Mapeamento de C e P com base na mineralogia estimada por Vis-NIR-SWIR
Carregando...
Data
2021-03-10
Autores
Braz, Kathleen Fernandes
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O uso irracional dos solos tem causado desequilíbrios em todo o ecossistema. A influência em outros componentes do meio, como na água, no ar e nos ciclos biogeoquímicos dos elementos, já é notória. O presente trabalho teve como objetivo estimar os estoques de C e P em solos do Planalto Ocidental Paulista, a partir da mineralogia da fração argila dada por espectroscopia de reflectância difusa, e estabelecer e mapear áreas específicas de manejo. A área de estudo foi o Planalto Ocidental Paulista, que ocupa cerca de 13 milhões de ha. Foram coletadas 600 amostras para realização de análises físicas, mineralógicas e espectrais. Foram avaliados: a mineralogia, química, granulometria e física dos solos, os estoques de C e P, e os isótopos de C, em amostras estratificadas, com base na geologia, geomorfologia e pedologia da área. Todos os dados foram submetidos a análises estatísticas e geoestatísticas. Os resultados obtidos permitiram obter os valores de referência de estoque de C e P, baseados na composição mineralógica do solo. Solos de basalto apresentam os maiores conteúdos de C e estoques do elemento, quando comparados aos solos de arenito. Solos de arenito, em formas da paisagem altamente dissecadas apresentaram os maiores estoques de P, seguido pelos solos de basalto. É possível gerar mapas de estoques de C e P, e mapas de conteúdo de C a partir de diferentes análises matemáticas. A aprendizagem de máquina permite boa estimativa dos conteúdos de C para diferentes geologias. Estes resultados permitirão o planejamento para definição de áreas de manejo específico para fins de conservação dos recursos naturais. Sendo possível atingir este objetivo utilizando sensores, como o espectro, que permitem a quantificação de um grande número amostral, de forma rápida, prática, barata e menos poluente.
The irrational use of land has caused imbalances throughout the ecosystem. The influence on other components of the environment, such as water, air and the biogeochemical cycles of the elements, is already notorious. The present work aimed to estimate the C and P stocks in soils of the Planalto Ocidental Paulista, from the mineralogy of the clay fraction given by diffuse reflectance spectroscopy, and to establish and map specific management areas. The study area was the Planalto Ocidental Paulista, which occupies about 13 million ha. 600 samples were collected to perform physical, mineralogical and spectral analyses. The mineralogy, chemistry, particle size and physics of soils, C and P stocks, and C isotopes were evaluated in stratified samples, based on the geology, geomorphology and pedology of the area. All data were subjected to statistical and geostatistical analyses. The results obtained allowed to obtain the reference values of C and P stock, based on the mineralogical composition of the soil. Basalt soils have the highest C contents and element stocks when compared to sandstone soils. Sandstone soils in highly dissected landscape forms had the highest P stocks, followed by basalt soils. It is possible to generate C and P stock maps, and C content maps from different mathematical analyses. Machine learning allows good estimation of C contents for different geologies. These results will allow planning for the definition of specific management areas for the purpose of conserving natural resources. It is possible to achieve this goal using sensors such as the spectrum, which allow the quantification of a large sample number, in a fast, practical, cheap and less polluting way.
The irrational use of land has caused imbalances throughout the ecosystem. The influence on other components of the environment, such as water, air and the biogeochemical cycles of the elements, is already notorious. The present work aimed to estimate the C and P stocks in soils of the Planalto Ocidental Paulista, from the mineralogy of the clay fraction given by diffuse reflectance spectroscopy, and to establish and map specific management areas. The study area was the Planalto Ocidental Paulista, which occupies about 13 million ha. 600 samples were collected to perform physical, mineralogical and spectral analyses. The mineralogy, chemistry, particle size and physics of soils, C and P stocks, and C isotopes were evaluated in stratified samples, based on the geology, geomorphology and pedology of the area. All data were subjected to statistical and geostatistical analyses. The results obtained allowed to obtain the reference values of C and P stock, based on the mineralogical composition of the soil. Basalt soils have the highest C contents and element stocks when compared to sandstone soils. Sandstone soils in highly dissected landscape forms had the highest P stocks, followed by basalt soils. It is possible to generate C and P stock maps, and C content maps from different mathematical analyses. Machine learning allows good estimation of C contents for different geologies. These results will allow planning for the definition of specific management areas for the purpose of conserving natural resources. It is possible to achieve this goal using sensors such as the spectrum, which allow the quantification of a large sample number, in a fast, practical, cheap and less polluting way.
Descrição
Palavras-chave
Ciência do solo, Mineralogia de argila, Ciclos biogeoquímicos, Segurança alimentar