The data set was used to predict soil organic carbon and soil moisture in the vertical as well as horizontal domain, i.e. volumetrically, by using a weighted conditioned Latin Hypercube Sampling design for selecting the calibration samples and the geophysical covariates derived from electromagnetic induction (EMI) and a gamma-ray spectrometer with different intercoil spacings and, thus, different penetration depths and footprints of the signal. The study site is an agricultural field of 58 ha about 70 km north of Leipzig, Saxony-Anhalt, Germany. Present soil types are Gleysols and Gleyic Cambisols consisting of alluvial loam (loam and clay)over Holocene sediments of fluvial sand (LAGB, 2014). The geophysical measurements were recorded with three EMI sensors (EM38-DD, Geonics Limited, ON, CA; CMD-Explorer and CMD-Mini-Explorer, both GF Instruments, CZ) and a gamma-ray spectrometer (GS CAR, GF Instruments, CZ) in August 2016. EMI sensors measure the apparent electric conductivity (ECa in mSm-1). All EMI sensors captured 5 records s-1 in any dipole orientation. The gamma-ray spectrometer is equipped with a 4l NaI(Tl)-crystal and automatic peak-stabilization to measure the concentration of potassium (40K), uranium (238U) and thorium (232Th). The device has 512 channels with an energy range from 100 keV to 3 MeV. Measurements were captured every 5 seconds. 40K, 238U and 232Th were measured as counts per second. The concentration of 40K (in %) and 238U and 232Th (both in ppm) was calculated corresponding to the decay rate at specific energy levels. The concentration of 40K, 238U and 232Th was used to calculate the dose rate (nGyh 1; IAEA, 2003). For determination of soil organic carbon, the samples were dried at 40 °C for 24 h, sieved (<2 mm), ground and root fragments were removed. Total carbon was determined with dry combustion using an ELTRA CHS-580A Helios analyser (ELTRA GmbH, GER). Soil moisture was measured gravimetrically with drying at 90 °C for 24 h.

Supplement to: Rentschler, Tobias; Werban, Ulrike; Ahner, Mario; Behrens, Thorsten; Gries, Phillipp; Scholten, Thomas; Teuber, Sandra; Schmidt, Karsten (2020): 3D mapping of soil organic carbon content and soil moisture with multiple geophysical sensors and machine learning. Vadose Zone Journal, 19(1)