2.5D SPATIO-TEMPORAL MAPPING OF SOIL ORGANIC CARBON STOCK IN BRAZIL

• Grant number: 21/10063-6
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: April 01, 2022
• End date: September 02, 2026
• Field of knowledge: Agronomical Sciences - Agronomy - Soil Science
• Principal Investigator: José Alexandre Melo Demattê
• Grantee: Nícolas Augusto Rosin
• Host Institution: Escola Superior de Agricultura Luiz de Queiroz (ESALQ). Universidade de São Paulo (USP). Piracicaba , SP, Brazil
• Associated scholarship(s): 23/11467-9 - Digital mapping of spatio-temporal dynamic of soil organic carbon stock in Brazil and world's soil potential productivity, BE.EP.DR

Abstract

Soil is the largest reservoir of carbon in earth surface. Soil organic carbon (SOC) is fundamental to soil quality and crop productivity, being directly associated with climate change mitigation and food security. Brazil has an expressive agricultural production area and a great potential for SOC sequestration. However, the spatio-temporal distribution of the SOC stock in the territory is poorly known, making difficult to implement public policies for low-C agriculture. The objective of this research is to map the 2.5D spatio-temporal distribution (present, past and future) of the SOC in brazilian soils with detailed resolution for layers 0-5, 5-15, 15-30, 30-60 and 60-100 cm, and identify the main factors controlling their dynamics through the use of different covariates derived from remote sensing and modeling strategies. Three mapping approaches will be tested for each 5-year period in the present and past: 1) Calibration and validation of the SOC predictive model for each period; 2) Calibration of a model in the recent period (2019-2023), which will be used to predict the SOC for each period; 3) Calibration of a unique model, which will be used to predict the SOC for each period. Several covariates, static (soil and relief attributes) and dynamic (climate, land use and land cover, vegetation indices, reflectance of exposed soil and vegetation and frequency of exposed soil) in the time, obtained from remote sensing, will also be tested. The models for future predictions will follow the approach 2 and will rely on obtained/simulated covariates for future scenarios. Final maps will be validated with existing maps and sampling points. It is expected to obtain and make available inedited maps, with detailed resolution, of the COS stock with a time interval of 5 years between 1984 and 2023 and of 20 years between 2021 and 2100 for the Brazilian territory, which can serve as subsidy for public policies.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

ROSIN, NICOLAS AUGUSTO; DE MELLO, DANILO CESAR; BONFATTI, BENITO R.; HARTEMINK, ALFRED E.; FERREIRA, TIAGO O.; SILVERO, NELIDA E. Q.; POPPIEL, RAUL ROBERTO; MENDES, WANDERSON DE S.; VELOSO, GUSTAVO VIEIRA; FRANCELINO, MARCIO ROCHA; et al. Mapping soil thickness using a mechanistic model and machine learning approaches. CATENA, v. 249, p. 17-pg., 2024-12-10. (14/22262-0, 21/10063-6, 21/05129-8)

ROSIN, NICOLAS AUGUSTO; DEMATTE, JOSE A. M.; DE CARVALHO, HUDSON WALLACE PEREIRA; RODRIGUEZ-ALBARRACIN, HEIDY SOLEDAD; ROSAS, JORGE TADEU FIM; NOVAIS, JEAN JESUS; DALMOLIN, RICARDO S. D.; ALVES, MARCELO RODRIGO; FALCIONI, RENAN; TZIOLAS, NIKOLAOS; et al. Spatializing soil elemental concentration as measured by X-ray fluorescence analysis using remote sensing data. CATENA, v. 240, p. 14-pg., 2024-03-23. (14/22262-0, 21/10063-6, 21/05129-8)