Chicken Creek Catchment Research Articles

One-dimensional deep learning inversion of electromagnetic induction data using convolutional neural network

SUMMARY Conventional geophysical inversion techniques suffer from several limitations including computational cost, nonlinearity, non-uniqueness and dimensionality of the inverse problem. Successful inversion of geophysical data has been a major challenge for decades. Here, a novel approach based on deep learning (DL) inversion via convolutional neural network (CNN) is proposed to instantaneously estimate subsurface electrical conductivity (σ) layering from electromagnetic induction (EMI) data. In this respect, a fully convolutional network was trained on a large synthetic data set generated based on 1-D EMI forward model. The accuracy of the proposed approach was examined using several synthetic scenarios. Moreover, the trained network was used to find subsurface electromagnetic conductivity images (EMCIs) from EMI data measured along two transects from Chicken Creek catchment (Brandenburg, Germany). Dipole–dipole electrical resistivity tomography data were measured as well to obtain reference subsurface σ distributions down to a 6 m depth. The inversely estimated models were juxtaposed and compared with their counterparts obtained from a spatially constrained deterministic algorithm as a standard code. Theoretical simulations demonstrated a well performance of the algorithm even in the presence of noise in data. Moreover, application of the DL inversion for subsurface imaging from Chicken Creek catchment manifested the accuracy and robustness of the proposed approach for EMI inversion. This approach returns subsurface σ distribution directly from EMI data in a single step without any iterations. The proposed strategy simplifies considerably EMI inversion and allows for rapid and accurate estimation of subsurface EMCI from multiconfiguration EMI data.

The Influence of Geostatistical Prior Modeling on the Solution of DCT-Based Bayesian Inversion: A Case Study from Chicken Creek Catchment

Low frequency loop-loop electromagnetic induction (EMI) is a widely-used geophysical measurement method to rapidly measure in situ the apparent electrical conductivity (ECa) of variably-saturated soils. Here, we couple Bayesian inversion of a quasi-two-dimensional electromagnetic (EM) model with image compression via the discrete cosine transform (DCT) for subsurface electrical conductivity (EC) imaging. The subsurface EC distributions are obtained from multi-configuration EMI data measured with a CMD-Explorer sensor along two transects in the Chicken Creek catchment (Brandenburg, Germany). Dipole-dipole electrical resistivity tomography (ERT) data are used to benchmark the inferred EC fields of both transects. We are especially concerned with the impact of the DCT truncation method on the accuracy and reliability of the inversely-estimated EC images. We contrast the results of two different truncation approaches for model parametrization. The first scenario considers an arbitrary selection of the dominant DCT coefficients and their prior distributions (a commonly-used approach), while the second methodology benefits from geostatistical simulation of the EMI data pseudosection. This study demonstrates that DCT truncation based on geostatistical simulations facilitates a robust selection of the dominant DCT coefficients and their prior ranges, resulting in more accurate subsurface EC imaging from multi-configuration EMI data. Results based on geostatistical prior modeling present an excellent agreement between the EMI- and ERT-derived EC fields of the Chicken Creek catchment.

A web-based platform for terrestrial data repository from Chicken Creek catchment

Exploring hydrological and ecological processes plays a key role in understanding ecosystem developments. In this respect, the constructed catchment, Chicken Creek, has been established for fundamental and interdisciplinary scientific research. Since 2005, an ongoing monitoring program has been launched to measure hydrological, biological, meteorological, and pedological parameters during the ecological development of the site. This comprehensive and multidisciplinary monitoring program has produced a diverse large data set. Handling such complex data for research purposes can be a cumbersome task. Consequently, we developed an online data portal (https://www.b-tu.de/chicken-creek/apps/datenportal/) to efficiently handle the data from Chicken Creek catchment. The portal was constructed using Shiny package of the R programming language. This platform provides a web-based data repository allowing for data discovery, download, visualization, and analysis. The data include time series of different parameters from installed sensors, data from laboratory analyses, vegetation data, data from sampling campaigns, and aerial photos. This platform demonstrates the relevancy and potentiality of the R-Shiny for constructing an online data portal to be used for multidisciplinary scientific purposes. The Chicken Creek data portal thus provides a comprehensive and reliable database to give scientists a fast and easy access to all collected data.

Machine learning to estimate soil moisture from geophysical measurements of electrical conductivity

ABSTRACTSoil water content (θ) is a key variable in different earth science disciplines since it mediates the water and energy exchange between the surface and atmosphere. Electrical and electromagnetic geophysical techniques have been widely used to estimate soil electrical conductivity (σ) and soil moisture. However, obtaining the relationship is not straightforward due to the non‐linearity and also dependency on many different soil and environmental properties. The purpose of this paper is to determine if artificial neural network is an appropriate machine learning technique for relating electrical conductivity to soil water content. In this respect, time‐lapse electrical resistivity tomography measurements were carried out along a transect in the Chicken Creek catchment (Brandenburg, Germany). To ensure proper retrieval of the σ and θ, reference values were measured near the beginning of the transect via an excavated pit using 5TE capacitance sensors installed at different depths. We explored robustness and pertinence of the artificial neural network approach in comparison with Rhoades model (as a commonly used petrophysical relationship) to convert the inversely estimated σ from electrical resistivity tomography to the θ. The proposed approach was successfully validated and benchmarked by comparing the estimated values with the reference data. This study showed the superiority of the artificial neural network approach to the Rhoades model to obtain relationship. In particular, artificial neural network allowed for more accurate estimation of the temporal wetting front than the petrophysical model. The proposed methodology thus offers a great promise for deriving spatiotemporal soil moisture patterns from geophysical data and obtaining the in situ relationship, taking into account the non‐linear variations of the soil moisture.

Water Balance Dynamics during Ten Years of Ecological Development at Chicken Creek Catchment

Core Ideas Water balance was calculated for the constructed catchment Chicken Creek. Influence of abiotic and biotic factors feedback mechanisms on the hydrology was determined. Different phases were derived during the development of the ecohydrological system. Difficulties in quantitatively closing the water balance of catchments arise when upscaling point measurements and from insufficient knowledge of the physical boundaries, inner structure, and storage volumes of natural catchments. In addition, there is a strong need for generalizing the relationship between catchment characteristics and hydrological response. Therefore, experimental catchments with well‐known boundaries and conditions could contribute valuable data to hydrological and critical zone research. One of the most well‐established and largest constructed catchments is the Chicken Creek catchment (6 ha including a pond, Brandenburg, Germany) representing an initial ecosystem undergoing highly dynamic ecological development starting from clearly defined starting conditions. Directly after completion of the construction, extensive monitoring equipment was installed to track the ecosystem development and to capture the spatiotemporal variability of meteorological, hydrological, ecological, and soil conditions and vegetation succession. In this study, we focused on the water balance dynamics of the Chicken Creek catchment for the period 2005 to 2015 as influenced by ecological development. Water storage in the catchment was calculated from a three‐dimensional model of groundwater volumes, soil moisture measurements, and water level recordings of the pond. The catchment water balance equation was resolved for evapotranspiration, the only part that was not measured directly. Time series of meteorological, hydrological, and ecological data for 10 yr enabled us to characterize the transient development of the catchment and to evaluate the effect of different feedback mechanisms on catchment hydrology.

Impact of modellers' decisions on hydrological a priori predictions

Abstract. In practice, the catchment hydrologist is often confronted with the task of predicting discharge without having the needed records for calibration. Here, we report the discharge predictions of 10 modellers – using the model of their choice – for the man-made Chicken Creek catchment (6 ha, northeast Germany, Gerwin et al., 2009b) and we analyse how well they improved their prediction in three steps based on adding information prior to each following step. The modellers predicted the catchment's hydrological response in its initial phase without having access to the observed records. They used conceptually different physically based models and their modelling experience differed largely. Hence, they encountered two problems: (i) to simulate discharge for an ungauged catchment and (ii) using models that were developed for catchments, which are not in a state of landscape transformation. The prediction exercise was organized in three steps: (1) for the first prediction the modellers received a basic data set describing the catchment to a degree somewhat more complete than usually available for a priori predictions of ungauged catchments; they did not obtain information on stream flow, soil moisture, nor groundwater response and had therefore to guess the initial conditions; (2) before the second prediction they inspected the catchment on-site and discussed their first prediction attempt; (3) for their third prediction they were offered additional data by charging them pro forma with the costs for obtaining this additional information. Holländer et al. (2009) discussed the range of predictions obtained in step (1). Here, we detail the modeller's assumptions and decisions in accounting for the various processes. We document the prediction progress as well as the learning process resulting from the availability of added information. For the second and third steps, the progress in prediction quality is evaluated in relation to individual modelling experience and costs of added information. In this qualitative analysis of a statistically small number of predictions we learned (i) that soft information such as the modeller's system understanding is as important as the model itself (hard information), (ii) that the sequence of modelling steps matters (field visit, interactions between differently experienced experts, choice of model, selection of available data, and methods for parameter guessing), and (iii) that added process understanding can be as efficient as adding data for improving parameters needed to satisfy model requirements.

Dynamics, chemical properties and bioavailability of DOC in an early successional catchment

Abstract. The dynamics of dissolved organic carbon (DOC) have been intensively studied in mature ecosystems, but little is known about DOC dynamics and the significance of DOC as a substrate for microbial activity in early-successional catchments. We determined the concentration, chemical composition, source, radiocarbon age, and bioavailability of DOC along the hydrological flow path from soil solution to a downstream pond in a recently constructed catchment (Chicken Creek Catchment, Germany). Soil solution, upwelling ground water, stream water, subsurface water in an alluvial fan, and pond water all had high DOC concentrations (averages: 6.0–11.6 mg DOC L–1), despite small carbon stocks in both vegetation and soil of the catchment. Solid-state CPMAS 13C NMR of DOC in upwelling ground water revealed a higher proportion of aromatic compounds (32%) and a lower proportion of carbohydrates (33%) than in pond water (18% and 45%, respectively). The average 14C age of DOC in upwelling ground water was 2600 to 2900 yr, while organic matter of the Quaternary substrate of the catchment had a 14C age of 3000 to 16 000 yr. Both the 14C age data and 13C NMR spectra suggest that DOC partly derived from organic matter of the Quaternary substrate (about 40 to 90% of the C in the DOC), indicating that both recent and old C of the DOC can support microbial activity during early ecosystem succession. However, in a 70 day incubation experiment, only about 11% of the total DOC was found to be bioavailable. This proportion was irrespective of the water type. Origin of the microbial communities within the catchment (enriched from soil, stream sediment or pond water) also had only a marginal effect on overall DOC utilization.

Dynamics of initial ecosystem development at the artificial catchment Chicken Creek, Lusatia, Germany

In autumn 2005, the artificial catchment Chicken Creek was completed in an open-cast lignite mine in Lusatia, Germany. The 6 ha area has been constructed as a two-layer system consisting of a clay aquiclude and a sandy aquifer at the top. After construction, the site was left to an unrestricted and unmanaged succession. A comprehensive environmental monitoring program started immediately after the site was completed. Time series of essential environmental parameters were recorded with high temporal and spatial resolution. This paper presents selected time series of the past 6-year ecosystem development. Important changes registered in this period allow for the definition of distinctive phases of the still ongoing initial ecosystem evolution. A primary, short, but pronounced geo-phase—characterized by surface runoff, excessive erosion and sedimentation as well as very rapid immigration of biota—was followed by a hydrological dominated phase with processes, such as groundwater recharge. At the end of the study period, biotic processes became more evident. It can be concluded that the artificial catchment offers unique opportunities for interdisciplinary research on the establishment of an ecological system with rapidly growing complexity. The highly dynamic development of the Chicken Creek catchment provides the possibility to observe manifold changes within short time and to detect feedbacks and their modifications between different ecosystem compartments.

Root growth of <i>Lotus corniculatus</i> interacts with P distribution in young sandy soil

Abstract. Large areas of land are restored with unweathered soil substrates following mining activities in eastern Germany and elsewhere. In the initial stages of colonization of such land by vegetation, plant roots may become key agents in generating soil formation patterns by introducing gradients in chemical and physical soil properties. On the other hand, such patterns may be influenced by root growth responses to pre-existing substrate heterogeneities. In particular, the roots of many plants were found to preferentially proliferate into nutrient-rich patches. Phosphorus (P) is of primary interest in this respect because its availability is often low in unweathered soils, limiting especially the growth of leguminous plants. However, leguminous plants occur frequently among the pioneer plant species on such soils, as they only depend on atmospheric nitrogen (N) fixation as N source. In this study we investigated the relationship between root growth allocation of the legume Lotus corniculatus and soil P distribution on recently restored land. As test sites, the experimental Chicken Creek Catchment (CCC) in eastern Germany and a nearby experimental site (ES) with the same soil substrate were used. We established two experiments with constructed heterogeneity, one in the field on the experimental site and the other in a climate chamber. In addition, we conducted high-density samplings on undisturbed soil plots colonized by L. corniculatus on the ES and on the CCC. In the field experiment, we installed cylindrical ingrowth soil cores (4.5 × 10 cm) with and without P fertilization around single two-month-old L. corniculatus plants. Roots showed preferential growth into the P-fertilized ingrowth-cores. Preferential root allocation was also found in the climate chamber experiment, where single L. corniculatus plants were grown in containers filled with ES soil and where a lateral portion of the containers was additionally supplied with a range of different P concentrations. In the high-density samplings, we excavated soil-cubes of 10 × 10 × 10 cm size from the topsoil of 3 mini-plot areas (50 × 50 cm) each on the ES and the CCC on which L. corniculatus had been planted (ES) or occurred spontaneously (CCC) and for each cube separated the soil attached to the roots (root-adjacent soil) from the remaining soil (root-distant soil). Root length density was negatively correlated with labile P (resin-extractable P) in the root-distant soil of the CCC plots and with water-soluble P in the root-distant soil of the ES plots. The results suggest that P depletion by root uptake during plant growth soon overrode the effect of preferential root allocation in the relationship between root density and plant-available soil P heterogeneity.

Impact of dumped sediment structures on hydrological modelling in the artificial Chicken Creek catchment, Germany

Revealing the hydrological impact of sediment structures promises a better understanding of the influence of the spatial variability of sediment properties on the hydrological patterns and processes at the catchment scale. To improve the knowledge of structure-process interactions in initial ecosystems, the 6-ha artificial Chicken Creek Catchment in Germany was investigated by the Transregional Collaborative Research Centre 38 (SFB/TRR 38). Sediment structures called pour-ribs, which are dumped by stackers during the construction process, lead to differently compacted sediment zones, which increase the spatial variability of sediments’ hydraulic properties. Although levelled afterwards by bulldozers, the majority of these structures remain in the subsurface. To analyse the effects of pour-ribs on the hydrological catchment’s behaviour, the process-based spatially distributed Water balance Simulation Model (WaSiM-ETH) was applied. The results show that the consideration of pour-ribs improves the runoff simulation and significantly affects the simulated soil moisture patterns and, thereby, the initial stage of the ecosystem development. Compacted zones act as hydraulic barriers and inhibit subsurface lateral water flow, whereas non-compacted zones constitute areas with increased water storage capacity. Both effects cause reduced catchment runoff. Moreover, disregarding of the pour-ribs was identified as a source of model uncertainty in previous studies. A further outcome of this study is the importance of a global sensitivity analysis as a tool for model improvement. Finally, the results stress the importance of considering the variability of sediment properties for hydrological modelling at the catchment scale.

A structure generator for modelling the initial sediment distribution of an artificial hydrologic catchment

Abstract. Artificially-created hydrological catchments are characterised by sediment structures from technological construction processes that can potentially be important for modelling of flow and transport and for understanding initial soil and ecosystem development. The subsurface spatial structures of such catchments have not yet been sufficiently explored and described. Our objective was to develop a structure generator programme for modelling the 3-D spatial distribution patterns of dumped sediments depending on the technical earth-moving and deposition processes. We are focussing in a first step on integrating sediment dumping, particle size, and bulk density modification processes on the catchment scale. For the model development, the artificially-constructed hydrological catchment "Chicken Creek" located in Lower Lusatia, Germany, served as an example. The structure generator describes 3-D technological sediment distributions at two scales: (i) for a 2-D-vertical cross-section, texture and bulk density distributions are generated within individual spoil cones that result from mass dumping, particle segregation, and compaction and (ii) for the whole catchment, the spoil cones are horizontally arranged along trajectories of mass dumping controlled by the belt stacker-machine relative to the catchment's clay layer topography. The generated 3-D texture and bulk density distributions are interpolated and visualised as a gridded 3-D-volume body using 3-D computer-aided design software. The generated subsurface sediment distribution for the Chicken Creek catchment was found to correspond to observed patterns already without calibration. Spatial aggregation and interpolation in the gridded volume body modified the generated distributions towards more uniform (unimodal) distributions and lower values of the standard deviations. The modelling approach is generally applicable to all situations where large masses of unconsolidated sediment are moved and dumped thereby allowing generation of basic soil structures and patterns of hydrological systems.

Observation of hydrological processes and structures in the artificial Chicken Creek catchment

Hydrological catchments are the essential observation spaces for hydrological and ecological studies on a landscape scale. Generally, natural catchments are characterised by high complexity and heterogeneity with uncertainties in boundaries, interior structures and properties. Artificially created, spatially and structurally well-defined hydrological catchments might be a suitable tool for enabling progress in hydrological research. “Chicken Creek”, one of the largest artificial catchments, with clearly defined boundary conditions and extended monitoring facilities, was set up on a dump site associated with opencast mining activities in East Germany. This catchment, left to undisturbed succession, has enabled the observation of ecosystem development from its very beginnings. A sophisticated measurement concept covering hydrological, meteorological and soil physical parameters was designed. It aimed at the detection and description of the hydrological processes that form structures and are controlled by structures at the catchment scale. In mature natural catchments that have already achieved a stage of dynamic equilibrium with evolved runoff structures and land cover, hot spots of hydrological interest are normally known or easily identifiable prior to the establishment of any monitoring concept. For a newly created artificial catchment, there is no recorded experience regarding hydrological development and behaviour. Under the premise of looking at a defined, initially unstructured, homogenous, isotropic, “virgin” area, a primarily grid-based observation concept was chosen as the initial approach. As components of the progressive development of the ecosystem, new soil and runoff structures have been formed, and vegetation has spread on the initially uncovered surface. As a function of the changed hydrological properties and the improved system knowledge, the initial observation programme was iteratively extended and adjusted by the addition of more structure- and process-related installations. In this study, the design of the experimental observation set-up, the gathered experiences and the selected results of a large number of the most various hydrological data are presented. The combined analysis of measurement data and qualitative information worked to determine that the precipitation-induced surface runoff generation and the filling of the catchment storage were the main relevant hydrological processes in the initial development phase of the Chicken Creek catchment. Formed by the dominating surface runoff, erosion gullies are the most easily distinguishable structures influencing the runoff process. Despite the well-known catchment boundaries and the extensive monitoring, the catchment water balance has not been adequately closed so far. Due to the catchment’s properties (porosity, morphology, land cover), which are less homogeneous than initially assumed, the determination of the catchment’s evapotranspiration and changes in storage remains the challenge for understanding the Chicken Creek catchment.

Modelling subsurface drainage pathways in an artificial catchment

The relationship between rainfall and runoff in catchments is a nonlinear function largely controlled by hydrologically connected hillslope elements. In this study, we investigate material and geometrical hillslope properties determining connection of hillslope elements to form subsurface drainage pathways in an artificial catchment (Chicken Creek) and we examine the impact of the drainage pathway connectivity on the catchment runoff. We developed a three-dimensional model based on percolation theory to simulate the drainage pathways in the catchment and to quantify the contribution from subsurface flow paths to total catchment outflow. The simulations show that the degree of drainage pathway connectivity depends on small-scale variability of soil depth, soil hydraulic properties and topography. Additionally, connectivity was very sensitive to changes of soil water content in the catchment with dramatic change of subsurface flow for small changes in soil water content. This threshold-like behaviour of flow paths connectivity accounted for the sudden availability of a large amount of drainable water, postulating a nonlinear rainfall–runoff relationship as confirmed by first observations in the Chicken Creek catchment. The model reproduced subsurface flow patterns, which were indicated by soil moisture measurements and visual observations in the catchment.

Grope in the Dark – Hydrological modelling of the artificial Chicken Creek catchment without validation possibilities

Environmental modelling is often reduced to a calibration–validation exercise in which a thorough understanding of the processes involved is less important than sophisticated parameter estimation tools. A problem occurs when no data are available for calibration and validation. To compare the ability of different models and modellers to describe the hydrological behaviour of ungauged catchments, a model comparison was initiated by Transregional Collaboration Research Centre 38 (SFB/TRR 38). The comparison was conducted in several consecutive stages, in which knowledge of the investigated artificial catchment of approximately 6 ha was enhanced stepwise. For modelling purposes, all basic input data were given, but neither discharge nor soil moisture measurements were provided. This hard information was supplemented stepwise by soft (qualitative) information mostly obtained by a field inspection after the first modelling stage. In this study, the process-based distributed Water balance Simulation Model (WaSiM-ETH) was applied. An initial perceptual model was developed without site-specific knowledge. Significant modifications to the perceptual model were conducted using the available soft information. These modifications resulted in more reliable water fluxes. The main changes, such as the importance of a sealed topsoil layer, vegetation development, initial conditions, and the effects of a small pond on discharge behaviour, were related to process understanding. Although hard information on the water balance was still missing, the qualitative behaviour of the model seemed to be realistic. This underlines the importance of soft information for successful modelling, which is especially important if ungauged catchments are to be investigated.

Comparative predictions of discharge from an artificial catchment (Chicken Creek) using sparse data

Abstract. Ten conceptually different models in predicting discharge from the artificial Chicken Creek catchment in North-East Germany were used for this study. Soil texture and topography data were given to the modellers, but discharge data was withheld. We compare the predictions with the measurements from the 6 ha catchment and discuss the conceptualization and parameterization of the models. The predictions vary in a wide range, e.g. with the predicted actual evapotranspiration ranging from 88 to 579 mm/y and the discharge from 19 to 346 mm/y. The predicted components of the hydrological cycle deviated systematically from the observations, which were not known to the modellers. Discharge was mainly predicted as subsurface discharge with little direct runoff. In reality, surface runoff was a major flow component despite the fairly coarse soil texture. The actual evapotranspiration (AET) and the ratio between actual and potential ET was systematically overestimated by nine of the ten models. None of the model simulations came even close to the observed water balance for the entire 3-year study period. The comparison indicates that the personal judgement of the modellers was a major source of the differences between the model results. The most important parameters to be presumed were the soil parameters and the initial soil-water content while plant parameterization had, in this particular case of sparse vegetation, only a minor influence on the results.