2021), (i.e., resolutions used in the LSM/GHM community). To
therefore providing the opportunity for the two modeling this end, machine learning and deep learning techniques are
categories (ISSHMs and LSM/GHMs) to start meeting in the recently being explored to emulate complex subsurface physical
middle while resolving regional to continental scale surface- processes (Radmanesh et al., 2020; Tran et al., 2021) and also to
subsurface hydrological processes. Moving LSM/GHMs to link modeled estimates with indirect measurements of the state
higher resolutions and/or ISSHMs to larger scales is posing huge variables (e.g., to link moisture states to radiances observations,
challenges, mainly represented by uncertainties due to the lack Section Earth observations).

Frontiers in Water 03 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

Recent advances and outstanding aggregated observation of multiple water storage components.
challenges in observation data The assimilation of GRACE data in an LSM allows us to
downscale the TWS observations spatially and temporally into
Earth observations its various water storage components such as surface and root
Earth observation data (aka satellite or remote sensing) is zone soil moisture and groundwater table (Girotto et al., 2016;
essential to monitor hydrologic variables such as soil moisture Kumar et al., 2016).
and groundwater table because it provides a bird’s-eye view of Another interesting opportunity is given by processes
the Earth’s dynamics. In the last several decades, the number linking changes in GW level to deformation of the Earth’s crust
of spaceborne sensors has rapidly increased. These observing and thus vertical land motion (Erban et al., 2014; Darvishi
systems can detect subsurface hydrologic states such as the et al., 2021). As opposed to tectonic and sediment compaction,
soil moisture at various spatial and temporal resolutions (e.g., the vertical land motion associated with surface water or
from ∼1 to 40 km resolutions, Entekhabi et al., 2010; Fang groundwater table changes can be visible at sub-annual temporal
et al., 2021). Despite this apparent wealth of new data, there scales. Some GW systems are more susceptible to compaction
are lingering methodological challenges to translate these types compared to others. In some cases, depletion and recharge of
of observations into water table depth. In fact, the basic aquifers can cause vertical land motion with rates up to ∼10
remote sensing principle involves the direct interaction between cm/year (Carlson et al., 2020). Elastic deformation typically
an incident radiation (e.g., natural solar radiation or active results in small-magnitude recoverable displacements (on the
radiation sent from a sensor) and the targets of interest. mm to cm scale) of the land surface (Shirzaei et al., 2020). When
Groundwater is a hidden resource and therefore cannot be the effective stress exceeds a stress threshold, the deformation
directly observed and measured from space. Ground-based is inelastic and results in permanent (irreversible) subsidence.
water table depth monitoring and reporting are scarce and Recent studies have quantified the elastic contributions to
variable across the globe, and frequently limited to developed vertical land motion and related it to GW (Chaussard and
regions. For this reason, little is still known about the global Farr, 2019; Smith and Knight, 2019; Hsu et al., 2020; Lu
patterns of the water table depth and its interactions with the et al., 2020; Ojha et al., 2020; Tangdamrongsub and Šprlák,
land, urban and natural ecosystems, hindering our capabilities 2021). Deriving accurate global soil compaction parameters and
to fully exploit the potential of water table information in data modules to be coupled to hydrological models can be complex,
assimilation for global-scale integrated hydrological modeling. if not impossible. For data assimilation, this limitation can be
Possible ways forward are offered by the fact that addressed with the development of an artificial intelligence (i.e.,
groundwater can be indirectly observed through other neural network) forward model to link vertical deformation
quantities, for example evaporation fluxes (Miralles et al., to changes in the variable of interest (i.e., groundwater table)
2011) and total terrestrial water storage (TWS) (Famiglietti and (Smith and Majumdar, 2020; Naghibi et al., 2022).
Rodell, 2013). In addition to improved estimates of the driving forces,
Water table dynamics are linked to the atmosphere as detailed above, an accurate, detailed and spatio-temporal
through evapotranspiration and recharge. From a water assessment of groundwater fluxes requires also a reliable
budget perspective, the accurate account of water losses estimate of soil and aquifer parameters, to avoid errors in
via evapotranspiration and runoff should lead to improved parameters leading to over- or underestimation of fluxes.
estimates of recharge fluxes, and thus water table depth. Starting from prior parameter distributions that can be extracted
Recent studies have shown that assimilation of remotely sensed from globally available datasets (e.g., Batjes, 1997; Hengl et al.,
evapotranspiration estimates can inform the groundwater table 2017), data assimilation allows for their dynamic update, which
dynamics (Gelsinari et al., 2020). can lead to improved model predictions (e.g., Hung et al., 2022).
Since 2002, TWS estimates have been derived from the
Gravity Recovery and Climate Experiment (GRACE) and
its follow-on mission (GRACE-FO). These missions provide Intermediate-Scale sensing technologies
valuable information on groundwater conditions beyond what One possible way forward to expand the capabilities of DA
can be seen at the surface (Li et al., 2012; Scanlon et al., for integrated hydrological modeling is offered by novel sensing
2018) because it measures the water storage changes in the technologies at intermediate scales that go beyond the local
entire terrestrial water storage (i.e., including snow, vegetation support scale typical of ground observations. While traditional
biomass, surface soil moisture, root-zone soil moisture, and soil moisture, water table, and pressure head observations have
GW). However, limitations still exist for using GRACE data a support volume of a few cubic centimeters or decimeters,
in operational groundwater dynamics monitoring. The major fiber optics sensing (FOS), airborne electro-magnetic methods
limitations are related to (i) the data delivery latency; (ii) (AEMs), cosmic ray neutron sensing (CRNS), and unmanned
the relatively coarse spatial and temporal resolutions of aerial vehicles (UAVs) represent relatively cheap (compared
GRACE observations; and (iii) the fact that TWS is an to the possible extent of the surveys) options to collect

Frontiers in Water 04 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

FIGURE 1
Data assimilation combines the strengths of modeled and observed hydrological estimates. Integrated surface-subsurface hydrological models
(ISSHMs) typically operate within spatial scales ranging from hillslope to local-scale or meso-scale catchments and its data assimilation
techniques primarily incorporate local ground-based observations. Land Surface (LSMs) and Global Hydrological models (GHMs) operate within
continental and global spatial scales and their data assimilation applications utilize remotely sensed observations. This mini review discusses
possible new directions for data assimilation technologies which could potentially bridge the scale gaps between the ISSHM and LSM/GHM
scientific communities.

hydrologically relevant data at scales ranging from tens of meters Vittecoq et al., 2019). So far, AEM surveys have been typically
to hundreds of kilometers. conducted to get a static picture of the subsurface structure and
Fiber optic sensing can be used to measure, among other parameters. However, if repeated over time (despite their costs),
variables, strain, displacement, pressure, and temperature with these surveys might provide valuable datasets to be used in data
high spatial resolution for distances of several kilometers (Leone, assimilation studies for integrated hydrological modeling, such
2022). Fiber optic distributed temperature sensing (FO-DTS), as variations of groundwater levels over large areas, which could
in particular, represents a promising technique to measure complement information provided by GRACE.
soil water at high spatial resolutions (<1 m), thanks to the Cosmic ray neutron probes are used in several fields of
relationship between soil thermal properties and soil moisture hydrology as they offer the possibility of estimating average
content (Sayde et al., 2010, 2014; Steele-Dunne et al., 2010; snow water equivalent (Schattan et al., 2017) and soil moisture
Striegl and Loheide, 2012; Dong et al., 2017; Vidana Gamage over areas of 130–240 m of radius and to depths of ∼15–
et al., 2018). However, to the best of our knowledge, FOS data 80 cm, depending on the soil moisture itself (Köhli et al.,
have never been used for DA studies in integrated hydrological 2015). Using an Ensemble Adjustment Kalman Filter, Patil et al.
modeling and still represent an underexploited opportunity. (2021) assimilated neutron counts for a 655 km2 catchment into
Much progress has been made in the past decade the Noah-MP land surface model, showing that incorporating
to assimilate hydrogeophysical observations in hydrological information from CRNS can improve parameter and soil
models. In particular, Electrical Resistivity (ER) data, thanks to moisture estimates and paving the way for future applications
their sensitivity to soil moisture and salinity, have demonstrated also in ISSHMs.
their usefulness in parameterizing subsurface systems (e.g., The increasing use of UAVs in all science disciplines,
Camporese et al., 2015; Crestani et al., 2015; Manoli et al., 2015; whereby multiple types of sensors can be easily deployed
Tso et al., 2020). However, all these studies have focused on local- over relatively large areas, also provides us with plenty of
scale systems, while hydrogeophysics now offers the opportunity opportunities to assimilate high-resolution data at intermediate
to survey large extensions of land with AEM methods. They scales. Assimilation of UAV-collected data is becoming common
potentially allows for the estimation of key groundwater in weather modeling (Sun et al., 2020; Jensen et al., 2021) and
parameters, such as hydraulic conductivity, with remotely crop modeling (Yu et al., 2019; Peng et al., 2021). However,
sensed geophysical data collected over thousands of square incorporation of UAV data in physics-based hydrological
kilometers, while still being able to capture heterogeneities modeling is still largely unexplored. Tang et al. (2018) used
at scales of interest for physics-based hydrological models river bathymetry data collected through UAV surveys, in
(∼101 m) (e.g., Christensen et al., 2017; Knight et al., 2018; conjunction with measurements of hydraulic head assimilated

Frontiers in Water 05 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

via EnKF, to improve model predictions and parameters with Compared to traditional EnKF implementations, DA
HydroGeoSphere (Brunner and Simmons, 2012), an ISSHM, for algorithms based on PF or transport methods require additional
a river catchment in Switzerland. Given the capabilities of UAV- computational effort. However, due to the increased availability
mounted sensors to measure a wide range of hydrological (e.g., of ever more powerful computers, we expect this is going to be
soil moisture) and vegetation (e.g., leaf area index) variables, we less of an issue in the near future.
see much room for DA studies in physics-based hydrological
modeling in the next few years.
Conclusions
Physics-based integrated hydrological modeling have been
Data assimilation for non-gaussian declined in different ways by two scientific communities. On
systems one hand, integrated surface-subsurface hydrological models
(ISSHMs), solving flow and transport processes in detail,
A common approach in data assimilation is the ensemble are typically applied to local to meso-scale catchments. On
Kalman filter (EnKF; Evensen, 2003) that dynamically and the other hand, land surface and global hydrological models
sequentially updates model error covariance information by (LSMs/GHMs) focus on continental to global scale applications,
producing an ensemble of model predictions, which are paying more attention to hydrological processes at the land-
individual model realizations perturbed by the assumed model atmosphere interface and representing subsurface processes
error (Reichle et al., 2002). EnKF systems have been targeted with simplified approaches. Accordingly, data assimilation has
by both ISSHMs (e.g., Camporese et al., 2009b; Pasetto et al., long been used by the two communities in different ways,
2012) and LSM/GHMs (de Rosnay et al., 2014; Kumar et al., exploiting data types consistent with their respective scales, i.e.,
2016; Girotto et al., 2019). Some common challenges include ground-based measurements for ISSHMs and remote sensing
the need of spatially localizing EnKF covariances to remove observations for LSMs/GHMs.
spurious correlations, and most importantly its assumption of In this mini review, we advocate for the two communities
multi-gaussianity, which must be met for both system states to find common ground for advancing data assimilation
and observations. If this assumption is not satisfied, the EnKF at intermediate scales (Figure 1). This can be done by
only gives sub-optimal results or, in the worst cases, fails to pursuing new hydrological modeling approaches that can result
give an adequate representation of the variables of interest. in less computational requirements for ISSHMs and more
Other, arguably more sophisticated methods, include particle accurate process representation for LSMs/GHMs. Also, we gave
filter (PF) techniques (e.g., Crisan, 2001) or combinations of an overview of opportunities for exploiting state-of-the-art
EnKF and PF (e.g., Stordal et al., 2011). Like the EnKF, the PF technologies providing data types at intermediate scales that
is a sequential Monte Carlo approach, but it does not depend can be of interest for assimilation in both categories of models.
on the assumption of Gaussian distributions, in principle being Finally, as non-gaussianity is a common issue in hydrological
able to accommodate every probability distribution functions, models and makes the most used DA algorithms sub-optimal, we
including multimodal ones. However, PF techniques typically suggest exploring alternative and innovative DA methods that
require larger ensembles to characterize the full probability can lead to a more plausible representation of parameter and
distribution of the state variables and consequently their process spatial variability.
uncertainties via resampling sets of state variables. Due to Ultimately, these developments can contribute to bridge
this disadvantage, PF applications in physics-based hydrological the gap between the two DA and modeling communities, with
models have been limited to a few cases (Pasetto et al., 2012; obvious benefits for both.
Zhang et al., 2017).
Interesting new developments for non-Gaussian DA Author contributions
applications in physics-based hydrological modeling could
derive from applications of the transport theory (Villani, All authors listed have made a substantial, direct,
2009). Ning et al., 2014 showed how transportation metrics and intellectual contribution to the work and approved it
in the framework of optimal mass transport can be used with for publication.
variational data assimilation to obtain optimal estimates of
the system state in the presence of systematic model errors or
bias, for which no prior knowledge is available. More recently, Conflict of interest
Ramgraber et al. (2021) introduced their work on non-linear
smoothing, whereby transport methods are applied as a pathway The authors declare that the research was conducted in the
for non-linear/non-Gaussian generalization of two different absence of any commercial or financial relationships that could
ensemble Kalman-type smoothers. be construed as a potential conflict of interest.

Frontiers in Water 06 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

Publisher’s note organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
All claims expressed in this article are solely those of the claim that may be made by its manufacturer, is not guaranteed
authors and do not necessarily represent those of their affiliated or endorsed by the publisher.

References
Balica, S. F., Wright, N. G., and Van der Meulen, F. (2012). A flood vulnerability Condon, L. E., Markovich, K. H., Kelleher, C. A., McDonnell, J. J., Ferguson, G.,
index for coastal cities and its use in assessing climate change impacts. Nat. Haz. and McIntosh, J. C. (2020b). Where is the bottom of a watershed? Water Resourc.
64, 73–105. doi: 10.1007/s11069-012-0234-1 Res. 56, e2019WR026010. doi: 10.1029/2019WR026010
Batelis, S.-C., Rahman, M., Kollet, S., Woods, R., and Rosolem, R. (2020). Condon, L. E., and Maxwell, R. M. (2019). Simulating the sensitivity of
Towards the representation of groundwater in the Joint UK land environment evapotranspiration and streamflow to large-scale groundwater depletion. Sci. Adv.
simulator. Hydrol. Process. 34, 2843–2863. doi: 10.1002/hyp.13767 5, eaav4574. doi: 10.1126/sciadv.aav4574
Batjes, N. H. (1997). A world dataset of derived soil properties by Crestani, E., Camporese, M., and Salandin, P. (2015). Assessment of
FAO–UNESCO soil unit for global modelling. Soil Use Manage. 13, 9–16. hydraulic conductivity distributions through assimilation of travel time
doi: 10.1111/j.1475-2743.1997.tb00550.x data from ERT-monitored tracer tests. Adv. Water Resour. 84, 23–36.
doi: 10.1016/j.advwatres.2015.07.022
Beven, K. J., and Cloke, H. L. (2012). Comment on: Hyperresolution
global land surface modeling: meeting a grand challenge for monitoring Crisan, D. (2001). “Particle Filters - A Theoretical Perspective,” in Sequential
Earth’s terrestrial water by Eric F Wood et al. Water Resour. Res. 48, 1801. Monte Carlo Methods in Practice, eds A. Doucet, N. de Freitas and Gordon, N. (New
doi: 10.1029/2011WR010982 York, NY: Statistics for Engineering and Information Science). Available online at:
https://doi.org/10.1007/978-1-4757-3437-9_2
Bierkens, M. F., Bell, V. A., Burek, P., Chaney, N., Condon, L. E., David,
C. H., et al. (2015). Hyper-resolution global hydrological modelling: what Darvishi, M., Destouni, G., Aminjafari, S., and Jaramillo, F. (2021). Multi-Sensor
is next? “Everywhere and locally relevant.” Hydrol. Process. 29, 310–320. InSAR assessment of ground deformations around lake mead and its relation to
doi: 10.1002/hyp.10391 water level changes. Remote Sens. 13, 406. doi: 10.3390/rs13030406
Botto, A., Belluco, E., and Camporese, M. (2018). Multi-source data assimilation de Rosnay, P., Balsamo, G., Albergel, C., Muñoz-Sabater, J., and Isaksen, L.
for physically based hydrological modeling of an experimental hillslope. Hydrol. (2014). Initialisation of land surface variables for numerical weather prediction.
Earth Syst. Sci. 22, 4251–4266. doi: 10.5194/hess-22-4251-2018 Surv. Geophys. 35, 607–621. doi: 10.1007/s10712-012-9207-x
Brunner, P., and Simmons, C. T. (2012). HydroGeoSphere: a fully Dong, J., Agliata, R., Steele-Dunne, S., Hoes, O., Bogaard, T., Greco, R., et al.
integrated, physically based hydrological model. Ground Water 50, 170–176. (2017). The impacts of heating strategy on soil moisture estimation using actively
doi: 10.1111/j.1745-6584.2011.00882.x heated fiber optics. Sensors 17, 2102. doi: 10.3390/s17092102
Camporese, M., Cassiani, G., Deiana, R., Salandin, P., and Binley, A. (2015). Elalem, S., and Pal, I. (2015). Mapping the vulnerability hotspots over Hindu-
Coupled and uncoupled hydrogeophysical inversions using ensemble K alman Kush Himalaya region to flooding disasters. Weather Clim. Extrem. 8, 46–58.
filter assimilation of ERT-monitored tracer test data. Water Resour. Res. 51, doi: 10.1016/j.wace.2014.12.001
3277–3291. doi: 10.1002/2014WR016017
Entekhabi, D., Njoku, E. G., O’Neill, P. E., Kellogg, K. H., Crow, W. T., Edelstein,
Camporese, M., Paniconi, C., Putti, M., and Salandin, P. (2009a). Comparison of W. N., et al. (2010). The soil moisture active passive (SMAP) mission. Proc. IEEE
data assimilation techniques for a coupled model of surface and subsurface flow. 98, 704–716. doi: 10.1109/JPROC.2010.2043918
Vadose Zone J. 8, 837–845. doi: 10.2136/vzj2009.0018
Erban, L. E., Gorelick, S. M., and Zebker, H. A. (2014). Groundwater extraction,
Camporese, M., Paniconi, C., Putti, M., and Salandin, P. (2009b). Ensemble land subsidence, and sea-level rise in the Mekong Delta, Vietnam. Environ. Res.
Kalman filter data assimilation for a process-based catchment scale model Lett. 9, 084010. doi: 10.1088/1748-9326/9/8/084010
of surface and subsurface flow. Water Resour. Res. 45, e2008WR007031.
Evensen, G. (2003). The ensemble Kalman filter: theoretical
doi: 10.1029/2008WR007031
formulation and practical implementation. Ocean Dyn. 53, 343–367.
Carlson, G., Shirzaei, M., Werth, S., Zhai, G., and Ojha, C. (2020). Seasonal and doi: 10.1007/s10236-003-0036-9
long-term groundwater unloading in the central valley modifies crustal stress. J.
Evensen, G., Vossepoel, F. C., and van Leeuwen, P. J. (2022). Data Assimilation
Geophys. Res. Solid Earth 125, e2019JB018490. doi: 10.1029/2019JB018490
Fundamentals: A Unified Formulation of the State and Parameter Estimation
Chaussard, E., and Farr, T. G. (2019). A new method for isolating elastic from Problem. Springer Nature. doi: 10.1007/978-3-030-96709-3 Available online at:
inelastic deformation in aquifer systems: application to the San Joaquin Valley, CA. https://link.springer.com/book/10.1007/978-3-030-96709-3
Geophys. Res. Lett. 46, 10800–10809. doi: 10.1029/2019GL084418
Famiglietti, J. S., and Rodell, M. (2013). Water in the balance. Science 340,
Christensen, N. K., Ferre, T. P. A., Fiandaca, G., and Christensen, S. (2017). 1300–1301. doi: 10.1126/science.1236460
Voxel inversion of airborne electromagnetic data for improved groundwater model
Fang, B., Lakshmi, V., Cosh, M. H., and Hain, C. (2021). Very high spatial
construction and prediction accuracy. Hydrol. Earth Syst. Sci. 21, 1321–1337.
resolution downscaled SMAP radiometer soil moisture in the CONUS using
doi: 10.5194/hess-21-1321-2017
VIIRS/MODIS data. IEEE J. Select. Top. Appl. Earth Observ. Remote Sens. 14,
Clark, M. P., Fan, Y., Lawrence, D. M., Adam, J. C., Bolster, D., Gochis, D. J., 4946–4965. doi: 10.1109/JSTARS.2021.3076026
et al. (2015). Improving the representation of hydrologic processes in earth system
Fatichi, S., Vivoni, E. R., Ogden, F. L., Ivanov, V. Y., Mirus, B., Gochis,
models. Water Resour. Res. 51, 5929–5956. doi: 10.1002/2015WR017096
D., et al. (2016). An overview of current applications, challenges, and future
Collier, N., Hoffman, F. M., Lawrence, D. M., Keppel-Aleks, G., Koven, C. trends in distributed process-based models in hydrology. J. Hydrol. 537, 45–60.
D., Riley, W. J., et al. (2018). The international land model benchmarking doi: 10.1016/j.jhydrol.2016.03.026
(ILAMB) system: design, theory, and implementation. J. Adv. Model. Earth Syst.
Felsberg, A., De Lannoy, G. J., Girotto, M., Poesen, J., Reichle, R. H., and Stanley,
10, 2731–2754. doi: 10.1029/2018MS001354
T. (2021). Global soil water estimates as landslide predictor: the effectiveness
Condon, L. E., Atchley, A. L., and Maxwell, R. M. (2020a). of SMOS, SMAP and GRACE observations, land surface simulations and data
Evapotranspiration depletes groundwater under warming over the assimilation. J. Hydrometeorol. 22, 1065–1084. doi: 10.1175/JHM-D-20-0228.1
contiguous United States. Nat. Commun. 11, 873. doi: 10.1038/s41467-020-14
Forrester, M. M., and Maxwell, R. M. (2020). Impact of lateral groundwater
688-0
flow and subsurface lower boundary conditions on atmospheric boundary
Condon, L. E., Kollet, S., Bierkens, M. F., Fogg, G. E., Maxwell, R. M., Hill, layer development over complex terrain. J. Hydrometeorol. 21, 1133–1160.
M. C., et al. (2021). Global groundwater modeling and monitoring: opportunities doi: 10.1175/JHM-D-19-0029.1
and challenges. Water Resourc. Res. 57, e2020WR029500. doi: 10.1029/2020WR0
Furman, A. (2008). Modeling coupled surface–subsurface flow processes: a
29500
review. Vadose Zone J. 7, 741–756. doi: 10.2136/vzj2007.0065

Frontiers in Water 07 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

Gebler, S., Kurtz, W., Pauwels, V. R. N., Kollet, S. J., Vereecken, H., and Kumar, S. V., Jasinski, M., Mocko, D. M., Rodell, M., Borak, J., Li, B., et al.
Hendricks Franssen, H.-J. (2019). Assimilation of high-resolution soil moisture (2019). NCA-LDAS land analysis: development and performance of a multisensor,
data into an integrated terrestrial model for a small-scale head-water catchment. multivariate land data assimilation system for the national climate assessment. J.
Water Resour. Res. 55, 10358–10385. doi: 10.1029/2018WR024658 Hydrometeorol. 20, 1571–1593. doi: 10.1175/JHM-D-17-0125.1
Gelsinari, S., Doble, R., Daly, E., and Pauwels, V. R. (2020). Feasibility of Kumar, S. V., Zaitchik, B. F., Peters-Lidard, C. D., Rodell, M., Reichle, R.,
improving groundwater modeling by assimilating evapotranspiration rates. Water Li, B., et al. (2016). Assimilation of gridded GRACE terrestrial water storage
Resourc. Res. 56, e2019WR025983. doi: 10.1029/2019WR025983 estimates in the North American land data assimilation system. J. Hydrometeorol.
17, 1951–1972. doi: 10.1175/JHM-D-15-0157.1
Getirana, A., Rodell, M., Kumar, S., Beaudoing, H. K., Arsenault, K., Zaitchik, B.,
et al. (2020). GRACE improves seasonal groundwater forecast initialization over Leone, M. (2022). Advances in fiber optic sensors for soil moisture monitoring:
the United States. J. Hydrometeorol. 21, 59–71. doi: 10.1175/JHM-D-19-0096.1 a review. Results Optics 7, 100213. doi: 10.1016/j.rio.2022.100213
Girotto, M., De Lannoy, G. J., Reichle, R. H., and Rodell, M. (2016). Assimilation Li, B., Rodell, M., Zaitchik, B. F., Reichle, R. H., Koster, R. D., and van Dam, T. M.
of gridded terrestrial water storage observations from GRACE into a land surface (2012). Assimilation of GRACE terrestrial water storage into a land surface model:
model. Water Resour. Res. 52, 4164–4183. doi: 10.1002/2015WR018417 evaluation and potential value for drought monitoring in western and central
Europe. J. Hydrol. 446, 103–115. doi: 10.1016/j.jhydrol.2012.04.035
Girotto, M., Reichle, R., Rodell, M., and Maggioni, V. (2021). Data assimilation
of terrestrial water storage observations to estimate precipitation fluxes: a synthetic Liu, Y., and Margulis, S. A. (2019). Deriving Bias and uncertainty in MERRA-
experiment. Remote Sens. 13, 1223. doi: 10.3390/rs13061223 2 snowfall precipitation over high mountain Asia. Front. Earth Sci. 7, 280.
doi: 10.3389/feart.2019.00280
Girotto, M., Reichle, R. H., Rodell, M., Liu, Q., Mahanama, S., and De Lannoy, G.
J. (2019). Multi-sensor assimilation of SMOS brightness temperature and GRACE Lu, C.-Y., Hu, J.-C., Chan, Y.-C., Su, Y.-F., and Chang, C.-H. (2020). The
terrestrial water storage observations for soil moisture and shallow groundwater relationship between surface displacement and groundwater level change and its
estimation. Remote Sens. Environ. 227, 12–27. doi: 10.1016/j.rse.2019.04.001 hydrogeological implications in an alluvial fan: case study of the Choshui River,
Taiwan. Remote Sens. 12, 3315. doi: 10.3390/rs12203315
Gochis, D. J., Barlage, M., Dugger, A., FitzGerald, K., Karsten, L., McAllister, M.,
et al. (2018). The WRF-Hydro modeling System Technical Description,(Version 5.0). Manoli, G., Rossi, M., Pasetto, D., Deiana, R., Ferraris, S., Cassiani, G.,
NCAR Technical Note 107. et al. (2015). An iterative particle filter approach for coupled hydro-geophysical
inversion of a controlled infiltration experiment. J. Comput. Phys. 283, 37–51.
He, X., Lucatero, D., Ridler, M.-E., Madsen, H., Kidmose, J., Hole, Ø.,
doi: 10.1016/j.jcp.2014.11.035
et al. (2019). Real-time simulation of surface water and groundwater with data
assimilation. Adv. Water Resour. 127, 13–25. doi: 10.1016/j.advwatres.2019.03.004 Maxwell, R. M. (2013). A terrain-following grid transform and preconditioner
for parallel, large-scale, integrated hydrologic modeling. Adv. Water Resour. 53,
Hengl, T., Mendes de Jesus, J., Heuvelink, G. B., Ruiperez
109–117. doi: 10.1016/j.advwatres.2012.10.001
Gonzalez, M., Kilibarda, M., Blagoti,ć, A., et al. (2017).
SoilGrids250m: global gridded soil information based on machine Maxwell, R. M., and Condon, L. E. (2016). Connections between
learning. PLoS ONE 12, e0169748. doi: 10.1371/journal.pone.016 groundwater flow and transpiration partitioning. Science 353, 377–380.
9748 doi: 10.1126/science.aaf7891
Hsu, Y.-J., Fu, Y., Bürgmann, R., Hsu, S.-Y., Lin, C.-C., Tang, C.-H., et al. Maxwell, R. M., Putti, M., Meyerhoff, S., Delfs, J.-O., Ferguson, I. M., Ivanov, V.,
(2020). Assessing seasonal and interannual water storage variations in Taiwan et al. (2014). Surface-subsurface model intercomparison: a first set of benchmark
using geodetic and hydrological data. Earth Planet. Sci. Lett. 550, 116532. results to diagnose integrated hydrology and feedbacks. Water Resour. Res. 50,
doi: 10.1016/j.epsl.2020.116532 1531–1549. doi: 10.1002/2013WR013725
Hung, P. C., Schalge, B., Baroni, G., Vereecken, H., and Hendricks Franssen, H. Miralles, D. G., Holmes, T. R. H., De Jeu, R. A. M., Gash, J. H.,
(2022). Assimilation of groundwater level and soil moisture data in an integrated Meesters, A., and Dolman, A. J. (2011). Global land-surface evaporation
land surface-subsurface model for southwestern Germany. Water Resourc. Res. 58, estimated from satellite-based observations. Hydrol. Earth Syst. Sci. 15, 453–469.
e2021WR031549. doi: 10.1029/2021WR031549 doi: 10.5194/hess-15-453-2011
Jensen, A. A., Pinto, J. O., Bailey, S. C., Sobash, R. A., de Boer, G., Houston, A. Naghibi, S. A., Khodaei, B., and Hashemi, H. (2022). An integrated InSAR-
L., et al. (2021). Assimilation of a coordinated fleet of uncrewed aircraft system machine learning approach for ground deformation rate modeling in arid areas.
observations in complex terrain: EnKF system design and preliminary assessment. J. Hydrol. 608, 127627. doi: 10.1016/j.jhydrol.2022.127627
Mthly. Weather Rev. 149, 1459–1480. doi: 10.1175/MWR-D-20-0359.1
Navari, M., Margulis, S. A., Tedesco, M., Fettweis, X., and Alexander, P.
Keune, J., Gasper, F., Goergen, K., Hense, A., Shrestha, P., Sulis, M., et al. (2016). M. (2018). Improving greenland surface mass balance estimates through the
Studying the influence of groundwater representations on land surface-atmosphere assimilation of MODIS albedo: a case study along the K-transect. Geophys. Res.
feedbacks during the European heat wave in 2003. J. Geophys. Res. Atmosph. 121, Lett. 45, 6549–6556. doi: 10.1029/2018GL078448
13301–13325. doi: 10.1002/2016JD025426
Ning, L., Carli, F. P., Ebtehaj, A. M., Foufoula-Georgiou, E., and Georgiou, T.
Keune, J., Sulis, M., Kollet, S., Siebert, S., and Wada, Y. (2018). Human water use T. (2014). Coping with model error in variational data assimilation using optimal
impacts on the strength of the continental sink for atmospheric water. Geophys. mass transport. Water Resour. Res. 50, 5817–5830. doi: 10.1002/2013WR014966
Res. Lett. 45, 4068–4076. doi: 10.1029/2018GL077621
Ojha, C., Werth, S., and Shirzaei, M. (2020). Recovery of aquifer-systems in
Khaki, M., Hendricks Franssen, H.-J., and Han, S. C. (2020). Multi-mission Southwest US following 2012–2015 drought: evidence from InSAR, GRACE and
satellite remote sensing data for improving land hydrological models via data groundwater level data. J. Hydrol. 587, 124943. doi: 10.1016/j.jhydrol.2020.124943
assimilation. Sci. Rep. 10, 18791. doi: 10.1038/s41598-020-75710-5
Or, D. (2020). The tyranny of small scales—on representing soil processes
Knight, R., Smith, R., Asch, T., Abraham, J., Cannia, J., Viezzoli, A., et al. (2018). in global land surface models. Water Resour. Res. 56, e2019WR024846.
Mapping aquifer systems with airborne electromagnetics in the Central Valley of doi: 10.1029/2019WR024846
California. Groundwater 56, 893–908. doi: 10.1111/gwat.12656
Or, D., Lehmann, P., and Assouline, S. (2015). Natural length scales define the
Köhli, M., Schrön, M., Zreda, M., Schmidt, U., Dietrich, P., and Zacharias, range of applicability of the R ichards equation for capillary flows. Water Resour.
S. (2015). Footprint characteristics revised for field-scale soil moisture Res. 51, 7130–7144. doi: 10.1002/2015WR017034
monitoring with cosmic-ray neutrons. Water Resour. Res. 51, 5772–5790.
Paniconi, C., Marrocu, M., Putti, M., and Verbunt, M. (2003). Newtonian
doi: 10.1002/2015WR017169
nudging for a richards equation-based distributed hydrological model. Adv. Water
Kollet, S., Gasper, F., Brdar, S., Goergen, K., Hendricks-Franssen, H.-J., Keune, Resour. 26, 161–178. doi: 10.1016/S0309-1708(02)00099-4
J., et al. (2018). Introduction of an experimental terrestrial forecasting/monitoring
Paniconi, C., and Putti, M. (2015). Physically based modeling in catchment
system at regional to continental scales based on the terrestrial systems modeling
hydrology at 50: survey and outlook. Water Resour. Res. 51, 7090–7129.
platform (v1. 1.0). Water 10, 1697. doi: 10.3390/w10111697
doi: 10.1002/2015WR017780
Kollet, S. J., Maxwell, R. M., Woodward, C. S., Smith, S., Vanderborght,
Pasetto, D., Camporese, M., and Putti, M. (2012). Ensemble Kalman filter versus
J., Vereecken, H., et al. (2010). Proof of concept of regional scale hydrologic
particle filter for a physically-based coupled surface–subsurface model. Adv. Water
simulations at hydrologic resolution utilizing massively parallel computer
Resour. 47, 1–13. doi: 10.1016/j.advwatres.2012.06.009
resources. Water Resour. Res. 46, e2009WR008730. doi: 10.1029/2009WR008730
Patil, A., Fersch, B., Hendricks Franssen, H.-J., and Kunstmann, H. (2021).
Kumar, S. V., Dong, J., Peters-Lidard, C. D., Mocko, D., and Gómez,
Assimilation of cosmogenic neutron counts for improved soil moisture
B. (2017). Role of forcing uncertainty and background model error
prediction in a distributed land surface model. Front. Water 115, 729592.
characterization in snow data assimilation. Hydrol. Earth Syst. Sci. 21, 2637–2647.
doi: 10.3389/frwa.2021.729592
doi: 10.5194/hess-21-2637-2017

Frontiers in Water 08 frontiersin.org

Camporese and Girotto 10.3389/frwa.2022.948832

Peng, X., Han, W., Ao, J., and Wang, Y. (2021). Assimilation of LAI derived from Stordal, A. S., Karlsen, H. A., Nævdal, G., Skaug, H. J., and Vallès, B. (2011).
UAV multispectral data into the SAFY model to estimate maize yield. Remote Sens. Bridging the ensemble Kalman filter and particle filters: the adaptive Gaussian
13, 1094. doi: 10.3390/rs13061094 mixture filter. Comput. Geosci. 15, 293–305. doi: 10.1007/s10596-010-9207-1
Radmanesh, F., Golabi, M. R., Khodabakhshi, F., Farzi, S., and Zeinali, M. (2020). Striegl, A. M., and Loheide, S. P. II. (2012). Heated distributed temperature
Modeling aquifer hydrograph: performance review of conceptual MODFLOW and sensing for field scale soil moisture monitoring. Groundwater 50, 340–347.
simulator models. Arab. J. Geosci. 13, 1–9. doi: 10.1007/s12517-020-5230-2 doi: 10.1111/j.1745-6584.2012.00928.x
Ramgraber, M., Weatherl, R., Blumensaat, F., and Schirmer, M. (2021). Sun, Q., Vihma, T., Jonassen, M. O., and Zhang, Z. (2020). Impact of assimilation
Non-Gaussian parameter inference for hydrogeological models using stein of radiosonde and UAV observations from the Southern Ocean in the polar WRF
variational gradient descent. Water Resour. Res. 57, e2020WR029339. model. Adv. Atmosph. Sci. 37, 441–454. doi: 10.1007/s00376-020-9213-8
doi: 10.1029/2020WR029339
Tang, Q., Schilling, O. S., Kurtz, W., Brunner, P., Vereecken, H., and
Rasmussen, J., Madsen, H., Jensen, K. H., and Refsgaard, J. C. (2015). Data Hendricks Franssen, H.-J. (2018). Simulating flood-induced riverbed
assimilation in integrated hydrological modeling using ensemble Kalman filtering: transience using unmanned aerial vehicles, physically based hydrological
evaluating the effect of ensemble size and localization on filter performance. modeling, and the ensemble kalman filter. Water Resour. Res. 54, 9342–9363.
Hydrol. Earth Syst. Sci. 19, 2999–3013. doi: 10.5194/hess-19-2999-2015 doi: 10.1029/2018WR023067
Rasmussen, J., Madsen, H., Jensen, K. H., and Refsgaard, J. C. (2016). Data Tangdamrongsub, N., and Šprlák, M. (2021). The assessment of hydrologic-and
assimilation in integrated hydrological modelling in the presence of observation flood-induced land deformation in data-sparse regions using GRACE/GRACE-FO
bias. Hydrol. Earth Syst. Sci. 20, 2103–2118. doi: 10.5194/hess-20-2103-2016 data assimilation. Remote Sens. 13, 235. doi: 10.3390/rs13020235
Reichle, R. H., McLaughlin, D. B., and Entekhabi, D. (2002). Hydrologic data Tran, H., Leonarduzzi, E., De la Fuente, L., Hull, R. B., Bansal, V.,
assimilation with the ensemble Kalman filter. Mthly. Weather Rev. 130, 103–114. Chennault, C., et al. (2021). Development of a deep learning emulator for a
doi: 10.1175/1520-0493(2002)130<0103:HDAWTE>2.0.CO;2 distributed groundwater–surface water model: ParFlow-ML. Water 13, 3393.
doi: 10.3390/w13233393
Ridler, M.-E., Zhang, D., Madsen, H., Kidmose, J., Refsgaard, J. C., and Jensen,
K. H. (2018). Bias-aware data assimilation in integrated hydrological modelling. Tso, C.-H. M., Johnson, T. C., Song, X., Chen, X., Kuras, O., Wilkinson,
Hydrol. Res. 49, 989–1004. doi: 10.2166/nh.2017.117 P., et al. (2020). Integrated hydrogeophysical modelling and data
assimilation for geoelectrical leak detection. J. Contam. Hydrol. 234, 103679.
Rummler, T., Wagner, A., Arnault, J., and Kunstmann, H. (2022). Lateral
doi: 10.1016/j.jconhyd.2020.103679
terrestrial water fluxes in the LSM of WRF-Hydro: benefits of a 2D groundwater
representation. Hydrol. Process. 36, e14510. doi: 10.1002/hyp.14510 Vidana Gamage, D. N., Biswas, A., Strachan, I. B., and Adamchuk, V. I. (2018).
Soil water measurement using actively heated fiber optics at field scale. Sensors 18,
Samaniego, L., Kumar, R., and Attinger, S. (2010). Multiscale parameter
1116. doi: 10.3390/s18041116
regionalization of a grid-based hydrologic model at the mesoscale. Water Resour.
Res. 46, e2008WR007327. doi: 10.1029/2008WR007327 Villani, C. (2009). Optimal Transport: Old and New, Vol. 338. Springer. Available
online at: https://link.springer.com/book/10.1007/978-3-540-71050-9
Sayde, C., Buelga, J. B., Rodriguez-Sinobas, L., El Khoury, L., English, M., van de
Giesen, N., et al. (2014). Mapping variability of soil water content and flux across Vittecoq, B., Reninger, P.-A., Lacquement, F., Martelet, G., and Violette, S.
1–1000 m scales using the actively heated fiber optic method. Water Resour. Res. (2019). Hydrogeological conceptual model of andesitic watersheds revealed by
50, 7302–7317. doi: 10.1002/2013WR014983 high-resolution heliborne geophysics. Hydrol. Earth Syst. Sci. 23, 2321–2338.
doi: 10.5194/hess-23-2321-2019
Sayde, C., Gregory, C., Gil-Rodriguez, M., Tufillaro, N., Tyler, S., van de Giesen,
N., et al. (2010). Feasibility of soil moisture monitoring with heated fiber optics. Wada, Y., Bierkens, M. F., Roo, A., de, Dirmeyer, P. A., Famiglietti, J. S.,
Water Resour. Res. 46, e2009WR007846. doi: 10.1029/2009WR007846 Hanasaki, N., et al. (2017). Human–water interface in hydrological modelling:
current status and future directions. Hydrol. Earth Syst. Sci. 21, 4169–4193.
Scanlon, B. R., Zhang, Z., Save, H., Sun, A. Y., Schmied, H. M., Van Beek, L. P.,
doi: 10.5194/hess-21-4169-2017
et al. (2018). Global models underestimate large decadal declining and rising water
storage trends relative to GRACE satellite data. Proc. Nat. Acad. Sci. U.S.A. 115, Wood, E. F., Roundy, J. K., Troy, T. J., Van Beek, L. P. H., Bierkens, M. F.,
E1080–E1089. doi: 10.1073/pnas.1704665115 Blyth, E., et al. (2011). Hyperresolution global land surface modeling: meeting a
grand challenge for monitoring Earth’s terrestrial water. Water Resour. Res. 47,
Schattan, P., Baroni, G., Oswald, S. E., Schöber, J., Fey, C., Kormann,
e2010WR010090. doi: 10.1029/2010WR010090
C., et al. (2017). Continuous monitoring of snowpack dynamics in alpine
terrain by aboveground neutron sensing. Water Resour. Res. 53, 3615–3634. Yu, X., Moraetis, D., Nikolaidis, N. P., Li, B., Duffy, C., and Liu, B.
doi: 10.1002/2016WR020234 (2019). A coupled surface-subsurface hydrologic model to assess groundwater
flood risk spatially and temporally. Environ. Model. Softw. 114, 129–139.
Shirzaei, M., Freymueller, J., Törnqvist, T. E., Galloway, D. L., Dura, T., and
doi: 10.1016/j.envsoft.2019.01.008
Minderhoud, P. S. (2020). Measuring, modelling and projecting coastal land
subsidence. Nat. Rev. Earth Environ. 2, 40–58. doi: 10.1038/s43017-020-00115-x Zeng, Y., Xie, Z., Liu, S., Xie, J., Jia, B., Qin, P., et al. (2018). Global land
surface modeling including lateral groundwater flow. J. Adv. Model. Earth Syst. 10,
Smith, R., and Knight, R. (2019). Modeling land subsidence using InSAR
1882–1900. doi: 10.1029/2018MS001304
and airborne electromagnetic data. Water Resour. Res. 55, 2801–2819.
doi: 10.1029/2018WR024185 Zhang, D., Madsen, H., Ridler, M. E., Kidmose, J., Jensen, K.
H., and Refsgaard, J. C. (2016). Multivariate hydrological data
Smith, R. G., and Majumdar, S. (2020). Groundwater storage loss associated with
assimilation of soil moisture and groundwater head. Hydrol.
land subsidence in Western United States mapped using machine learning. Water
Earth Syst. Sci. 20, 4341–4357. doi: 10.5194/hess-20-4341-
Resourc. Res. 56, e2019WR026621. doi: 10.1029/2019WR026621
2016
Smith, W. K., Fox, A. M., MacBean, N., Moore, D. J., and
Zhang, D., Madsen, H., Ridler, M. E., Refsgaard, J. C., and Jensen,
Parazoo, N. C. (2020). Constraining estimates of terrestrial carbon
K. H. (2015). Impact of uncertainty description on assimilating
uptake: new opportunities using long-term satellite observations
hydraulic head in the MIKE SHE distributed hydrological model.
and data assimilation. New Phytol. 225, 105–112. doi: 10.1111/nph.
Adv. Water Resour. 86, 400–413. doi: 10.1016/j.advwatres.2015.
16055
07.018
Steele-Dunne, S. C., Rutten, M. M., Krzeminska, D. M., Hausner,
Zhang, H., Hendricks Franssen, H.-J., Han, X., Vrugt, J. A., and
M., Tyler, S. W., Selker, J., et al. (2010). Feasibility of soil
Vereecken, H. (2017). State and parameter estimation of two land
moisture estimation using passive distributed temperature sensing.
surface models using the ensemble Kalman filter and the particle filter.
Water Resour. Res. 46, e2009WR008272. doi: 10.1029/2009WR0
Hydrol. Earth Syst. Sci. 21, 4927–4958. doi: 10.5194/hess-21-4927-
08272
2017

Frontiers in Water 09 frontiersin.org