Detalhes bibliográficos
| Ano de defesa: |
2020 |
| Autor(a) principal: |
SILVA, Diego Castro da
 |
| Orientador(a): |
LOPES, Pabrício Marcos Oliveira |
| Banca de defesa: |
NASCIMENTO, Cristina Rodrigues,
BRITO, José Ivaldo Barbosa de |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal Rural de Pernambuco
|
| Programa de Pós-Graduação: |
Programa de Pós-Graduação em Engenharia Agrícola
|
| Departamento: |
Departamento de Engenharia Agrícola
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
http://www.tede2.ufrpe.br:8080/tede2/handle/tede2/9080
|
Resumo: |
High concentrations of salts in the soil are one of the serious environmental problems that degrade the environment, make agricultural activities unfeasible and can lead to desertification. It occurs more frequently in irrigated perimeters located in arid and semi-arid regions, such as the Maniçoba irrigation project, the main irrigated fruit production project in the municipality of Juazeiro-BA. Monitoring salinized areas and assessing their impacts on land use and occupation, over time on a large scale, using Remote Sensing, can be an effective approach to support decision making in prevention and control activities of this phenomenon. The objective of this study was to analyze the spatial-temporal dynamics of salinity in the irrigated perimeter of Maniçoba using Landsat-8 and Sentinel-2 images, applying salinity and vegetation indices in conjunction with meteorological data. The study was carried out in agricultural areas with signs of salinity, where samples were collected for analysis of the electrical conductivity of the soil (CE). Images from the Landsat-8 and Sentinel-2 satellites were used in meteorological data from 2014 to 2019. Using the QGis 2.18.19 software, the images were pre-processed, atmospheric influences were corrected, converting the digital numbers into surface reflectance and calculating spectral bands to obtain the biophysical parameters: vegetation indices NDVI, SAVI, EVI and GDVI, and salinity indices SI-1, SI-2, SI-3 and IB, albedo, surface temperature and actual evapotranspiration. The interpolation techniques, digital classification of Maxver images were performed, their accuracy was assessed and the pixel values of 4 soil classes were extracted to cross-check information from the calculated variables. Multivariate statistics of principal component analysis (PCA), Pearson's correlation and descriptive statistics were used to assess the relationships between parameters and to quantify their behavior over time as a function of soil salinity. The meteorological information characterized the climatic conditions for the study period. The SI-1 and SI-3 salinity indices and GDVI and SAVI vegetation indices showed the best statistical responses. The (ACP) reduced the size of the data set and separated groups of variables of greater similarity, obtaining in the accumulated of CP2 values above 78% for the three areas. The EC demonstrated a strong relationship with the surface temperature, albedo and SI-1 and SI-3 indices, in addition to a strong indirect relationship with the GDVI and SAVI. The EC analyzes revealed that the areas are very degraded by salinity, mainly in exposed soils, followed by natural vegetation and agricultural area. The analysis of thematic maps generated from the GDVI, SAVI, SI-1 and SI-3 indexes, showed the changes that occurred in the use and occupation of the soil over time, due to the salinization of the soils, confirmed by the statistical analysis and graphs of spectral reflectance of the different classes. The joint application of Remote Sensing techniques proved to be effective in characterizing salinity at a spatial and temporal level, and meteorological data contributed to the understanding of the processes observed in the study. |
