Sporala red del conocimiento
EN
Universidad Nacional de Colombia

Bogotá - Ciencias Agrarias - Maestría en Ciencias Agrarias · 2020

Spatial modeling of soil organic carbon with NIR and environmental covariates in the municipality of Villavicencio

Soil is the largest reservoir of organic carbon in the biosphere after the ocean. Soil organic carbon accumulation is controlled by several natural and anthropogenic factors including climate, relief, vegetation and intrinsic soil properties. The interaction between these factors significantly modifies the impact of carbon concentration in the atmosphere under the response of CO2 fluxes from the soil. Assessing soil organic carbon content (%) for large areas is costly and time consuming due to its large spatial variability. Near-infrared reflectance spectroscopy is a fast and inexpensive tool to measure this property. In the present study, spatial regression modeling of some environmental covariates was used to estimate topsoil organic carbon (0-30 cm) in the Depositional piedmont of the municipality of Villavicencio (Meta). The georeferenced field observations were cross-referenced with raster information of 15 terrain parameters, climatic data, normalized vegetation index, sand, clay and silt content and soil spectra. The exploratory analysis of the data showed positive spatial autocorrelation for the study area, so it was necessary to consider them in the modeling. Elevation, valley depth, flow direction, vegetation index, soil spectra, and sand and clay contents were the most important predictors of carbon in the study area. With this proposal, the carbon in each of the polygons was estimated, which can be useful for taking actions in those regions where soil quality is compromised (Texto tomado de la fuente).

Texto completo 63 páginas con texto de 65

Leer la tesis completa Ficha en el repositorio

Contenido

Tesauro de su biblioteca