Chlorophyll a spatial inference using artificial neural network from multispectral images and in situ measurements

Considering the importance of monitoring the water quality parameters, remote sensing is a practicable alternative to limnological variables detection, which interacts with electromagnetic radiation, called optically active components (OAC). Among these, the phytoplankton pigment chlorophyll a is the most representative pigment of photosynthetic activity in all classes of algae. In this sense, this work aims to develop a method of spatial inference of chlorophyll a concentration using Artificial Neural Networks (ANN). To achieve this purpose, a multispectral image and fluorometric measurements were used as input data. The multispectral image was processed and the net training and validation dataset were carefully chosen. From this, the neural net architecture and its parameters were defined to model the variable of interest. In the end of training phase, the trained network was applied to the image and a qualitative analysis was done. Thus, it was noticed that the integration of fluorometric and multispectral data provided good results in the chlorophyll a inference, when combined in a structure of artificial neural networks.

Inferência espacial para mapeamento de macrófitas submersas: estudo de caso

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Métodos para extração de informações a partir de imagens multiespectrais de escalas grandes

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Desenvolvimento de método para inferência de características físicas da água associadas às variações espectrais. Caso de Estudo: Reservatório de Itupararanga/SP

Pós-graduação em Ciências Cartográficas - FCT

Geostatistics and remote sensing methods in the classification of images of areas cultivated with citrus

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Estimation of chlorophyll-a concentration and the trophic state of the Barra Bonita hydroelectric reservoir using oli/landsat-8 images

Reservoirs are artificial environments built by humans, and the impacts of these environments are not completely known. Retention time and high nutrient availability in the water increases the eutrophic level. Eutrophication is directly correlated to primary productivity by phytoplankton. These organisms have an important role in the environment. However, high concentrations of determined species can lead to public health problems. Species of cyanobacteria produce toxins that in determined concentrations can cause serious diseases in the liver and nervous system, which could lead to death. Phytoplankton has photoactive pigments that can be used to identify these toxins. Thus, remote sensing data is a viable alternative for mapping these pigments, and consequently, the trophic. Chlorophyll-a (Chl-a) is present in all phytoplankton species. Therefore, the aim of this work was to evaluate the performance of images of the sensor Operational Land Imager (OLI) onboard the Landsat-8 satellite in determining Chl-a concentrations and estimating the trophic level in a tropical reservoir. Empirical models were fitted using data from two field surveys conducted in May and October 2014 (Austral Autumn and Austral Spring, respectively). Models were applied in a temporal series of OLI images from May 2013 to October 2014. The estimated Chl-a concentration was used to classify the trophic level from a trophic state index that adopted the concentration of this pigment-like parameter. The models of Chl-a concentration showed reasonable results, but their performance was likely impaired by the atmospheric correction. Consequently, the trophic level classification also did not obtain better results.

Imagens aéreas multiespectrais na identificação de zonas de manejo em áreas de algodão para aplicação localizada de insumos

Crop yield is influenced by several factors with variability in time and space that are associated with the variations in the plant vigor. This variability allows the identification of management zones and site-specific applications to manage different regions of the field. The purpose of this study was the use of multispectral image for management zones identification and implications of site-specific application in commercial cotton areas. Multispectral airborne images from three years were used to classify a field into three vegetation classes via the Normalized Difference Vegetation Index (NDVI). The NDVI classes were used to verify the potential differences between plant physical measurements and identify management zones. The cotton plant measurements sampled in 8 repetitions of 10 plants at each NDVI class were Stand Count, Plant Height, Total Nodes and Total Bolls. Statistical analysis was performed with treatments arranged in split plot design with Tukey’s Test at 5% of probability. The images were classified into five NDVI classes to evaluate the relationship between cotton plant measurement results and sampling location across the field. The results have demonstrated the possibility of using multispectral image for management zones identification in cotton areas. The image classification into three NDVI classes showed three different zones in the field with similar characteristics for the studied years. Statistical differences were shown for plant height, total nodes and total bolls between low and high NDVI classes for all years. High NDVI classes contained plants with greater height, total nodes and total bolls compared to low NDVI classes. There was no difference in Stand Count between low and high NDVI classes for the three studied years. The final plant stand was the same between all NDVI classes for 2001 and 2003 as it was expected due to the conventional seeding application with the same rate of seeds for the entire field.