Kurt C. Kornelsen

Root‐zone soil moisture estimation using data‐driven methods

The soil moisture state partitions both mass and energy fluxes and is important for many hydro-geochemical cycles, but is often only measured within the surface layer. Estimating the amount of soil moisture in the root-zone from this information is difficult due to the nonlinear and heterogeneous nature of the various processes which alter the soil moisture state. Data-driven methods, such as artificial neural networks (ANN), mine data for nonlinear interdependencies and have potential for estimating root-zone soil moisture from surface soil moisture observations. To create an ANN root-zone model that was nonsite-specific and physically constrained, a training set was generated by forcing HYDRUS-1D with meteorological observations for different soil profiles from the unsaturated soil hydraulic database. Ensemble ANNs were trained to provide soil moisture at depths of 10, 20, and 50 cm below the surface using surface soil moisture observations and local meteorological information. Insights into the processes represented by the ANNs were derived from a clamping sensitivity analysis and by changing the ANNs input data. Further model testing based on synthetic soil moisture profiles from three McMaster Mesonet and three USDA soil climate analysis network sites suggests that ANNs are a flexible tool capable of predicting root-zone soil moisture with good accuracy. It was found that ANNs could well represent soil moisture as estimated by HYDRUS-1D, but performance was reduced in comparison to in situ soil moisture observations outside the training conditions. The transferability of the model appears limited to the same geographic region.

Comparison of Interpolation, Statistical, and Data-Driven Methods for Imputation of Missing Values in a Distributed Soil Moisture Dataset

AbstractMissing values in in situ monitoring data is a problem often encountered in hydrologic research and applications. Values in a data set may be missing because of sensor error or failure of data recording devices. Whereas various imputation techniques have focused on hydrometeorological data, very few studies have investigated gap-filling methods for soil moisture data. This paper aims to fill that gap by investigating well-established statistical and data-driven methods for infilling missing values in a high resolution, soil moisture time series. Since 2006, the authors collected hourly soil moisture data in the Hamilton-Halton Watershed, Southern Ontario, Canada at four research sites. Each site contained nine stations with time domain reflectometry (TDR) soil sensors at six soil depths. From these distributed data sets, the authors removed values randomly (∼5%) and systematically (∼20%) from the data to evaluate the effectiveness of the monthly average replacement (MAR), soil layer relative diffe...

Potential of bias correction for downscaling passive microwave and soil moisture data

Passive microwave satellites such as Soil Moisture and Ocean Salinity or Soil Moisture Active Passive observe brightness temperature (TB) and retrieve soil moisture at a spatial resolution greater than most hydrological processes. Bias correction is proposed as a simple method to disaggregate soil moisture to a scale more appropriate for hydrological applications. Temporal stability of soil moisture and TB was demonstrated at the Little Washita and Little River Experimental Watersheds using in situ observations and the Community Microwave Emissions Model. Decomposition of the mean square difference (MSD) between the watershed average soil moisture and TB showed that bias was a major contributor to differences between watershed average and local-scale soil moisture and TB, particularly at sites with high MSD. The mean RMSD between watershed average and local soil moisture was 0.04 m3 m−3 and 0.06 m3 m−3 at Little River and Little Washita, respectively. Following a simple bias correction the RMSD was reduced to 0.03 m3 m−3 at both sites. Considering multiple incidence angles at both horizontal and vertical polarization, bias correction of watershed average TBV reduced the RMSD by approximately 75% and 45% and TBH RMSD by 68% and 36% for Little River and Little Washita, respectively, at all incidence angles. Therefore, at subsatellite grid scale, bias correction can be considered a viable technique for downscaling passive microwave observations and soil moisture retrievals.

Design of an Optimal Soil Moisture Monitoring Network Using SMOS Retrieved Soil Moisture

Many methods have been proposed to select sites for grid-scale soil moisture monitoring networks; however, calibration/validation activities also require information about where to place grid representative monitoring sites. In order to design a soil moisture network for this task in the Great Lakes Basin (522 000 km2), the dual-entropy multiobjective optimization algorithm was used to maximize the information content and minimize the redundancy of information in a potential soil moisture monitoring network. Soil moisture retrieved from the Soil Moisture and Ocean Salinity (SMOS) mission during the frost-free periods of 2010-2013 were filtered for data quality and then used in a multiobjective search to find Pareto optimum network designs based on the joint entropy and total correlation measures of information content and information redundancy, respectively. Differences in the information content of SMOS ascending and descending overpasses resulted in distinctly different network designs. Entropy from the SMOS ascending overpass was found to be spatially consistent, whereas descending overpass entropy had many peaks that coincided with areas of high subgrid heterogeneity. A combination of both ascending and descending overpasses produced network designs that incorporated aspects of information from each overpass. Initial networks were designed to include 15 monitoring sites, but the addition of network cost as an objective demonstrated that a network with similar information content could be achieved with fewer monitoring stations.

Entropy Applications to Water Monitoring Network Design: A Review

Having reliable water monitoring networks is an essential component of water resources and environmental management. A standardized process for the design of water monitoring networks does not exist with the exception of the World Meteorological Organization (WMO) general guidelines about the minimum network density. While one of the major challenges in the design of optimal hydrometric networks has been establishing design objectives, information theory has been successfully adopted to network design problems by providing measures of the information content that can be deliverable from a station or a network. This review firstly summarizes the common entropy terms that have been used in water monitoring network designs. Then, this paper deals with the recent applications of the entropy concept for water monitoring network designs, which are categorized into (1) precipitation; (2) streamflow and water level; (3) water quality; and (4) soil moisture and groundwater networks. The integrated design method for multivariate monitoring networks is also covered. Despite several issues, entropy theory has been well suited to water monitoring network design. However, further work is still required to provide design standards and guidelines for operational use.

Application of SMOS Soil Moisture and Brightness Temperature at High Resolution With a Bias Correction Operator

The assimilation of soil moisture and brightness temperature (TB) are expected to improve the modeling of land surface processes, but are only available at a resolution that is far coarser than the scale of many hydrological processes. Due to systematic differences between model states and satellite observations, a bias correction operator is a necessary step in land data assimilation schemes and was evaluated as a method to disaggregate coarse-scale satellite observations to fine-scale model grid cells (~800 m). This was done by coupling the Modélisation Environmentale Communautaire-Surface Hydrology (MESH) Hydrological Land-Surface Scheme to the Community Microwave Emissions Model (CMEM) to simulate soil moisture and TB. By comparison, MESH-CMEM was found to be in good agreement with observations from the Soil Moisture and Ocean Salinity (SMOS) satellite at the scale of SMOS data products (R ≈ 0.55), with simulated TB being better correlated than soil moisture retrievals. Following bias correction, TB and soil moisture retrievals at 800-m resolution had comparable performance to coarse-resolution SMOS data. Bias correction of TB was more reliable than soil moisture. These findings indicate that both TB and soil moisture retrievals can be assimilated in a land surface model at moderate-to-high resolution with a simple observation operator.

Design of an optimum soil moisture monitoring network using SMOS

The monitoring of in situ soil moisture is an important task for hydro-climatic forecasting and the calibration/validation of satellite based soil moisture missions such as the Soil Moisture and Ocean Salinity (SMOS) mission. Many techniques have been explored to select a single site for a monitoring station that is representative of watershed characteristics. To objectively design an entire soil moisture monitoring network, this paper presents the dual-entropy multi-objective optimization (DEMO) system using soil moisture data retrieved from SMOS. The resulting optimal networks are created by maximizing the information content and minimizing the redundancy of information in each potential network.

Data-based disaggregation of SMOS soil moisture

The Soil Moisture and Ocean Salinity (SMOS) microwave radiometer is used to retrieve surface soil moisture with a grid resolution of 15 km. Due to various contributing factors SMOS soil moisture is known to have bias with respect to in situ soil moisture measurements and land surface models. For this reason it is common practice to match the cumulative distribution function (CDF) of retrieved soil moisture prior to analysis. Using the concept of temporal stability this study demonstrates that CDF matching is effective for correcting the bias at both grid and sub-grid scales with minimal impact on the time in-variant component of SMOS error.