• Media type: Electronic Conference Proceeding
  • Title: Geophysics-based subsurface conceptualization for improved prediction of plant productivity beyond the field scale
  • Contributor: Brogi, Cosimo [Author]; Huisman, Johan Alexander [Author]; Herbst, Michael [Author]; Weihermüller, Lutz [Author]; Vereecken, Harry [Author]
  • Published: Forschungszentrum Jülich: JuSER (Juelich Shared Electronic Resources), 2018
  • Published in: European Geosciences Union General Assembly 2018, EGU 2018, Vienna, Austria, 2018-04-09 - 2018-04-13
  • Language: English
  • Origination:
  • Footnote: Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
  • Description: A precise and reliable characterization of the shallow subsurface is vital in hydrological and agronomical modelling.Non-invasive geophysical methods such as electromagnetic induction (EMI) measurements in combinationwith soil sampling can provide subsurface information with a high spatial resolution. In principle, such geophysicaldata allow a generalization of the subsurface in homogeneous areas that share similar patterns of soil structuralorganization (layering and texture) using a reduced amount of resources. However, it is still challenging to deriverelevant subsurface information from such geophysical data sets, and the added value of such high-resolution soilinformation for the analysis of patterns in plant productivity has also not been investigated yet. In this study, weused an image classification method to classify high-resolution multi-configuration EMI measurements obtainedin an agricultural area of 102 ha near Selhausen (Germany) where the subsurface structure is known to affect cropproductivity during water stress periods. EMI measurements were collected in 2016 within a few days after harvestof each field and were automatically filtered, temperature corrected, and interpolated onto a 1 m resolutiongrid. The EMI data indicated four main sub-areas with characteristic subsurface heterogeneity and typical impacton plant productivity patterns. To delineate areas with similar subsurface structures, we stacked the ECa mapsobtained with different coil configurations in a single multiband image that was subsequently classified using animage classification methodology. In a second step, we selected one hundred soil sampling locations within thestudy area and obtained soil profile descriptions with type, depth, thickness, and texture of all soil horizons up to2 m depth. By combining the EMI and soil data, typical soil profiles with soil textural information were assignedto each of the classes obtained from the classification of EMI data. The proposed methodology was effective inproducing a high resolution ...
  • Access State: Open Access