Dados do Trabalho

Título

Detection of sand subfractions using bare soil reflectance data modeling

Resumo

The sand fraction of the soil is essential for soil functions because this attribute can affect drainage and water retention, in addition to functioning as a soil buffer against evaporation. Despite these, there are few studies about sand fractions, with the clay receiving more attention due its relation with soil chemical attributes. This work aims to evaluate the potential of Landsat and Sentinel satellites to predict the sand fraction. We created a bare soil image for Landsat and Sentinel in all agricultural areas in Brazil using images from 2017 to 2022. . We sampled the reflectance bands of these images for 3,541 samples (0-20 cm layer) with coarse and fine sand subfractions determined by wet laboratory analysis. We depict the spectral behavior of bands and used the cubist algorithm to estimate its content. . The reflectance values were inversely proportional to the particle size predominant on sand fraction. For Landsat, it was found R2 of 0.21 and 0.13 for fine and coarse sand, while for Sentinel it was found R2 of 0.14 for both subfractions. The results indicated a potential to assist this important fraction of soil and its usefulness in soil mapping and water dynamics. However, more studies are necessary to improve the accuracy of predictions.   

Palavras-chave

soil texture; pedometrics; remote sensing; soil health.

Instituição financiadora

The State of São Paulo Research Foundation (FAPESP) 2021/05129-8.

Agradecimentos

The authors thanks GeoCis Research Group for sampling and analyses.