Juliana de Abreu Araújo

Formação complementar

Idiomas

Participação em eventos

Produções bibliográficas

• ARAÚJO, JULIANA DE ABREU ; GALVÃO, LÊNIO SOARES ; DALAGNOL, RICARDO . Evaluating changes with vegetation cover in PRISMA's spectral, spatial, and temporal attributes and their performance for classifying savannahs in Brazil. Remote Sensing Applications-Society And Environment , v. 32, p. 101074, 2023.

• DE ABREU ARAÚJO, JULIANA ; GALVÃO, LÊNIO SOARES ; DALAGNOL, RICARDO . Sensitivity of hyperspectral vegetation indices to rainfall seasonality in the Brazilian savannahs: an analysis using PRISMA data. Remote Sensing Letters , v. 14, p. 277-287, 2023.

• ARAUJO, J. A. ; FREIRE, A. ; DALAGNOL, RICARDO ; GALVÃO, LÊNIO SOARES . Mapping irrigated rice using MSI/Sentinel-2 time series of vegetation indices and Random Forest. In: Simpósio Brasileiro de Geoinformática - GEOINFO, 2021, São José dos Campos. Proceedings XXII GEOINFO, 2021. p. 37-45.

• ARAUJO, J. A. ; MEDEIROS, L. C. C. . Armadilhas de ovos: Será esse método eficiente para o controle do mosquito da Dengue?. In: XXXVI Congreso Interamericano de Ingeniería Sanitaria y Ambiental, 2018, Guayaquil. Anais do XXXVI Congreso Interamericano de Ingeniería Sanitaria y Ambiental, 2018.

• ARAUJO, J. A. ; MEDEIROS, L. C. C. ; PORTELA, C. I. . Modelo baseado em agentes para simular e analisar a eficiência de diferentes estratégias de controle do mosquito Aedes aegypti via inundação por ovitrampas.. In: X Congreso Latinoamericano de Biomatemática, 2017, Cusco. X Congreso Latinoamericano de Biomatemática - Libro de Resumenes, 2017., 2017.

• PORTELA, C. I. ; ARAUJO, J. A. ; MEDEIROS, L. C. C. . Modelo baseado em agentes para simular e analisar a eficiência da estratégia de controle da dengue via a introdução da bactéria Wolbachia nos mosquitos Aedes aegypti.. In: X Congreso Latinoamericano de Biomatemática, 2017, Cusco. X Congreso Latinoamericano de Biomatemática - Libro de Resumenes, 2017., 2017.

• ULIANA, E. M. ; ARAUJO, J. A. ; SOUSA JUNIOR, M. F. ; AIRES, U. R. V. ; RAMOS FILHO, H. A. . Global Radiation Estimation Using Reanalysis Data and Artificial Intelligence Technique. MODELING EARTH SYSTEMS AND ENVIRONMENT , 2024.

• ULIANA, E. M. ; SOUSA JUNIOR, M. F. ; ARAUJO, J. A. ; AIRES, U. R. V. ; SILVA, D. D. ; ZANUZO, M. R. ; CRUZ, I. F. . Validation and Spatio-Temporal Analysis of Precipitation Data Obtained by CHIRPS Remote Sensing for the State of Mato Grosso, Brazil. Revista Brasileira de Climatologia , 2024.

• ARAUJO, J. A. . Innovative Fintech Solutions for Sustainable Development. 2024. (Apresentação de Trabalho/Outra).

Projetos de pesquisa

• 2021 - 2023

  ADVANCING SAVANNAH CLASSIFICATION IN BRAZIL THROUGH PRISMA HYPERSPECTRAL DATA ANALYSIS, Descrição: Using a multi-temporal sequence of PRISMA images obtained between May and October of 2020, over the Brasília National Park (BNP), the two major objectives of the present study were: (a) to investigate the sensitivity of hyperspectral vegetation indices (VIs) with rainfall seasonality in the savannahs; (b) to evaluate the changes in spectral, spatial and temporal attributes with savannah vegetation cover, and their performance to classify Cerrados physiognomies. Spectral metrics comprised 166 reflectance bands of PRISMA; 22 narrowband vegetation indices, associated with vegetation structure, biochemistry and physiology; 24 absorptions bands parameters (the depth, area, width, and asymmetry) of the major absorption bands centered at 680 nm (chlorophyll), 980 nm and 1200 nm (leaf water), and 1700 nm, 2100 nm, and 2300 nm (lignin-cellulose) calculated from the continuum removal method; and four fractions images derived from Spectral Mixture Analysis (SMA). The spatial attributes included eight Grey Level Co-occurrence Matrix (GLCM) metrics, obtained in a 3x3 filter window applied over the near infrared (NIR) reflectance band (864 nm). The temporal metrics came from browning and greening rates of vegetation change. They were calculated using data from May 11th and August 17th (browning), and from September 4th and October 3rd (greening). To reduce dimensionality of the hyperspectral data and remove redundant information, feature selection was performed using the Correlation based Feature Selection (CFS) technique.Then, the classification of seven savannah physiognomies was carried out using the Random Forest (RF) model, considering each set of individual metrics and their combination. To validate the classification, an available reference map was used, allowing the calculation,through Monte Carlo simulation, of the metrics of Overall Accuracy, F1-score, Recall, and Precision. The results showed that the hyperspectral VIs were sensitive to seasonal variations in rainfall. The VI sensitivity to increasing water deficit from May to July and after the first occurrences of rainfall in October was dependent on the vegetation index and on the savannah physiognomy under analysis. NDVI, EVI, MSI, SIPI and NDWI were the most sensitive indices to the increase in water deficit over grassland areas, while RVSI and RENDVI were the most sensitive indices to the beginning of the new seasonal cycle in October. When evaluating the RF classification using spectral, spatial, and temporal attributes based on PRISMA, the combined use of metrics showed the best overall accuracy (74 after feature selection). The highest F1-scores were observed for the woodland savannah, savannah grassland, and riparian forest classes, ranging from 0.72 to 0.91. The lowest F1-scores (between 0.20 and 0.28) were observed for the wooded savannah classes with or without Trembleias. Results indicated the importance of considering different hyperspectral metrics and multitemporal data for classifying savannahs in Brazil.. , Situação: Concluído; Natureza: Pesquisa. , Integrantes: Juliana de Abreu Araujo - Integrante / GALVÃO, LÊNIO SOARES - Coordenador / DALAGNOL, RICARDO - Integrante.

• 2015 - 2016

  Modelo Baseado em Agentes para simular e analisar a eficiência de diferentes estratégias de controle do mosquito Aedes aegypti via inundação por ovitrampas., Descrição: O estudo teve como objetivo desenvolver um Modelo Baseado em Agente para simular e analisar a técnica de monitoramento e controle do mosquito Aedes aegypti, inundação por ovitrampas. Investigando a quantidade, distribuição e frequência das armadilhas de ovos no ambiente na eficiência do controle vetorial.. , Situação: Concluído; Natureza: Pesquisa. , Alunos envolvidos: Graduação: (1) . , Integrantes: Juliana de Abreu Araujo - Coordenador / Liliam César de Castro Medeiros - Integrante.

Prêmios