@article{FranckeBaroniBrosinskyetal.2018, author = {Francke, Till and Baroni, Gabriele and Brosinsky, Arlena and Foerster, Saskia and Lopez-Tarazon, Jos{\´e} Andr{\´e}s and Sommerer, Erik and Bronstert, Axel}, title = {What Did Really Improve Our Mesoscale Hydrological Model?}, series = {Water resources research}, volume = {54}, journal = {Water resources research}, number = {11}, publisher = {American Geophysical Union}, address = {Washington}, issn = {0043-1397}, doi = {10.1029/2018WR022813}, pages = {8594 -- 8612}, year = {2018}, abstract = {Modelers can improve a model by addressing the causes for the model errors (data errors and structural errors). This leads to implementing model enhancements (MEs), for example, meteorological data based on more monitoring stations, improved calibration data, and/or modifications in process formulations. However, deciding on which MEs to implement remains a matter of expert knowledge. After implementing multiple MEs, any improvement in model performance is not easily attributed, especially when considering different objectives or aspects of this improvement (e.g., better dynamics vs. reduced bias). We present an approach for comparing the effect of multiple MEs based on real observations and considering multiple objectives (MMEMO). A stepwise selection approach and structured plots help to address the multidimensionality of the problem. Tailored analyses allow a differentiated view on the effect of MEs and their interactions. MMEMO is applied to a case study employing the mesoscale hydro-sedimentological model WASA-SED for the Mediterranean-mountainous Isabena catchment, northeast Spain. The investigated seven MEs show diverse effects: some MEs (e.g., rainfall data) cause improvements for most objectives, while other MEs (e.g., land use data) only affect a few objectives or even decrease model performance. Interaction of MEs was observed for roughly half of the MEs, confirming the need to address them in the analysis. Calibration and increasing the temporal resolution showed by far stronger impact than any of the other MEs. The proposed framework can be adopted in other studies to analyze the effect of MEs and, thus, facilitate the identification and implementation of the most promising MEs for comparable cases.}, language = {en} } @article{BaroniScheiffeleSchroenetal.2018, author = {Baroni, Gabriele and Scheiffele, Lena M. and Schr{\"o}n, Martin and Ingwersen, Joachim and Oswald, Sascha Eric}, title = {Uncertainty, sensitivity and improvements in soil moisture estimation with cosmic-ray neutron sensing}, series = {Journal of hydrology}, volume = {564}, journal = {Journal of hydrology}, publisher = {Elsevier}, address = {Amsterdam}, issn = {0022-1694}, doi = {10.1016/j.jhydrol.2018.07.053}, pages = {873 -- 887}, year = {2018}, abstract = {Cosmic-ray neutron sensing (CRNS) is a promising proximal soil sensing technique to estimate soil moisture at intermediate scale and high temporal resolution. However, the signal shows complex and non-unique response to all hydrogen pools near the land surface, providing some challenges for soil moisture estimation in practical applications. Aims of the study were 1) to assess the uncertainty of CRNS as a stand-alone approach to estimate volumetric soil moisture in cropped field 2) to identify the causes of this uncertainty 3) and possible improvements. Two experimental sites in Germany were equipped with a CRNS probe and point-scale soil moisture network. Additional monitoring activities were conducted during the crop growing season to characterize the soil-plant systems. This data is used to identify and quantify the different sources of uncertainty (factors). An uncertainty analysis, based on Monte Carlo approach, is applied to propagate these uncertainties to CRNS soil moisture estimations. In addition, a sensitivity analysis based on the Sobol' method is performed to identify the most important factors explaining this uncertainty. Results show that CRNS soil moisture compares well to the soil moisture network when these point-scale values are weighted to account for the spatial sensitivity of the signal and other sources of hydrogen (lattice water and organic carbon) are added to the water content. However, the performance decreases when CRNS is considered as a stand-alone method to retrieve the actual (non-weighted) volumetric soil moisture. The support volume (penetration depth and radius) shows also a considerable uncertainty, especially in relatively dry soil moisture conditions. Four of the seven factors analyzed (the vertical soil moisture profile, bulk density, incoming neutron correction and the calibrated parameter N0) were found to play an important role. Among the possible improvements identified, a simple correction factor based on vertical point-scale soil moisture profiles shows to be a promising approach to account for the sensitivity of the CRNS signal to the upper soil layers.}, language = {en} } @article{ErdalBaroniSanchezLeonetal.2019, author = {Erdal, Daniel and Baroni, Gabriele and S{\´a}nchez Le{\´o}n, Eduardo Emilio and Cirpka, Olaf A.}, title = {The value of simplified models for spin up of complex models with an application to subsurface hydrology}, series = {Computers \& geosciences : an international journal devoted to the publication of papers on all aspects of geocomputation and to the distribution of computer programs and test data sets ; an official journal of the International Association for Mathematical Geology}, volume = {126}, journal = {Computers \& geosciences : an international journal devoted to the publication of papers on all aspects of geocomputation and to the distribution of computer programs and test data sets ; an official journal of the International Association for Mathematical Geology}, publisher = {Elsevier}, address = {Oxford}, issn = {0098-3004}, doi = {10.1016/j.cageo.2019.01.014}, pages = {62 -- 72}, year = {2019}, abstract = {Spinning up large-scale coupled surface-subsurface numerical models can be a time and resource consuming task. If an uninformed initial condition is chosen, the spin-up can easily require 20 years of repeated simulations on high-performance computing machines. In this paper we compare the classical approach of starting from a fixed shallow depth to groundwater (here 3 m) with three more informed approaches for the definition of initial conditions in the spin up. In the first of these three approaches, we start from a known-steady state groundwater table, calculated with a 2-D groundwater model and the yearly net recharge, and combine it with an unsaturated zone that assumes hydrostatic conditions. In the second approach, we start from the same groundwater table combined with vertical profiles in the unsaturated zone with uniform vertical flow identical to the groundwater recharge. In the third approach we calculate a dynamic steady state from a simplified subsurface model combining a transient 2-D groundwater model with a limited number of 1-D transient unsaturated zone columns on top. Results for spinning-up a 3-D Parflow-CLM model using the different initial conditions show that large gains can be made by considering states in groundwater and the vadose zone that are consistent, i.e. where groundwater recharge and the vertical flux in the vadose zone agree. By this, the spin-up time was reduced from about 10 years to about 3 years of simulated time. In the light of seasonal fluctuations of net recharge, using the transient approach showed more stable results.}, language = {en} } @article{BaroniOrtuaniFacchietal.2013, author = {Baroni, Gabriele and Ortuani, B. and Facchi, A. and Gandolfi, C.}, title = {The role of vegetation and soil properties on the spatio-temporal variability of the surface soil moisture in a maize-cropped field}, series = {Journal of hydrology}, volume = {489}, journal = {Journal of hydrology}, number = {7}, publisher = {Elsevier}, address = {Amsterdam}, issn = {0022-1694}, doi = {10.1016/j.jhydrol.2013.03.007}, pages = {148 -- 159}, year = {2013}, abstract = {Soil moisture dynamics are affected by complex interactions among several factors. Understanding the relative importance of these factors is still an important challenge in the study of water fluxes and solute transport in unsaturated media. In this study, the spatio-temporal variability of surface soil moisture was investigated in a 10 ha flat cropped field located in northern Italy. Soil moisture was measured on a regular 50 x 50 m grid on seven dates during the growing season. For each measurement campaign, the spatial variability of the soil moisture was compared with the spatial variability of the soil texture and crop properties. In particular, to better understand the role of the vegetation, the spatio-temporal variability of two different parameters - leaf area index and crop height - was monitored on eight dates at different crop development stages. Statistical and geostatistical analysis was then applied to explore the interactions between these variables. In agreement with other studies, the results show that the soil moisture variability changes according to the average value within the field, with the standard deviation reaching a maximum value under intermediate mean soil moisture conditions and the coefficient of variation decreasing exponentially with increasing mean soil moisture. The controls of soil moisture variability change according to the average soil moisture within the field. Under wet conditions, the spatial distribution of the soil moisture reflects the variability of the soil texture. Under dry conditions, the spatial distribution of the soil moisture is affected mostly by the spatial variability of the vegetation. The interaction between these two factors is more important under intermediate soil moisture conditions. These results confirm the importance of considering the average soil moisture conditions within a field when investigating the controls affecting the spatial variability of soil moisture. This study highlights the importance of considering the spatio-temporal variability of the vegetation in investigating soil moisture dynamics, especially under intermediate and dry soil moisture conditions. The results of this study have important implications in different hydrological applications, such as for sampling design, ranking stability application, indirect measurements of soil properties and model parameterisation.}, language = {en} } @article{SchattanKoehliSchroenetal.2019, author = {Schattan, Paul and K{\"o}hli, Markus and Schr{\"o}n, Martin and Baroni, Gabriele and Oswald, Sascha Eric}, title = {Sensing area-average snow water equivalent with cosmic-ray neutrons: the influence of fractional snow cover}, series = {Water resources research}, volume = {55}, journal = {Water resources research}, number = {12}, publisher = {American Geophysical Union}, address = {Washington}, issn = {0043-1397}, doi = {10.1029/2019WR025647}, pages = {10796 -- 10812}, year = {2019}, abstract = {Cosmic-ray neutron sensing (CRNS) is a promising non-invasive technique to estimate snow water equivalent (SWE) over large areas. In contrast to preliminary studies focusing on shallow snow conditions (SWE <130 mm), more recently the method was shown experimentally to be sensitive also to deeper snowpacks providing the basis for its use at mountain experimental sites. However, hysteretic neutron response has been observed for complex snow cover including patchy snow-free areas. In the present study we aimed to understand and support the experimental findings using a comprehensive neutron modeling approach. Several simulations have been set up in order to disentangle the effect on the signal of different land surface characteristics and to reproduce multiple observations during periods of snow melt and accumulation. To represent the actual land surface heterogeneity and the complex snow cover, the model used data from terrestrial laser scanning. The results show that the model was able to accurately reproduce the CRNS signal and particularly the hysteresis effect during accumulation and melting periods. Moreover, the sensor footprint was found to be anisotropic and affected by the spatial distribution of liquid water and snow as well as by the topography of the nearby mountains. Under fully snow-covered conditions the CRNS is able to accurately estimate SWE without prior knowledge about snow density profiles or other spatial anomalies. These results provide new insights into the characteristics of the detected neutron signal in complex terrain and support the use of CRNS for long-term snow monitoring in high elevated mountain environments.}, language = {en} } @article{VillarreyesBaroniOswald2014, author = {Villarreyes, Carlos Andres Rivera and Baroni, Gabriele and Oswald, Sascha Eric}, title = {Inverse modelling of cosmic-ray soil moisture for field-scale soil hydraulic parameters}, series = {European journal of soil science}, volume = {65}, journal = {European journal of soil science}, number = {6}, publisher = {Wiley-Blackwell}, address = {Hoboken}, issn = {1351-0754}, doi = {10.1111/ejss.12162}, pages = {876 -- 886}, year = {2014}, abstract = {We used inverse modelling techniques and soil moisture measured by the cosmic-ray neutron sensing (CRS) to estimate root-zone soil hydraulic properties at the field scale. A HYDRUS-1D model was developed for inverse modelling and calibrated with parameter estimation software (PEST) using a global optimizer. Integral CRS measurements recorded from a sunflower farm in Germany comprised the model input. Data were transformed to soil water storage to enable direct model calibration with a HYDRUS soil-water balance. Effective properties at the CRS scale were compared against local measurements and other inversely estimated soil properties from independent soil moisture profiles. Moreover, CRS-scale soil properties were tested on the basis of how field soil moisture (vertical distribution) and soil water storage were reproduced. This framework provided good estimates of effective soil properties at the CRS scale. Simulated soil moisture at different depths at the CRS scale agreed with field observations. Moreover, simulated soil water storage at the CRS scale compared well with calculations from point-scale profiles, despite their different support volumes. The CRS-scale soil properties estimated with the inverse model were within the range of variation of properties identified from all inverse simulations at the local scale. This study demonstrates the potential of CRS for inverse estimation of soil hydraulic properties.}, language = {en} } @article{RiveraVillarreyesBaroniOswald2011, author = {Rivera Villarreyes, C. A. and Baroni, Gabriele and Oswald, Sascha Eric}, title = {Integral quantification of seasonal soil moisture changes in farmland by cosmic-ray neutrons}, series = {Hydrology and earth system sciences : HESS}, volume = {15}, journal = {Hydrology and earth system sciences : HESS}, number = {12}, publisher = {Copernicus}, address = {G{\"o}ttingen}, issn = {1027-5606}, doi = {10.5194/hess-15-3843-2011}, pages = {3843 -- 3859}, year = {2011}, abstract = {Soil moisture at the plot or hill-slope scale is an important link between local vadose zone hydrology and catchment hydrology. However, so far only a few methods are on the way to close this gap between point measurements and remote sensing. One new measurement methodology that could determine integral soil moisture at this scale is the aboveground sensing of cosmic-ray neutrons, more precisely of ground albedo neutrons. The present study performed ground albedo neutron sensing (GANS) at an agricultural field in northern Germany. To test the method it was accompanied by other soil moisture measurements for a summer period with corn crops growing on the field and a later autumn-winter period without crops and a longer period of snow cover. Additionally, meteorological data and aboveground crop biomass were included in the evaluation. Hourly values of ground albedo neutron sensing showed a high statistical variability. Six-hourly values corresponded well with classical soil moisture measurements, after calibration based on one reference dry period and three wet periods of a few days each. Crop biomass seemed to influence the measurements only to minor degree, opposed to snow cover which has a more substantial impact on the measurements. The latter could be quantitatively related to a newly introduced field neutron ratio estimated from neutron counting rates of two energy ranges. Overall, our study outlines a procedure to apply the ground albedo neutron sensing method based on devices now commercially available, without the need for accompanying numerical simulations and suited for longer monitoring periods after initial calibration.}, language = {en} } @article{SchroenKoehliScheiffeleetal.2017, author = {Schroen, Martin and Koehli, Markus and Scheiffele, Lena and Iwema, Joost and Bogena, Heye R. and Lv, Ling and Martini, Edoardo and Baroni, Gabriele and Rosolem, Rafael and Weimar, Jannis and Mai, Juliane and Cuntz, Matthias and Rebmann, Corinna and Oswald, Sascha Eric and Dietrich, Peter and Schmidt, Ulrich and Zacharias, Steffen}, title = {Improving calibration and validation of cosmic-ray neutron sensors in the light of spatial sensitivity}, series = {Hydrology and earth system sciences : HESS}, volume = {21}, journal = {Hydrology and earth system sciences : HESS}, publisher = {Copernicus}, address = {G{\"o}ttingen}, issn = {1027-5606}, doi = {10.5194/hess-21-5009-2017}, pages = {5009 -- 5030}, year = {2017}, abstract = {In the last few years the method of cosmic-ray neutron sensing (CRNS) has gained popularity among hydrologists, physicists, and land-surface modelers. The sensor provides continuous soil moisture data, averaged over several hectares and tens of decimeters in depth. However, the signal still may contain unidentified features of hydrological processes, and many calibration datasets are often required in order to find reliable relations between neutron intensity and water dynamics. Recent insights into environmental neutrons accurately described the spatial sensitivity of the sensor and thus allowed one to quantify the contribution of individual sample locations to the CRNS signal. Consequently, data points of calibration and validation datasets are suggested to be averaged using a more physically based weighting approach. In this work, a revised sensitivity function is used to calculate weighted averages of point data. The function is different from the simple exponential convention by the extraordinary sensitivity to the first few meters around the probe, and by dependencies on air pressure, air humidity, soil moisture, and vegetation. The approach is extensively tested at six distinct monitoring sites: two sites with multiple calibration datasets and four sites with continuous time series datasets. In all cases, the revised averaging method improved the performance of the CRNS products. The revised approach further helped to reveal hidden hydrological processes which otherwise remained unexplained in the data or were lost in the process of overcalibration. The presented weighting approach increases the overall accuracy of CRNS products and will have an impact on all their applications in agriculture, hydrology, and modeling.}, language = {en} } @misc{SchroenKoehliScheiffeleetal.2017, author = {Schr{\"o}n, Martin and K{\"o}hli, Markus and Scheiffele, Lena and Iwema, Joost and Bogena, Heye R. and Lv, Ling and Martini, Edoardo and Baroni, Gabriele and Rosolem, Rafael and Weimar, Jannis and Mai, Juliane and Cuntz, Matthias and Rebmann, Corinna and Oswald, Sascha Eric and Dietrich, Peter and Schmidt, Ulrich and Zacharias, Steffen}, title = {Improving calibration and validation of cosmic-ray neutron sensors in the light of spatial sensitivity}, series = {Postprints der Universit{\"a}t Potsdam : Mathematisch Naturwissenschaftliche Reihe}, journal = {Postprints der Universit{\"a}t Potsdam : Mathematisch Naturwissenschaftliche Reihe}, number = {636}, doi = {10.25932/publishup-41913}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-419134}, pages = {5009 -- 5030}, year = {2017}, abstract = {In the last few years the method of cosmic-ray neutron sensing (CRNS) has gained popularity among hydrologists, physicists, and land-surface modelers. The sensor provides continuous soil moisture data, averaged over several hectares and tens of decimeters in depth. However, the signal still may contain unidentified features of hydrological processes, and many calibration datasets are often required in order to find reliable relations between neutron intensity and water dynamics. Recent insights into environmental neutrons accurately described the spatial sensitivity of the sensor and thus allowed one to quantify the contribution of individual sample locations to the CRNS signal. Consequently, data points of calibration and validation datasets are suggested to be averaged using a more physically based weighting approach. In this work, a revised sensitivity function is used to calculate weighted averages of point data. The function is different from the simple exponential convention by the extraordinary sensitivity to the first few meters around the probe, and by dependencies on air pressure, air humidity, soil moisture, and vegetation. The approach is extensively tested at six distinct monitoring sites: two sites with multiple calibration datasets and four sites with continuous time series datasets. In all cases, the revised averaging method improved the performance of the CRNS products. The revised approach further helped to reveal hidden hydrological processes which otherwise remained unexplained in the data or were lost in the process of overcalibration. The presented weighting approach increases the overall accuracy of CRNS products and will have an impact on all their applications in agriculture, hydrology, and modeling.}, language = {en} } @article{PilzFranckeBaronietal.2020, author = {Pilz, Tobias and Francke, Till and Baroni, Gabriele and Bronstert, Axel}, title = {How to Tailor my process-based hydrological model?}, series = {Water resources research}, volume = {56}, journal = {Water resources research}, number = {8}, publisher = {American Geophysical Union}, address = {Washington}, issn = {0043-1397}, doi = {10.1029/2020WR028042}, pages = {24}, year = {2020}, abstract = {In the field of hydrological modeling, many alternative representations of natural processes exist. Choosing specific process formulations when building a hydrological model is therefore associated with a high degree of ambiguity and subjectivity. In addition, the numerical integration of the underlying differential equations and parametrization of model structures influence model performance. Identifiability analysis may provide guidance by constraining the a priori range of alternatives based on observations. In this work, a flexible simulation environment is used to build an ensemble of semidistributed, process-based hydrological model configurations with alternative process representations, numerical integration schemes, and model parametrizations in an integrated manner. The flexible simulation environment is coupled with an approach for dynamic identifiability analysis. The objective is to investigate the applicability of the framework to identify the most adequate model. While an optimal model configuration could not be clearly distinguished, interesting results were obtained when relating model identifiability with hydro-meteorological boundary conditions. For instance, we tested the Penman-Monteith and Shuttleworth \& Wallace evapotranspiration models and found that the former performs better under wet and the latter under dry conditions. Parametrization of model structures plays a dominant role as it can compensate for inadequate process representations and poor numerical solvers. Therefore, it was found that numerical solvers of high order of accuracy do often, though not necessarily, lead to better model performance. The proposed coupled framework proved to be a straightforward diagnostic tool for model building and hypotheses testing and shows potential for more in-depth analysis of process implementations and catchment functioning.}, language = {en} }