| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.53 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Este trabalho de projeto tem como objetivo a identificar e diagnosticar situações de stress no sobreiro com recurso a tecnologias de deteção remota, sendo realizado em duas áreas da Companhia das Lezírias e da Herdade da Machoqueira do Grou. A monitorização é feita a partir da avaliação da evolução temporal de diversos índices vegetativos calculados com imagens multiespectrais de alta resolução adquiridas pelo satélite Sentinel-2, obtendo assim a variabilidade temporal da assinatura espectral de sobreiros saudáveis e sobreiros identificados como mortos ou em perda de vitalidade. Para este estudo foram adquiridas imagens dos satélites Sentinel-2A e Sentinel-2B compreendidas entre 2016 e 2018, cuja área contém as duas áreas em estudo. A informação obtida pelas imagens permitiu, numa primeira fase, identificar a diferente assinatura espectral de cada uma das classes em época de chuvas e em época seca. Posteriormente foram calculados sete índices vegetativos (NDVI, CIRedEdge1, CIRedEdge2, CIGreen, GNDVI, EVI e SAVI) para cada uma das imagens, com o objetivo de reconhecer qual o índice que melhor se adequava para o estudo, e elaboradas séries temporais de 25 e 47 imagens, confirmando o resultado obtido primeiramente com as assinaturas espectrais. A análise temporal dos índices vegetativos permitiu observar a relação entre o estado vegetativo do sobreiro e a presença de precipitação, tal como a existência de vegetação vigorosa no solo em épocas de maior precipitação. Esta abordagem permite que sejam definidos intervalos de valores dos índices para os sobreiros saudáveis e para os sobreiros identificados como mortos ou em perda de vitalidade, a partir da média e desvio padrão obtidos para cada uma das classes. A metodologia proposta apresentou resultados satisfatórios, sendo que estes estão sujeitos a melhorias significativas nomeadamente na utilização de imagens de satélite com uma melhor resolução, que permitirão no futuro identificar remotamente sobreiros que possam estar em situação de stress, de forma a que seja diminuída a necessidade de deslocações ao terreno e os custos associados.
This project work aims to identify and diagnose stress situations in the cork oak using remote sensing technologies, being carried out in two areas of the Companhia das Lezírias and the Herdade da Machoqueira do Grou. The monitoring is done by evaluating the temporal evolution of several vegetative indexes calculated with high resolution multispectral images acquired by the Sentinel-2 satellite, thus obtaining the temporal variability of the spectral signature of healthy cork oaks and cork oaks identified as dead or in loss of vitality. For this study, images of the Sentinel-2A and Sentinel-2B satellites were acquired between 2016 and 2018, whose area contains the two areas under study. The information obtained by the images allowed, in a first phase, to identify the different spectral signature of each one of the classes in rainy season and in the dry season. Subsequently, seven vegetative indexes (NDVI, CIRedEdge1, CIRedEdge2, CIGreen, GNDVI, EVI and SAVI) were calculated for each of the images, with the objective of recognizing which index was best suited for the study, and elaborated time series of 25 and 47 images, confirming the result obtained first with the spectral signatures. The temporal analysis of vegetative indexes allowed to observe the relationship between cork oak vitality and the presence of precipitation, as well as the existence of vigorous vegetation in the soil in times of greater precipitation. This approach allows ranges of index values to be defined for healthy cork oaks and cork oaks identified as dead or loss of vitality, from the mean and standard deviation obtained for each class. The proposed methodology presented satisfactory results, and these are subject to significant improvements, especially in the use of satellite images with better resolution, which will allow in the future to remotely identify corns that may be in a stress situation, in order to reduce the need of travel to the land and associated costs.
This project work aims to identify and diagnose stress situations in the cork oak using remote sensing technologies, being carried out in two areas of the Companhia das Lezírias and the Herdade da Machoqueira do Grou. The monitoring is done by evaluating the temporal evolution of several vegetative indexes calculated with high resolution multispectral images acquired by the Sentinel-2 satellite, thus obtaining the temporal variability of the spectral signature of healthy cork oaks and cork oaks identified as dead or in loss of vitality. For this study, images of the Sentinel-2A and Sentinel-2B satellites were acquired between 2016 and 2018, whose area contains the two areas under study. The information obtained by the images allowed, in a first phase, to identify the different spectral signature of each one of the classes in rainy season and in the dry season. Subsequently, seven vegetative indexes (NDVI, CIRedEdge1, CIRedEdge2, CIGreen, GNDVI, EVI and SAVI) were calculated for each of the images, with the objective of recognizing which index was best suited for the study, and elaborated time series of 25 and 47 images, confirming the result obtained first with the spectral signatures. The temporal analysis of vegetative indexes allowed to observe the relationship between cork oak vitality and the presence of precipitation, as well as the existence of vigorous vegetation in the soil in times of greater precipitation. This approach allows ranges of index values to be defined for healthy cork oaks and cork oaks identified as dead or loss of vitality, from the mean and standard deviation obtained for each class. The proposed methodology presented satisfactory results, and these are subject to significant improvements, especially in the use of satellite images with better resolution, which will allow in the future to remotely identify corns that may be in a stress situation, in order to reduce the need of travel to the land and associated costs.
Descrição
Trabalho de rojecto de mestrado, Sistemas de Informação Geográfica - Tecnologias e Aplicações, Universidade de Lisboa, Faculdade de Ciências, 2019
Palavras-chave
Deteção remota Sentinel-2 Análise multitemporal Índices de Vegetação Montado de Sobreiro Trabalhos de projecto de mestrado - 2019
