A carregar...
Publicação
Supporting environmental management and climate change governance with satellite chlorophyll-a estimation from space
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 6.94 MB | Adobe PDF |
Orientador(es)
Resumo(s)
O fitoplâncton desempenha um papel fundamental na regulação do clima global e na manutenção da vida nos ecossistemas aquáticos. Estes organismos microscópicos, que realizam fotossíntese, são responsáveis por cerca de metade da produção de oxigénio do planeta e pela absorção significativa de dióxido de carbono atmosférico. No entanto, as alterações climáticas têm vindo a modificar profundamente as condições dos ecossistemas aquáticos, afetando a composição, distribuição e fenologia das comunidades fitoplanctónicas. Este trabalho de investigação centra-se na monitorização do fitoplâncton como indicador de variabilidade climática, com especial enfoque na concentração de Clorofila-a (Chl-a), um pigmento fotossintético amplamente utilizado como proxy da biomassa fitoplanctónica. A tese propõe o reforço da monitorização do fitoplâncton em regiões particularmente sensíveis às alterações climáticas, como os lagos boreais, onde o degelo sazonal e outras variações geofísicas têm impacto direto na produtividade primária. Através da análise de séries temporais de dados de satélite e de amostragens in situ, foi possível modelar padrões de floração e identificar alterações na fenologia do fitoplâncton ao longo de mais de três décadas. A tese demonstra que é possível utilizar imagens de satélites de alta resolução espacial que não foram concebidas inicialmente com o propósito de detetar Clorofila-a. O caso apresentado refere-se à utilização dos satélites da família Landsat (Landsat 5 e 7), com resolução espacial de 30 metros, onde foi possível estimar padrões sazonais de clorofila em pequenos lagos – porém, com limitações bem definidas devidas, sobretudo, à baixa resolução espectral destes satélites. O primeiro artigo da tese é então um estudo inovador que permitiu superar limitações associadas à baixa resolução de sensores tradicionais de cor do oceano, como o MODIS ou o MERIS, especialmente em corpos de água de pequena dimensão. Os resultados demonstram que é possível estimar com precisão a concentração média de Chl-a em lagos de diferentes estados tróficos, com coeficientes de determinação (R²) superiores a 0.5 em lagos eutróficos. A metodologia adotada envolveu a calibração de modelos empíricos e semi-empíricos de regressão linear multivariada, utilizando bandas espectrais visíveis e infravermelhas dos sensores Landsat, bem como índices espectrais conhecidos como o NDCI (Normalized Difference Chlorophyll Index) e o BRG (Blue-Red-Green index). Estes modelos foram validados com dados de campo recolhidos em lagos da Finlândia, abrangendo diferentes estados tróficos, condições óticas e geofísicas. Os resultados demonstraram que os modelos baseados em dados Landsat conseguem estimar a concentração média de Chl-a com elevada precisão, especialmente em lagos eutróficos, onde a biomassa fitoplanctónica é mais elevada e, consequentemente, o sinal espectral é mais pronunciado.
Phytoplankton play a fundamental role in regulating the global climate and sustaining life in aquatic ecosystems. These microscopic, photosynthetic organisms are responsible for approximately half of the planet's oxygen production and for the significant absorption of atmospheric carbon dioxide. However, climate change has profoundly altered the conditions of aquatic ecosystems, affecting the composition, distribution, and phenology of phytoplankton communities. This research focuses on monitoring phytoplankton as an indicator of climate variability, with a special emphasis on the concentration of Chlorophyll-a (Chl-a), a photosynthetic pigment widely used as a proxy for phytoplankton biomass. The thesis proposes enhancing phytoplankton monitoring in regions particularly sensitive to climate change, such as boreal lakes, where seasonal thawing and other geophysical variations have a direct impact on primary productivity. Through the analysis of time series from satellite data and in situ samples, it was possible to model bloom patterns and identify changes in phytoplankton phenology over more than three decades. The thesis demonstrates that it's possible to use high-spatial-resolution satellite images that were not initially designed to detect Chlorophyll-a. The presented case refers to the use of satellites from the Landsat family (Landsat 5 and 7), with a spatial resolution of 30 meters, which allowed for the estimation of seasonal chlorophyll patterns in small lakes. However, this method has well-defined limitations due mainly to the low spectral resolution of these satellites. The first article in the thesis is an innovative study that overcame the limitations associated with the low resolution of traditional ocean color sensors, such as MODIS or MERIS, especially in small bodies of water. The results show that it is possible to accurately estimate the mean Chl-a concentration in lakes of different trophic states, with coefficients of determination (R²) greater than 0.5 in eutrophic lakes. The adopted methodology involved the calibration of empirical and semi-empirical multivariate linear regression models using visible and infrared spectral bands from Landsat sensors, as well as spectral indices known as the NDCI (Normalized Difference Chlorophyll Index) and the BRG (Blue-Red-Green index). These models were validated with field data collected from lakes in Finland, covering diferente trophic states, optical conditions, and geophysical characteristics. The results demonstrated that Landsat-based models can estimate the mean Chl-a concentration with high precision, especially in eutrophic lakes, where phytoplankton biomass is higher and, consequently, the spectral signal is more pronounced.
Phytoplankton play a fundamental role in regulating the global climate and sustaining life in aquatic ecosystems. These microscopic, photosynthetic organisms are responsible for approximately half of the planet's oxygen production and for the significant absorption of atmospheric carbon dioxide. However, climate change has profoundly altered the conditions of aquatic ecosystems, affecting the composition, distribution, and phenology of phytoplankton communities. This research focuses on monitoring phytoplankton as an indicator of climate variability, with a special emphasis on the concentration of Chlorophyll-a (Chl-a), a photosynthetic pigment widely used as a proxy for phytoplankton biomass. The thesis proposes enhancing phytoplankton monitoring in regions particularly sensitive to climate change, such as boreal lakes, where seasonal thawing and other geophysical variations have a direct impact on primary productivity. Through the analysis of time series from satellite data and in situ samples, it was possible to model bloom patterns and identify changes in phytoplankton phenology over more than three decades. The thesis demonstrates that it's possible to use high-spatial-resolution satellite images that were not initially designed to detect Chlorophyll-a. The presented case refers to the use of satellites from the Landsat family (Landsat 5 and 7), with a spatial resolution of 30 meters, which allowed for the estimation of seasonal chlorophyll patterns in small lakes. However, this method has well-defined limitations due mainly to the low spectral resolution of these satellites. The first article in the thesis is an innovative study that overcame the limitations associated with the low resolution of traditional ocean color sensors, such as MODIS or MERIS, especially in small bodies of water. The results show that it is possible to accurately estimate the mean Chl-a concentration in lakes of different trophic states, with coefficients of determination (R²) greater than 0.5 in eutrophic lakes. The adopted methodology involved the calibration of empirical and semi-empirical multivariate linear regression models using visible and infrared spectral bands from Landsat sensors, as well as spectral indices known as the NDCI (Normalized Difference Chlorophyll Index) and the BRG (Blue-Red-Green index). These models were validated with field data collected from lakes in Finland, covering diferente trophic states, optical conditions, and geophysical characteristics. The results demonstrated that Landsat-based models can estimate the mean Chl-a concentration with high precision, especially in eutrophic lakes, where phytoplankton biomass is higher and, consequently, the spectral signal is more pronounced.
Descrição
Tese de doutoramento em Alterações Climáticas e Políticas de Desenvolvimento Sustentável (Ciências do Ambiente), Universidade de Lisboa, Faculdade de Ciências, 2026.
Palavras-chave
Chlorophyll Climate Change Phenology Environmental Monitoring Remote Sensing Clorofila Alterações Climáticas Fenologia Monitorização Ambiental Deteção Remota