02708naa a2200325 a 450000100080000000500110000800800410001910000180006024501130007826000090019152018140020065000080201465000260202265000180204865000240206665000170209065000220210770000200212970000170214970000170216670000190218370000260220270000160222870000170224470000130226170000200227470000180229470000220231277300480233411318912022-03-15       2022    bl uuuu     u00u1 u    #d1 aMENDES, W. S.  aThe Brazilian soil Mid-infrared Spectral LibrarybThe Power of the Fundamental Range.h[electronic resource]  c2022  aBrazil plays an important role in global agricultural production and environmental politics, and thus a national spectral library can bring light to new initiatives on soil monitoring. The middle infrared spectral range (MIR, 4000?600 cm− 1) includes the fundamental reflectance of soil organic and mineral components and thus can bring important information on soil characterization. This work aimed to build a national soil spectral library from the MIR, relate its descriptive and quantitative potential to soil assessment. We characterized up to 4309 soil samples from different depths (0?20, 40?60, 80?100, and 100?120 cm) including particle size distribution and chemical analyses (2963?4309 samples), and mineralogical analyses (278?861 samples). Unsupervised classification distinguished five soil spectra classes with distinct absorption shapes and features. The modeling results showed that MIR can accurately predict soil physicochemical attributes and revealed a strong association with environmental and geographical variables. There were no significant differences between raw and pre processed MIR spectral data in predicting soil attributes. In general, local spectral models presented better results than national ones, as expected. Consistent results were achieved for several soil attributes, such as mineralogy (R2 = 0.68?0.87), clay fraction (R2 = 0.55?0.91), organic carbon (R2 = 0.77?0.94), and base saturation (R2 = 0.68?0.87). It was possible to create a MIR spectral classification based only on the spectra and assess its relation with environmental and geographical variables. The information generated in this work may be the basis for future research as MIR spectra contain fundamental soil signatures that can be practically used for agronomic and environmental decisions.  aMIR  aProximal soil sensing  aSoil analysis  aSoil classification  aSoil quality  aSoil spectroscopy1 aDEMATTÊ, J. A.1 aROSIN, N. A.1 aTERRA, F. S.1 aPOPPIEL, R. R.1 aURBINA-SALAZAR, D. F.1 aBOECHAT, L.1 aSILVA, E. B.1 aCURI, N.1 aSILVA, S. H. G.1 aSANTOS, U. J.1 aVALLADARES, G. S.  tGeoderma, AMSTERDAMgn. 415, p. 1-12, 2022.