Skoči na glavni sadržaj

Agriculturae Conspectus Scientificus, Vol. 83 No. 1, 2018.

Izvorni znanstveni članak

Estimation of Soil Properties Based on Soil Colour Index

Jaroslav Novák orcid id orcid.org/0000-0002-0808-7863 ; Department of Agrosystems and Bioclimatology, Mendel University in Brno, Zemědělská 1, 613 00 Brno Czech Republic
Vojtěch Lukas ; Department of Agrosystems and Bioclimatology, Mendel University in Brno, Zemědělská 1, 613 00 Brno Czech Republic
Jan Křen ; Department of Agrosystems and Bioclimatology, Mendel University in Brno, Zemědělská 1, 613 00 Brno Czech Republic

Puni tekst: engleski pdf 631 Kb

str. 71-76

preuzimanja: 921

citiraj

Sažetak

Knowledge on soil properties is an important aspect in the implementation of precision agriculture. For this study was used an image taken by Sentinel 2 depicting fields of one farm with 8,261 ha in the South Moravian region of the Czech Republic. For the determination of soil properties, soil samples were taken at a density of 1 sample per 3 hectares and analyzed by Mehlich III methodology. The content of available nutrients phosphorus, potassium, magnesium and calcium have been determined together with soil pH, soil texture and sand. The specified sampling revealed high variability for phosphorus, potassium and calcium. Lower variability has been observed with magnesium and pH. An identification of bare soil area without vegetation cover was tested by different threshold values of Normalized vegetation difference index (NDVI) (0.15 – 0.3). The correlations between the multispectral bands and the soil properties were weak. In the analysis of soil samples was detected positive correlation (r = 0.505) between soil texture and Colour Index (CI). In area was found a negative correlation between CI and Ca (r = -0.618), then between CI and pH (r = -0.504). Weak correlation were found between CI, phosphorus and magnesium. At the level of lower NDVI values (0.16 - 0.15) we found correlation between CI and the sand content. The observed level of correlation found in the data of remote sensing can predict some soil properties in fields that have not been subjected to soil sampling and facilitate learning about soil properties for decisions in precision agriculture.

Ključne riječi

precision agriculture; soil properties; Mehlich III; Sentinel 2; NDVI; Colour Index

Hrčak ID:

197111

URI

https://hrcak.srce.hr/197111

Datum izdavanja:

22.3.2018.

Posjeta: 1.723 *