Relação entre índices de vegetação e atributos biofísicos do solo e da planta em vinhedo irrigado
Carregando...
Data
2022-02-22
Autores
Pereira, Anderson de Jesus
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O conhecimento da distribuição espacial do estado hídrico é um caminho para entender os fatores que interferem na produtividade da videira, auxiliar na redução da aplicação de água, possibilitar a identificação de zonas homogêneas e na melhoria da qualidade da uva. Nessa perspectiva, o objetivo deste estudo é identificar a variabilidade espacial do estado hídrico em videiras irrigadas por gotejamento no município de Ribeirão Preto-SP, mediante a utilização de dados multiespectrais e hiperespectrais obtidos por sensores suborbital e proximal, para observar a correlação entre os índices de vegetação (IV) e os atributos da planta e do solo coletados in loco. As imagens multiespectrais foram capturadas através do sensor MicaSense RedEdge-M e adquiriram os dados proximais por meio do sensor Crop Circle ACS-470 e do espectrorradiômetro FieldSpec 3. Os índices de vegetação foram elaborados e comparados com os dados coletados em campo, a exemplo do teor relativo de água (TRA), condutância estomática (gs) condutividade elétrica aparente do solo (CEa), conteúdo total de clorofila e umidade volumétrica do solo (θ). As correlações dos índices de vegetação (IV) suborbitais com a condutância estomática, o teor relativo de água e o conteúdo de clorofila total foram maiores do que as correlações com os índices multiespectrais proximais. O índice da diferença normalizada do vermelho limítrofe (NDRE) suborbital alcançou um R2 de 70 % para a (gs) e 57 % para o TRA, esses valores indica o melhor desempenho do NDRE. Assim, as imagens multiespectrais possibilitaram a construção de um mapa da variabilidade espacial do estado da água em um vinhedo irrigado. As assinaturas espectrais possibilitaram a formulação de 26 índices de vegetação com bandas de faixa estreita. Dentre os quais o índice de umidade (MSI), índice de água por diferença normalizada (NDWI) e o índice de água modificado (WI1) tiveram as melhores correlações com a (gs) e TRA. O índice de conteúdo de clorofila da copa (CCCI) e o índice de absorção de clorofila transformada em refletância (TCARI), ao serem relacionados com a clorofila total, alcançaram um R2 de 77 % e de 73 %, respectivamente. A integração de planos de informações possibilitou a definição de uma zona de manejo (ZM), em que comprovou a confiabilidade da metodologia empregada, pois houve diferenças significativas entre os atributos adotados para validar a delimitação das zonas de manejo. Assim, há potencial para mapear a variabilidade do estado hídrico pelos sensores multiespectral suborbital e proximal, entretanto o processamento digital das imagens aéreas produziram informações mais consistentes em relação aos dados proximais.
The knowledge of the spatial distribution of the water status is a way to understand the factors that interfere in the productivity of the vine, to help in the reduction of water application, to enable the identification of homogeneous zones and to improve the quality of the grape. In this perspective, the goal of this study is to identify the spatial variability of the water status in drip irrigated vines in Ribeirão Preto, SP, using multispectral and hyperspectral data obtained by suborbital and proximal sensors, to observe the correlation between the indices vegetation (IV) and plant and soil attributes collected in loco. The multispectral images were captured using the MicaSense RedEdge-M sensor and proximal data were acquired using the Crop Circle ACS-470 sensor and the FieldSpec 3 spectroradiometer. The vegetation indices were prepared and compared with data collected in the field, the example of the relative water content (TRA), stomatal conductance (gs), soil apparent electrical conductivity (ECa), total chlorophyll content and soil volumetric moisture (θ). Correlations of suborbital vegetation indices (IV) with stomatal conductance, relative water content and total chlorophyll content were greater than correlations with proximal multispectral indices. The suborbital red border normalized difference index (NDRE) reached a R2 of 70% for a (gs) and 57% for the TRA, these values indicate the best performance of the NDRE. Thus, the multispectral images enabled the construction of a map of the spatial variability of the water state in an irrigated vineyard. The spectral signatures allowed the formulation of 26 vegetation indices with narrow bands. Among which the moisture index (MSI), normalized difference water index (NDWI) and modified water index (WI1) had the best correlations with a (gs) and TRA. The crown chlorophyll content index (CCCI) and the transformed chlorophyll absorption into reflectance index (TCARI), when related to the total chlorophyll, reached an R2 of 77 % and 73 %, respectively. The integration of information plans made it possible to define a management zone (MZ), which proved the reliability of the methodology used, as there were significant differences between the attributes adopted to validate the delimitation of management zones. Thus, there is potential to map the variability of the water status by the suborbital and proximal multispectral sensors, however the digital processing of the aerial images produced more consistent information in relation to the proximal data.
The knowledge of the spatial distribution of the water status is a way to understand the factors that interfere in the productivity of the vine, to help in the reduction of water application, to enable the identification of homogeneous zones and to improve the quality of the grape. In this perspective, the goal of this study is to identify the spatial variability of the water status in drip irrigated vines in Ribeirão Preto, SP, using multispectral and hyperspectral data obtained by suborbital and proximal sensors, to observe the correlation between the indices vegetation (IV) and plant and soil attributes collected in loco. The multispectral images were captured using the MicaSense RedEdge-M sensor and proximal data were acquired using the Crop Circle ACS-470 sensor and the FieldSpec 3 spectroradiometer. The vegetation indices were prepared and compared with data collected in the field, the example of the relative water content (TRA), stomatal conductance (gs), soil apparent electrical conductivity (ECa), total chlorophyll content and soil volumetric moisture (θ). Correlations of suborbital vegetation indices (IV) with stomatal conductance, relative water content and total chlorophyll content were greater than correlations with proximal multispectral indices. The suborbital red border normalized difference index (NDRE) reached a R2 of 70% for a (gs) and 57% for the TRA, these values indicate the best performance of the NDRE. Thus, the multispectral images enabled the construction of a map of the spatial variability of the water state in an irrigated vineyard. The spectral signatures allowed the formulation of 26 vegetation indices with narrow bands. Among which the moisture index (MSI), normalized difference water index (NDWI) and modified water index (WI1) had the best correlations with a (gs) and TRA. The crown chlorophyll content index (CCCI) and the transformed chlorophyll absorption into reflectance index (TCARI), when related to the total chlorophyll, reached an R2 of 77 % and 73 %, respectively. The integration of information plans made it possible to define a management zone (MZ), which proved the reliability of the methodology used, as there were significant differences between the attributes adopted to validate the delimitation of management zones. Thus, there is potential to map the variability of the water status by the suborbital and proximal multispectral sensors, however the digital processing of the aerial images produced more consistent information in relation to the proximal data.
Descrição
Palavras-chave
variabilidade hídrica, viticultura de precisão, sensoriamento proximal, sensoriamento remoto, zona de manejo, water variability, precision viticulture, proximal sensing, remote sensing, management zone