Enter Zoom Meeting

HS2.2.4

Improving hydrological process understanding and model prediction using soil moisture data

The importance of soil moisture for the hydrological systems dynamics is undebated. A great deal of observations and research have been invested in the last decades to improve the knowledge of soil water status as well its spatial and temporal variation within a given hydrological system. In that effort, several types of soil moisture data have become available, spanning from in-situ observations, radar data, cosmic ray studies to several satellite products.

Although spatial and temporal patterns of soil moisture are the result of processes that hydrological models typically capture, the application of the currently available soil moisture information for improving models is progressing only slowly. This is partly due to a gap between the information content provided by the available data and the information required to improve models. Furthermore, some essential parts of soil water storage at the larger scale, like that of the root zone, is typically assessed using combination of models and data, resulting in a lack of independent information for validation.

This session invites contributions dealing with closing these gaps. This could, for example, be achieved by progress in the descriptions of the processes causing the spatial and temporal variations in soil moisture or by more efficiently using information from available data to improve model predictions across scales. The session is explicitly open for research across all relevant hydrological scales: local, hillslope, catchment up to the continental scale, and deal with both the vertical and lateral flow processes.

Examples for suitable contributions are (but are not limited to):
- The role of soil moisture in the functioning of hydrological systems
- Methods and case studies on improving the predictive power of models using soil moisture data
- Deriving process knowledge from soil moisture data that can be used to improve hydrological models
- Evaluating the suitability of given soil moisture data types for representing hydrologic processes

Co-organized by SSS10
Convener: Anke Hildebrandt | Co-conveners: Josie Geris, Markus Hrachowitz, Daniele Penna
Presentations
| Tue, 24 May, 13:20–15:55 (CEST)
 
Room 2.17

Tue, 24 May, 13:20–14:50

Chairpersons: Anke Hildebrandt, Daniele Penna

13:20–13:23
Welcome

13:23–13:33
|
EGU22-4465
|
solicited
|
Highlight
Mie Andreasen et al.

Since the 1990´s, and in particular during the last decade, afforestation has become a common water management practice. Afforestation improves the quality of the groundwater resource by reducing the leaching of nutrients and pesticides in the soil. Furthermore, planting of trees is also used to capture carbon from the atmosphere as an integral element of carbon emission mitigation, for biodiversity restoration and for biofuel. With the more extensive implementation of afforestation, it is important to understand the hydrological responses and to predict and quantify these adequately using hydrological modelling.

The hydrology of the forest system is characterized by high spatial variability. The forest vegetation intercepts and redistribute a considerable fraction of the precipitation resulting in an uneven input of water at the forest floor. The transpiration and soil evaporation vary in space according to the tree root distribution and soil texture. All these factors influence the soil moisture in the unsaturated zone, the percolation, and the groundwater recharge. Hydrological models are often used to estimate the groundwater recharge rate and to obtain information of the timing of the recharge to ensure sustainable groundwater exploitation and sufficient streamflow. The high spatial variability makes it difficult to predict forest hydrology and it is important that the observations are representative of the forest plot to assess the performance of the hydrological model.

In this study, we predict the water balance for bare ground conditions and for a coniferous forest to examine the hydrological responses to afforestation. We use a physically based and spatially distributed hydrological model with an energy-based description of evapotranspiration processes (MIKE SHE SVAT). The forest model was calibrated against timeseries of throughfall and point-scale soil moisture. Simulated soil moisture is evaluated against forest plot cosmic-ray neutron and point-scale estimates. Further assessment of the model is obtained through comparison to time-series of forest plot eddy-covariance evapotranspiration estimates and observation-based and predicted interception loss. We find that the forest plot and point-scale soil moisture estimates differ which in turn affects the assessment of the reliability of the model performance. The hydrological responses of afforestation are significant, influencing the total evapotranspiration, the soil moisture and the groundwater recharge.  

How to cite: Andreasen, M., Christiansen, J. R., Sonnenborg, T. O., Stisen, S., and Looms, M. C.: Assessing modelled hydrological responses to afforestation using hectometre-scale cosmic-ray neutron soil moisture, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4465, https://doi.org/10.5194/egusphere-egu22-4465, 2022.

13:33–13:40
|
EGU22-3151
|
Highlight
Flora Branger et al.

Soil moisture is a critical control of process-based hydrologic models. This variable has so far been little used, mainly due to the difficulty to extract information from in-situ soil moisture observations that can be directly compared to simulated model variables. The concept of hydrological signature is now being increasingly used for the evaluation of hydrological models. However, hydrological signatures based on soil moisture are still rarely used.

We propose nine soil moisture signatures, encompassing three levels of hydrological time response (storm event response : rising time, normalized amplitude, response type, rising limb density, seasonal response : dates and durations of seasonal transitions, average characteristic values : distribution type, field capacity and wilting point). These signatures were applied to datasets from six in-situ observatories around the world with contrasted climates and land uses. The obtained values were analysed to assess whether the signatures could discriminate between land uses and could be interpreted in terms of hydrological processes.

Results showed that differences could be found between land uses for most signatures, and that these differences could be attributed to flow pathways or soil wetness, hence indicating that the signatures are good indicators of key hydrological processes and potentially useful for model evaluation.

How to cite: Branger, F., Araki, R., Wiekenkamp, I., and McMillan, H.: Improving hydrological process understanding and model prediction using soil moisture data, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3151, https://doi.org/10.5194/egusphere-egu22-3151, 2022.

13:40–13:47
|
EGU22-3514
Pierre Baguis et al.

We investigate the possibilities to improve hydrological simulations by assimilating active radar backscatter observations from the Advanced Scatterometer (ASCAT) in the hydrological model SCHEME. This effort is motivated by the great need of accurate initial model states in hydrological forecasting and the potential to improve them by using remotely sensed data of land surface processes. ASCAT data assimilation is enabled by coupling the Water Cloud Model (WCM) with the SCHEME model. We calibrated the WCM over two catchments in Belgium exhibiting different hydrological regimes. We explore a data assimilation system based on the Ensemble Kalman Filter (EnKF) whereby the observation operator is given by the coupling of WCM and SCHEME models. This coupling underlines the advantage of using backscatter data for assimilation purposes instead of a soil moisture product carrying its own climatology. In the present study we focus on optimising the EnKF for the task, unveil the main challenges and investigate possible solutions including methods to address the biases affecting the data assimilation procedure.

How to cite: Baguis, P., Carrassi, A., Roulin, E., Vannitsem, S., Van den Bergh, J., Modanesi, S., and Lievens, H.: Assimilation of backscatter observations in a hydrological model: a case study in Belgium using ASCAT data, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3514, https://doi.org/10.5194/egusphere-egu22-3514, 2022.

13:47–13:54
|
EGU22-11356
|
ECS
Friedrich Boeing et al.

The 2018-2020 consecutive drought events in Germany resulted in impacts related with several sectors such as agriculture, forestry, water management, industry, energy production and transport. The key to increase preparedness for extreme drought events are high-resolution information systems. A major national operational drought information system is the German Drought Monitor (GDM), launched in 2014 [1]. It provides daily soil moisture (SM) simulated with the mesoscale hydrological model (mHM) and its related soil moisture index [2] at a spatial resolution of 4×4 km². The release of the new soil map BUEK200 allowed us to increase its model resolution to ≈1.2×1.2 km², which is used now for the second version of the GDM [3].

To explore the ability of the GMD-v2 to provide drought information at one-kilometer scale, we evaluated mHM soil moisture simulations against an unprecedented large sample of soil moisture observations from 40 locations across Germany. These SM observations are obtained from single profile measurements, spatially distributed sensor networks, cosmic-ray neutron stations, and lysimeters over a wide range of climatic conditions, vegetation types and soil depths. Specifically, the study aimed at answering two research questions: 1) how well do high-resolution German-wide soil moisture simulations capture the dynamics in observed soil moisture that constitute the basis for the near real-time soil moisture drought monitoring system? 2) Does the mHM simulations obtained with the high spatial resolution data set provide soil moisture estimates with greater model efficiency than those obtained in the coarser resolution?

The results showed that the agreement of simulated and observed SM dynamics is especially high during the vegetation period (0.84 median Spearman correlation(r)) and lower in winter (0.59 median r). Moderate but significant improvements between the low- and high-resolution GDM versions to observed SM were found in correlations for autumn (+0.07 median r) and winter (+0.12 median r). The spatially distributed sensor networks outperformed single profile measurements with higher than average correlation values especially for the 25–60 cm depth, which supports the closer scale match of spatially distributed measurements to the simulations. The results indicate areas for potential improvement and shows limitations from both: model parameterization (e.g., improvement of local scale hydrological processes) and observations methodology (e.g., reduction of measurement errors). Finally, the results of this study underline the fact that nationwide drought information systems depend both on appropriate simulations of the water cycle and a broad, high-quality observational soil moisture database.

References:

[1] Zink, M. et al. doi: 10.1088/1748-9326/11/7/074002 , 2016

[2] Samaniego et al. doi: 10.1175/jhm-d-12-075.1 2013

[3] Boeing, et al. doi: 10.5194/hess-2021-402 2021 (in revision)

How to cite: Boeing, F., Rakovec, O., Kumar, R., Samaniego, L., Schrön, M., Hildebrandt, A., Rebmann, C., Thober, S., Müller, S., Zacharias, S., Bogena, H., Schneider, K., Kiese, R., and Marx, A.: High resolution soil drought simulations evaluated at an unprecedented broad-range of soil moisture networks in Germany, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11356, https://doi.org/10.5194/egusphere-egu22-11356, 2022.

13:54–14:01
|
EGU22-11810
|
ECS
|
Highlight
Soham Adla et al.

Poverty reduction programs across the world have invested in the agriculture sector, specifically in agricultural technology. Irrigation remains a crucial input to agriculture, and the lack of access to supplemental irrigation aggravates the distress of farmers, particularly, smallholders. Crop simulation models use parameters like crop characteristics, environmental conditions and management practices in combination with the local input data, to compute the 'yield response of crops to water', to better inform irrigation decision-making, for saving resources and/or increasing yield. Soil moisture data can be critical to develop more representative crop models by influencing soil hydraulic parameter estimation, and consequently improving the simulation of soil water movement. The dearth of cost-effective soil moisture sensors is a limitation to their effective incorporation in crop modelling, but calibrating them against primary or secondary standards can expand their scope of application. This study applies different calibration techniques on the low-cost capacitance based soil moisture sensor, Spectrum SM100. Calibration techniques include segmented linear regression, polynomial regression, spline regression, and machine learning algorithms such as support vector regression, random forest regression, multi-layer perceptron, extreme learning machine and support vector categorization. Independent soil moisture data are taken as both continuous and categorical variables, are calibrated both in the laboratory and field, and validated using field data. Field data is obtained from an experimental field in Kanpur (India) during a wheat cropping season in 2018. The experimental site is representative of an intensively managed rural landscape in the Ganga river basin, India. The calibrated soil moisture data are subsequently used in the  crop-water productivity model FAO Aquacrop to tune its soil hydraulic properties. Various models are developed with soil hydraulic parameter sets estimated using the calibrated soil moisture data. The respective performances of these models are compared with the default model performance (with parameters derived from the literature), based on outputs of interest such as above ground biomass, crop yield and water use efficiency. A representative crop model is then used to develop scenarios of irrigation scheduling, with varying degrees of water stress. Results indicate that calibrating the soil moisture sensors in laboratory conditions alone is not sufficient to parameterize soil hydraulic properties, and adequate parameterization requires sensor calibration in field conditions. Further, a cost-benefit analysis is conducted to assess and critically discuss the tradeoffs between the cost of soil moisture monitoring and the obtained crop yield.

How to cite: Adla, S., Bruckmaier, F., Arias Rodriguez, L. F., Tripathi, S., Disse, M., and Pande, S.: Analysing the impact of calibrating a low-cost soil moisture sensor on FAO Aquacrop model performance., EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11810, https://doi.org/10.5194/egusphere-egu22-11810, 2022.

14:01–14:08
|
EGU22-5078
Gertraud Meißl et al.

Since 2009, we have continuously monitored soil moisture in the Eastern Alpine torrent catchment of the Brixenbach (Tyrol, Austria). The measurement network is one of the rare with a high spatial resolution and long temporal coverage and consists of eight sites with three frequency-domain (FD) sensors 10 cm below soil surface. The resulting data allowed us to analyse the precipitation-runoff reaction of the catchment depending on the antecedent soil moisture content. In Meißl et al. (2020) we found:

  • The site-specific soil moisture medians correlate with altitude, but don’t correlate with sites’ slope, the topographic index nor the specific upslope area.
  • In contrast to the results of other authors who analysed much shorter time series, the scatter plot of the spatial standard deviation of soil moisture against the spatial mean does not show a convex shape. We found that progressive drying during rainless periods leads to increasing spatial variability of soil moisture contents at mean soil moisture values<40 vol%. Above about 42 vol% the spatial variability of soil moisture contents decreases.
  • The most exceptional out of the 547 analysed rainfall-runoff events took place at rainfall event types with high precipitation sum and long duration, but low intensity or at events with medium precipitation sum, short duration, but high intensity.
  • 244 precipitation events triggered a significant increase in soil moisture (≥ 0.5 vol%) and a total runoff of at least three cubic metres. During these events, the Brixenbach catchment showed a clear threshold behaviour: Discharge coefficients above 0.23 were only observed when the spatial mean soil moisture exceeded 43.5 vol% at the eight sites. Looking at the individual sites, this threshold is also more or less clearly visible, but at different levels. The level of the spatial mean of all sites thus depends strongly on the number and local characterstics of the sites used.
  • If we define the relative soil moisture as proportion of the maximum soil moisture content of the site during the whole measurement period, the threshold ranges between 0.65 and 0.80 with the sites’ mean of 0.72, which can be interpreted as saturation deficit of 0.28.
  • At moist conditions, event streamflow peaked prior to soil moisture, which can be explained by increased surface flow volumes at higher soil moisture as well as already initialized subsurface flow paths.

The analyses of the long-term soil moisture time series provide a valuable insight into the hydrological system of the Brixenbach catchment and may help to identify critical conditions, which may lead to floods, also under changed conditions in future.

References: Meißl G., Zieher Th., Geitner C. (2020): Runoff response to rainfall events considering initial soil moisture – Analysis of 9-year records in a small Alpine catchment (Brixenbach valley, Tyrol, Austria). Journal of Hydrology: Regional Studies, August 2020, 100711. https://doi.org/10.1016/j.ejrh.2020.100711

How to cite: Meißl, G., Zieher, T., and Geitner, C.: Rainfall-runoff reaction controlled by soil moisture thresholds in a small Alpine catchment, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5078, https://doi.org/10.5194/egusphere-egu22-5078, 2022.

14:08–14:15
|
EGU22-2369
|
ECS
|
Mirna Švob et al.

When simulating soil water content (SWC) and dynamics, the reservoir cascade scheme (RCS) approach is considered appropriate in cases when number of parameters for model calibration and validation is limited. This is often the case in Mediterranean karst soils, where due to high rockiness and shallow soil depths it is often difficult to set dense measurement network. In this study, a 1-D model which simulates SWC using RCS approach was developed for a location in central Spain. The soil on the studied site has silt loam texture and is developed on dolomite marbles. The model simulates SWC at daily resolution for six layers in soil that range from 0-50 cm depth, and has three different configurations. Configuration 1 considers only basic RCS module, while configurations 2 and 3 simulate preferential flows in soil as well. Therefore Configuration 2 considers RCS module together with continuous preferential flow module, where between 1 and 5% of available SWC is drained from each soil layer every day. Configuration 3 considers discontinuous preferential flows in addition to two previous modules. Discontinuous preferential flows are active in cases of rainfall events that occur during prolonged dry periods. Simulated SWC values are compared with SWC values measured at five depths in soil, so model parameters are iteratively adjusted to optimize the model results. The simulation produced the best results when implementing Configuration 3: when matrix flow and two kinds of preferential flow are assumed. The model shows that preferential flows could significantly contribute to recharge and should be given more attention in soil hydrological models, especially in karst terrains. It is expected that the model can be implemented in a wide range of locations with karst soils, since it requires limited number of input parameters, but in the same time provides a detailed simulation of soil drainage processes and recharge.

How to cite: Švob, M., Domínguez-Villar, D., and Krklec, K.: Water dynamics in karst soil: Modelling matrix and preferential flow using reservoir cascade scheme approach, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2369, https://doi.org/10.5194/egusphere-egu22-2369, 2022.

14:15–14:22
|
EGU22-10926
Huade Guan et al.

Since the first stoma appeared about 400 million years ago, moisture exchange between lands and the atmosphere extends into the root zone. However, due to its invisibility from the surface, root distribution and its temporal variation are difficult to estimate, which greatly hinders investigation of root zone moisture dynamics, soil-plant water relations, and transpiration modelling. Plant water potential reflects dynamic water condition in vegetation, which is determined by moisture supply in the root zone, atmospheric demand, and plant physiological control. Thus, dynamic water potential can provide a “periscope” to observe root zone hydraulic conditions. Based on this hydraulic connection in the soil-plant-atmosphere continuum (SPAC), plant individuals work very likely as “observation wells” to the whole root zone at predawn, and as “pumping test wells” in daytime. Meanwhile, stable isotopic composition of water in plant xylem approximately reflects the isotopic signature of bulk root accessible moisture. These hydraulic and isotopic root-zone periscopes provide information to estimate root-zone and plant hydraulic states and their dynamics, and hydraulic properties. In this presentation, we will show how this root-zone periscope concept, based on continuous monitoring of plant water potential, sapflow, and/or isotopic composition of xylem water, has been successfully applied in SPAC model development, root water uptake model improvement, transpiration model parameterisation, as well as investigation of ecohydrological separation.

How to cite: Guan, H., Deng, Z., Wang, H., Xu, X., Yang, Y., Liu, N., Luo, Z., Zhang, C., Hutson, J., Zhang, X., He, X., and Simmons, C.: Root-zone “Periscope” and its applications for investigating plant-soil water relations and transpiration modelling , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10926, https://doi.org/10.5194/egusphere-egu22-10926, 2022.

14:22–14:29
|
EGU22-12864
Theresa Blume et al.

Ecohydrological consequences of dry years are difficult to predict. To understand the underlying drivers and responses, extensive monitoring over longer periods of time is a prerequisite. We are here providing an overview of multi-year monitoring of different forest stands in the TERENO Observatory NorthEast Germany. These forest stands include pure oak, beech and pine stands as well as mixed stands and the experimental design also allows the comparison of sites with and without accessibility to groundwater. Monitoring covers a large number of variables with high temporal resolution, such as soil moisture and groundwater dynamics but also sapflow and tree growth. Due to the deep groundwater levels and the high conductivity of the sandy soils, water storage dynamics in the large unsaturated zone and the deeper root zone are of special importance. Soil moisture monitoring therefore extends down to a depth of 2m. We provide an overview of the ecohydrological responses of this forest system to the extreme summer and fall of 2018 as well as the progression in the following years.

How to cite: Blume, T., Güntner, A., Weiler, M., and Heinrich, I.: Ecohydrological responses to a series of dry years at the TERENO Observatory NorthEast Germany, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12864, https://doi.org/10.5194/egusphere-egu22-12864, 2022.

14:29–14:36
|
EGU22-6880
|
ECS
Siang-Heng Wang and Jehn-Yih Juang
14:36–14:43
|
EGU22-5387
Roberto Corona et al.

Soil moisture content influences the partitioning of net radiation into latent and sensible heat fluxes that, in turn, affect the dynamics of the atmospheric boundary layer depth and concomitant generation of precipitation. The interactive effect between soil moisture and precipitation has been found to be stronger in areas where soil moisture temporal variability is enhanced such as in arid and semiarid regions, and in transitional regions between dry and wet climate. For this reason, variability in soil moisture at multiple time scales continues to draw attention in climate science and hydrology. In this work, the soil moisture variability at multiple scales for a typical Mediterranean ecosystem, has been quantified using the spectrum of soil moisture.The case study is the Orroli site in Sardinia (Italy), a typical semi-arid Mediterranean ecosystem which is an experimental site for the ALTOS European project of the PRIMA MED program.The spectrum of root-zone soil moisture content for this Mediterranean ecosystem is analyzed using 14-years of half-hourly measurements. A distinguishing hydro-climatic feature in such ecosystems is that sources (mainly rainfall) and sinks (mainly evapotranspiration) of soil moisture are roughly out of phase with each other. For over 4 decades of time scales and 7 decades of energy, the canonical shape of the measured soil moisture spectrum is shown to be approximately Lorentzian determined by the soil moisture variance and its memory but with two exceptions: the occurrences of a peak at diurnal-to daily time scales and weaker peak at near annual time scales. Model calculations and spectral analysis demonstrate that diurnal and seasonal variations in hydroclimate forcing responsible for variability in evapotranspiration had minor impact on the normalized shape of the soil moisture spectrum. However, their impact was captured by adjustments in the temporal variance. These findings indicate that precipitation and not evapotranspiration variability dominates the multi-scaling properties of soil moisture variability consistent with prior climate model simulations. Furthermore, the soil moisture memory inferred by the annual peak of soil moisture (340 d) is consistent with climate model simulations, while the memory evaluated from the loss function of a linearized mass balance approach leads to a smaller value (50 d), highlighting the effect of weak non-stationarity on soil moisture variability. Spatial variability in infiltration rates introduce some whitening of rainfall temporal auto-correlation recovering a spectral decay in soil moisture spectra consistent with  f2 at sub-weekly time scales, where f is the frequency or inverse time scale.

How to cite: Corona, R., Montaldo, N., and Katul, G. G.: Multiscale analysis of soil moisture variability for a typical semi-arid Mediterranean ecosystem, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5387, https://doi.org/10.5194/egusphere-egu22-5387, 2022.

14:43–14:50
|
EGU22-6685
Ernesto Lopez-Baeza et al.

The Valencia Anchor Station (VAS) was set up by the University of Valencia at the very end of the year 2001 starting its operations on 1st January 2002. Since then, uninterruptedly, the Climatology from Satellites Group (GCS) has developed a constant activity addressed to the difficult task of characterising an area sufficiently large as to also serve as a reference site for the scientific validation of current low and middle spatial resolution remote sensing instruments onboard the missions NASA CERES (Clouds and the Earth’s Radiant Energy System) and SMAP (Soil Moisture Active and Passive), EUMETSAT GERB (Geostationary Earth Radiation Budget), ESA SMOS (Soil Moisture and Ocean Salinity), ESA-EUMETSAT EPS (EUMETSAT Polar System) MetOp, EC-ESA Copernicus Sentinel-1, -2, -3 and is getting ready for the now close ESA-JAXA EarthCARE (Earth Clouds, Aerosols and Radiation Explorer) launch and for the current GNSS-R, -iR terrestrial applications from the Galileo, BeiDou, GPS and GLONASS constellations. All these instruments have in common their middle/large footprint sizes for which a sufficiently large validation area up to about 50 x 50 km2 needs to be robustly equipped and fully characterised from different viewpoints such as soils, vegetation, atmospheric parameters, etc. This presentation shows the fundamentals of the methodologies used for the validation of surface radiation, soil moisture and biophysical vegetation parameters, and a brief summary of the field campaigns developed for the characterisation of the site in the context of the models used and some of the achievements so far obtained. Detailed account of the validation exercises for the different parameters under consideration is also given in different sessions of this EGU 22 Assembly. The paper also emphasises the role of the distributed soil measurements carried out over a large vineyard field in relation to the rest of significant parameters from a dense FAPAR (Fraction of Absorbed Photosynthetically Active Radiation) network and from an eddy-covariance station, together with the complete surface radiation network and land surface and atmospheric temperatures, also provided by the VAS. It is worth noting the role of the collaborative interdisciplinary international teams associated to the Climatology from Satellites Group in the framework of the different Missions and Agencies above mentioned, the qualitative upgrading of the VAS as a GBOV (Ground-Based Observations for the Validation of Copernicus Land Products) supersite and its future prospects by incorporating artificial intelligence and data semantics techniques. The VAS is currently jointly run by LISITT (Integrated Laboratory of Intelligent Systems and Technologies of Traffic Information), a research and development group integrated into the IRTIC (Research Institute on Robotics and Information and Communication Technologies) and by UV-ERS (Environmental Remote Sensing Group) of the Faculty of Physics, both from the University of Valencia. This guarantees the envisaged new developments planned for the VAS to offer data and products in an optimum user-friendly format by using artificial intelligence and data semantics methods.

How to cite: Lopez-Baeza, E., Garcia Rodriguez, D., Albero Peralta, E., Garcia-Celda, A., Asensi Ortega, V., Catalan Alcober, D. J., and Martinez Dura, J. J.: The Valencia Anchor Station: 20 Years of Uninterrupted Scientific Activity on Validation of Low and Middle Resolution Earth Observation Remote Sensing Data and Products, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6685, https://doi.org/10.5194/egusphere-egu22-6685, 2022.

Tue, 24 May, 15:10–16:40

Chairpersons: Anke Hildebrandt, Josie Geris

15:10–15:17
|
EGU22-8625
Ester Carbo et al.

The prediction of spatial and temporal variation of soil water content brings numerous benefits in the studies of soil. However, it requires a considerable number of covariates to be included in the study, complicating the analysis. Integrated nested Laplace approximations (INLA) with stochastic partial differential equation (SPDE) methodology is a possible approach that allows the inclusion of covariates in an easy way.

The current study has been conducted using INLA-SPDE to study soil moisture in the area of the Valencia Anchor Station (VAS), soil moisture validation site for the European Space Agency SMOS (Soil Moisture and Ocean Salinity). The data used were collected in a typical ecosystem of the semiarid Mediterranean conditions, subdivided into physio-hydrological units (SMOS units) which presents a certain degree of internal uniformity with respect to hydrological parameters and capture the spatial and temporal variation of soil moisture at the local fine scale. The use of the INLA-SPDE methodology presents the possibility to analyze the significance of different covariates having spatial and temporal effects and has allowed us to fit spatial and temporal hierarchical models that are too complicated to be fitted by maximum likelihood methods.

The models allow to analyze the influence of hydrodynamic properties on VAS soil moisture (texture, porosity/bulk density and soil organic matter and land use) filtering out the effect of spatial and temporal variation. With the goal of understanding the factors that affect the variability of soil moisture in the SMOS pixel (50 km x 50 km), five states of soil moisture are proposed (moisture values range from values near saturation to moisture values near wilting point). Regarding the different models for the different Soil Moisture states, a general pattern was detected where both porosity and organic matter are two significant elements in all the cases. The model with all covariates and spatial effect has the lowest DIC value. In addition, the correlation coefficient was close to 1 for the relationship between observed and predicted values.

The findings of this study demonstrate an advancement in that framework, demonstrating that it is faster than previous methodologies, provides significance of individual covariates, is reproducible, and is easy to compare with models. The use of this methodology permits to design more efficient sampling campaigns for future SMOS missions. In addition, it also allows to construct soil moisture maps in a more sensible and efficient way.

 

How to cite: Carbo, E., Juan, P., Añó, C., Chaudhuri, S., Diaz-Avalos, C., and López-Baeza, E.: Modeling soil moisture at the Valencia Anchor Station (VAS), Eastern Spain., EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8625, https://doi.org/10.5194/egusphere-egu22-8625, 2022.

15:17–15:24
|
EGU22-10703
Pierre Ferreira do Prado et al.

The urgency on detailing surface soil moisture content worldwide, especially in agricultural soils, is well established. The efforts of the European Space Agency (ESA), regarding the Sentinel-1 mission, facilitated a synthetic aperture radar (SAR) sensor that, in conjunction with machine-learning-based methods, can be useful  and fruitful responding to this technological demand. This paper aims at  exploring the possibility of the Valencia Anchor Station, near the city of Valencia, Eastern Spain, to provide 1 km x 1 km soil moisture products using its ground-based reference meassurements. The results suggest that, among several options, an artificial neural network using the Levenberg-Maquardt learning algorithm, based on soil moisture recovery from Sentinel-1 SAR radar data should be preferred for this site. Among other options, the so called fine-tree regression also presented relevant results. All of this allows us to gain insights into the complexity of the relation SAR´s backscatter – surface soil moisture relation for this site, also aiming at the potential extension of this knowledge to other sites where Sentinel-1 data is available, for example, in framework of the Joint Research Center "Ground-Based Observations for Validation (GBOV) of Copernicus Global Land Products" Project.

How to cite: Ferreira do Prado, P., Silveira Duarte, I. C., and Lopez-Baeza, E.: Using Matlab´s Supervised Machine-Learning Tools to Retrieve Surface Soil Moisture from Sentinel-1 SAR Data Over the Valencia Anchor Station (Spain), EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10703, https://doi.org/10.5194/egusphere-egu22-10703, 2022.

15:24–15:31
|
EGU22-9849
|
ECS
Ehsan Jalilvand et al.

Irrigation is the largest human intervention in the water cycle that can modulate climate extremes. Despite the importance of irrigation, global irrigation water use (IWU) remains largely unknown. Microwave remote sensing offers a low-cost solution to quantify IWU by monitoring the changes in the soil moisture caused by irrigation. However, high-resolution satellite soil moisture data has fewer observations and might miss irrigation events. This study tests a method to quantify the IWU by assimilating high resolution (~1km), but less frequent SMAP-Sentinel1 (SMAP-S1) remotely sensed soil moisture with a land surface model. We use a particle batch smoother (PBS) to assimilate the SMAP-S1 soil moisture data with the VIC (4.2d) land surface model. It is important to remove the biases between the model and the satellite observations prior to the data assimilation, so we also evaluate the impact of model calibration during the irrigation or rainy season on the quantified irrigation. Moreover, we conducted a synthetic experiment in which the uncertainty due to the noise in assimilated soil moisture data, the frequency of the satellite observations, and the knowledge of irrigation timing was investigated. We will present the results of these studies.

How to cite: Jalilvand, E., Abolafia-Rosenweig, R., Das, N., and Tajrishy, M.: Quantifying the irrigation water use by assimilating SMAP-Sentinel1 1km soil moisture data using a particle batch smoother approach, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9849, https://doi.org/10.5194/egusphere-egu22-9849, 2022.

15:31–15:38
|
EGU22-10810
|
ECS
Sungmin Oh et al.

Soil moisture information is valuable for a wide range of applications in various fields such as hydrology, agriculture, and climate. Although spatially continuous soil moisture data can be obtained from satellite observations or model simulations, each type of data has its own uncertainty and bias. In this study, we use machine learning as a hydrologic model and generate a gridded soil moisture dataset—SoMo.ml—that can complement existing soil moisture datasets (O and Orth, 2021). We train a Long Short-Term Memory neural network model using in-situ measurements to extrapolate daily soil moisture dynamics in space and in time. The first version of the data, SoMo.ml v1, provides multilayer soil moisture (0-10cm, 10-30cm, and 30-50 cm) at 0.25° and daily resolutions for the period 2000-2019; it has been actively used for drought analysis, data comparison, and other relevant research. The dataset is freely available from https://www.bgc-jena.mpg.de/geodb/. Given the growing needs for this unique soil moisture dataset, SoMo.ml v2 is currently under development, which aims to provide soil moisture data over Europe with a higher spatial resolution (0.1°). In this presentation, we will introduce the SoMo.ml datasets and show examples of data applications in other studies.

 

Reference: O and Orth, Global soil moisture data derived through machine learning trained with in-situ measurements. Sci Data, (2021). https://doi.org/10.1038/s41597-021-00964-1

How to cite: Oh, S., Orth, R., and Park, S. K.: Machine learning-based multilayer soil moisture datasets: SoMo.ml, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10810, https://doi.org/10.5194/egusphere-egu22-10810, 2022.

15:38–15:45
|
EGU22-11462
|
ECS
Junhan Zeng et al.

Soil moisture (SM) plays an important role in hydrological processes and land-atmospheric interactions, and serves as an important boundary condition for the weather forecasting and climate modeling. Influenced by global environment change, SM changes significantly at local scales which rises the great need of high-resolution SM products to provide locally relevant information. However, the three SM estimation approaches, namely in-situ observation, remote sensing retrieval and land surface modeling, all have their disadvantages. Although recent works produce a combined SM products by merging the in-situ observations and several land surface simulation products, the long-term high-resolution SM product integrating multivariate data including remote sensing products is still lacking. In this study, high-resolution land surface modeling, high-resolution remote sensing products and SM observations from more than 2000 stations will be combined to generate spatially continuous and temporally complete soil moisture data in China by using the random forest algorithm. We first performed land surface simulations by using the Conjunctive Surface-Subsurface Process version 2 (CSSPv2) model forced by three meteorological forcings including the China Meteorological Administration Land Data Assimilation System version 2.0 (CLDASv2.0), ERA5 and GLDASv2.1. The validations over 2090 in situ stations during 2012–2017 showed that CLDASv2.0/CSSPv2 soil moisture simulation performed better than ERA5 and GLDASv2.1 reanalysis products, with an increased correlation of 26%–68% and reduced errors of 14%–24% at the daily time scale. The improvements mostly originate from the use of an advanced LSM because CLDASv2.0/CSSPv2 only increased the correlation by 5%–35% and decreased the errors by up to 9% when compared with ERA5/CSSPv2 and GLDASv2.1/CSSPv2. Due to the high accuracy of CLDASv2.0/CSSPv2 product, it will be used as a background to fuse the in-situ observations and satellite remote sensing soil moisture. The 70% of the observation site data, remote sensing products and CLDASv2.0/CSSPv2 product will be used to train the random forest model and generate a high resolution soil moisture product from 2008 to 2017, and another 30% of the site data will be used to evaluate the accuracy of the results. Such a SM product can describe the spatial and temporal distribution characteristics of soil moisture heterogeneity more accurately, and thus provide sufficient data support for scientific research and social development.

How to cite: Zeng, J., Xing, Y., Ji, P., and Shi, C.: Multi-source soil moisture data fusion based on high-resolution land surface simulation and machine learning, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11462, https://doi.org/10.5194/egusphere-egu22-11462, 2022.

15:45–15:52
|
EGU22-11052
|
Highlight
Dominik Michel et al.

Climate projections indicate an increasing risk of dry and hot episodes in Central Europe, including in Switzerland. However, models display a large spread in projections of changes in summer drying, highlighting the importance of related observations to evaluate climate models and constrain projections. Land hydrological variables play an essential role for these projections. This is particularly the case for soil moisture and land evaporation, which are directly affecting the development of droughts and heatwaves in both present and future.
The recent 2020 spring as well as 2015 and 2018 summer droughts in Switzerland have highlighted the importance of monitoring and assessing changes of soil moisture and land evaporation, which are strongly related to drought impacts on agriculture, forestry, and ecosystems.
The only Switzerland-wide soil moisture monitoring programme currently in place is the SwissSMEX (Swiss Soil Moisture Experiment) measurement network. It was initiated in 2008 and comprises 19 soil moisture measurement profiles at 17 different sites (grassland, forest and arable land). Since 2017, seven grassland SwissSMEX sites are complemented with land evaporation measurements from mini-lysimeters.
Here we analyze long-term satellite-based drought parameters, namely ASCAT Soil Water Index (SWI) derived from an H SAF test data set and LSA SAF Meteosat land evaporation products. We compare the satellite-based datasets with the SwissSMEX in-situ measurements of soil moisture and lysimeter land evaporation. The comparison of in-situ soil moisture and land evaporation data with the satellite parameters shows strong agreement in terms of anomalies. SWI indicates high correlations of 0.6 to 0.8 with in-situ measurements. The Meteosat land evaporation products strongly agree with measurements, with correlations of 0.7 and 0.9 for potential and actual land evaporation, respectively (Burgstall et al.).
These analyses provide useful insights in order to provide near-real time monitoring, enhance process understanding and for a better preparedness for future droughts.

References:
Burgstall, A. et al., Climatological drought monitoring in Switzerland using EUMETSAT SAF satellite products, Remote Sensing, in preparation.

How to cite: Michel, D., Burgstall, A., Hirschi, M., Anke Duguay-Tetzlaff, A., and Seneviratne, S. I.: Assessing the suitability of remote sensing estimates of soil moisture and land evaporation in Switzerland for a better preparedness for projected drying trends, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11052, https://doi.org/10.5194/egusphere-egu22-11052, 2022.

15:52–15:55
Wrap-up