01764nam a2200217 a 450000100080000000500110000800800410001910000220006024501300008226001410021252009950035365000280134865000150137665000180139165000260140965300230143570000210145870000230147970000190150270000250152121400332022-02-24       2022    bl uuuu     u00u1 u    #d1 aVASQUES, G. de M.  aManagement zones from proximal soil sensors capture within-field soil property and terrain variations.h[electronic resource]  aIn: PEDOMETRICS BRAZIL, 2., 2021, Rio de Janeiro. Annals [...]. Rio de Janeiro: Embrapa Solos, 2022. Não paginado. Evento online.c2022  aThe objectives are to delineate soil management zones from soil proximal sensor data, and compare soil property values among zones in a 72-ha crop field in southeastern Brazil. Apparent electrical conductivity (aEC) and magnetic susceptibility (aMS), and equivalent Th (eTh) and U (eU) were measured across the field by a Geonics EM38-MK2 and a Medusa MS1200 sensors, respectively. These properties were kriged and used as input for delineating three management zones by fuzzy k-means clustering. Soil properties were measured at 0-10 cm at 72 sites, and their means were compared among the zones. Soil clay, organic C and exchangeable Ca and Mg vary significantly among the zones, according to Brown-Forsythe and Games-Howell tests (p=0.05), while pH, available P and exchangeable K do not. Zone delineation from proximal sensor data constitutes an efficient data-driven approach to separate the field into meaningful parts for soil, irrigation and crop management based on soil variation.  aElectrical conductivity  aGeophysics  aGeostatistics  aPrecision agriculture  aGamma radiometrics1 aRODRIGUES, H. M.1 aOLIVEIRA, R. P. de1 aHERNANI, L. C.1 aTAVARES, S. R. de L.