Navegando por Assunto "Sensoriamento remoto"

Agora exibindo 1 - 20 de 30

Análise da distribuição espacial do índice de umidade do solo em regiões semiáridas a partir de dados de sensoriamento remoto
(2019-11-26) Santos, Jadiene Moura dos; Silva, Emanuel Araújo; Oliveira, Cinthia Pereira de; http://lattes.cnpq.br/8148643000907549; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/5414923091157764
Soil moisture represents a fraction of water that is at a surface level of the earth where there is interaction with the atmosphere through evapotranspiration. It is a fundamental variable in the functioning of several processes that act in the terrestrial system, besides characterizing the desertification of semiarid and arid regions. This course conclusion work aimed to evaluate the spatial distribution of the soil moisture index in a dry tropical forest area, in the city of Floresta/PE, through orbital images. The methodology was applied for four distinct dates (11/21/15, 11/23/16, 12/12/17 and 11/13/18) and data processing to obtain the Normalized Difference Vegetation Index (NDVI), Surface Temperature (Ts) and Soil Moisture Index (IUS) were performed using Qgis software. In addition, the time series of the annual precipitation of the municipality of Floresta / PE were classified into dry, normal or rainy years, using the quantile method and the monthly precipitation analysis in relation to the climatological normals from 2015 to 2018, using the data obtained from the APAC website. Finally, the point cloud distribution was distributed between IUSxNDVI and IUSxTs. The results showed that the classification from 1999 to 2018 eight years behaved as normal and six years as dry and rainy, where the values for dry years ranged from 149.50 to 349.20mm, the normal ones from 392.70 to 538.1mm and the rainy ones from 559.10 to 750.60mm. For the years 2015, 2016, 2017 and 2018, the accumulated monthly precipitation values were 223.00mm, 395.10mm, 399.20mm and 653.50mm, respectively. The NDVI values in exposed soil plus thin vegetation ranged from 0.124 to 0.323, in arboreal vegetation between 0.351 to 0.649 and in surrounding water bodies -0.072. At surface temperature, minimum values of 23.80 ° C and maximum values of 44.93 ° C were found. For the soil moisture index, 0.240 were found in exposed soil and thin vegetation, 0.417 to 0.746 in tree vegetation and 0.821 in water. In the distribution of the IUS point cloud with NDVI and Ts biophysical parameters, it was observed that 2015 and 2016 had no pixels in the negative NDVI region and Ts pixels were above 40 ° C. On the other hand, in 2017 and 2018 due to the presence of water there was a reduction in Ts, where most pixels were below 40 ° C. The Quantis method allowed to identify in an interval of twenty years an irregular pattern between years classified as dry, normal and rainy. The monthly precipitation of the four showed great variability in relation to the climatological normal of the municipality. The NDVI allowed to identify the presence of water bodies, exposed soil with herbaceous and arboreal vegetation in the farm Itapemirim / PE. The achievement of low surface temperature values on the imaged dates is associated with high NDVI and soil moisture values. IUS values were lower in exposed soil and more expressive in water and under tree vegetation, due to the rapid response of dry tropical forests during the rainy season during the imaging date. Point cloud distribution showed increasing behavior for IUSxNDVI and inverse for IUSxTs.
Análise das características dos índices de vegetação em um plantio de Eucalyptus spp. utilizando imagens do Sentinel-2A
(2023-09-14) Silva, Adailton Domingos Salustiano da; Silva, Emanuel Araújo; Sá, Vânia Aparecida de; http://lattes.cnpq.br/5807408661337266; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/7511858370212406
Análise de correlação entre a produção primária bruta do sensor MODIS e balanço hídrico do SWAT para bacia hidrográfica do Riacho do Pontal, Pernambuco
(2021-07-19) Brito, Pedro Vinícius da Silva; Silva, Antonio Samuel Alves da; http://lattes.cnpq.br/0249875496935177; http://lattes.cnpq.br/2946714997867399
The correlation between gross primary production and the water balance of the Riacho do Pontal watershed, located in the state of Pernambuco, Brazil, was evaluated. For this, the components of precipitation (PRECIP), water production (WYLD), soil water storage (SW), surface runoff (SURQ), potential evapotranspiration (PET) and actual evapotranspiration (ET) of the water balance were used simulated by the SWAT model and the MOD17A2H remote sensing product of Gross Primary Productivity (GPP) from the MODIS (MODerate Resolution Imaging Spectroradiometer) sensor. The data obtained by the MODIS sensor and the SWAT model were analyzed and then correlated using Spearman correlation coefficient. The results show that the minimum GPP with runoff had the lowest correlation value of the entire analysis (of 0.13), a negative correlation with the potential evapotranspiration (of -0.63) and positive with the other components of the balance water, ranging from 0.22 with precipitation to 0.38 with actual evapotranspiration. With the average GPP, there was a negative correlation for potential evapoternspiration (from -0.74), and with the other components of the water balance, there was a positive correlation, ranging from 0.36 with runoff to 0.6 with real evapotranspiration. At maximum GPP, there was again a negative correlation with potential evapotranspiration (from -0.76), and a positive correlation with all other components of the water balance, ranging from 0.28 with runoff to 0.46 with actual evapotranspiration and water production. The existence of a correlation between the gross primary production and the water balance was observed, even if it is low, as occurred between the minimum GPP and the runoff. The correlation between GPP and real and potential evapotranspiration was greater than the correlation between GPP and the other components of the water balance, with the exception of the maximum GPP with water production, which had the same value as the maximum GPP with the real evapotranspiration. Thus, monitoring evapotranspiration in semiarid regions is of great importance for predicting gross primary production. And according to the predictions of the Intergovernmental Panel on Climate Change (IPCC), which predicts the increase of extreme events in semiarid regions, it is possible to say that if climate change scenarios come to occur, there is a strong trend in production primary gross, in semiarid regions, decrease.
Análise de degradação ambiental na bacia do Rio Moxotó através de imagens de satélite
(2023-04-19) Aires, Giovanna da Cunha; Nascimento, Cristina Rodrigues; http://lattes.cnpq.br/9289129949520610; http://lattes.cnpq.br/1000446236815556
The disorderly exploitation of natural resources. the inappropriate use of soil and irrational deforestation has been causing numerous environmental problems, which leads to environmental degradation, with that remote sensing has been gaining space through environmental monitoring in space and time with efficiency and at low cost, allowing planning and decision-making for sustainable use of natural resources. In view of this, the objective was to analyze, through a time series of LANDSAT satellite images, spaced at intervals of five years, the environmental degradation suffered in the Moxot6 River Basin. The climate analysis was based on a monthly series of twenty-one years of data (2000-2021), using the Quantile technique. The satellite images analyzed were from MapBiomas and processed in QGIS software. The change detection analysis was also carried out through the difference image, resulting from techniques based on the observation of the pixels and the change that occurred with them over time. With this, it was observed that the rainy season in the Moxoto River Basin lasts four months, from January to April, representing about 72% of the annual rainfall and presenting an average of total annual precipitation of 482.65mm. Through the analysis of land use and cover, it was possible to verify the growth of agriculture in the basin, through the classes of temporary crops, with a growth of 97.13% between the years 2005 and 2021, and the classes of perennial crops and areas of pasture, with growth between 2000 and 2021 of 97.64% and 5.35% respectively. The savanna vegetation is the most extensive in the basin, however the areas of forest formation are a minority in the basin, thus having a low presence of dense vegetation. Regarding the degradation of the Moxot6 River Basin, there was a decrease of 98.74% from the year 2000 to 2019, with the municipalities most affected by the degradation being Ibimirim-PE, Inaji-PE, Custodia-PE, Math Grande-AL, Piraconha- AL and Delmiro Gouveia-AL. The regeneration classes showed an increase of 32.63% from 2000 to 2019. Through the difference image analysis, it was also possible to verify permanence of recovered areas, however the transition from degraded to recovered areas was lower than expected. With that, remote sensing through LANDSAT satellite images, enabled the classification and analysis of land use classes, from the years 2000, 2005, 2010, 2015, 2020 and 2021, as well as the analysis of degraded and recovered areas from the years 2000, 2005, 2010, 2015 and 2019 satisfactorily so that it is possible to carry out adequate monitoring of the transformations undergone by the Moxoto River basin. With the studies, the need for environmental recovery projects in the basin was verified, in addition to constant monitoring to assist in public policies for this region.
Análise temporal do uso e cobertura da terra do município de Macaparana - Pernambuco
(2020-11-03) Moura, Lucas Araujo; Duarte, Simone Mirtes Araújo; Moreira, Giselle Lemos; http://lattes.cnpq.br/6171199372079024; http://lattes.cnpq.br/5876968040869585; http://lattes.cnpq.br/2567696308015910
The geotechnologies linked to remote sensing are essential tools to understand the use and occupation of a territory, in addition to effectively and economically assisting in the monitoring of natural resources. Through data from different years, it is possible to create a study of the main factors of degradation of natural resources. Thus, the present work aimed at elaborating a temporal analysis of the municipality of Macaparana - PE for the years 2007 and 2018. All the geoprocessing activities for the supervised classification and vegetation index generation were computed through the software Qgis version 2.18.10 and 3.10.9, the supervised classification was performed through the Semi Automatic Classification Plugin (SCP), where several samples were selected in the bands compiling for the due years, and through the MaxVer, the classes of exposed soil/urban area, agricultural culture, forests and water resources were computed. For the accuracy of the data, the kappa index was performed. The kappa index for the years 2007 and 2018 was 0.49 and 0.79, showing that it is a good mapping. And through the maps generated and the quantified classes, where the exposed soil of the area increased by 48%, the vegetation had a decrease of 35%, the agricultural crop had its area reduced by 10% and the water resources increased about 303%. There was a big change in the results of water resources due to the amount of clouds in the image of 2018, which hindered the classification, but when going to the field it was possible to observe, lack of vegetation in the area around the water resources, which may occur a process of silting up the rivers. It was possible to establish a relationship between the exposed soil and the agricultural crop, the lack of effective management in search of greater productivity instead of cutting more areas to plant, justified with the data of the vegetation area, which decreased to give space to new agricultural crops. And relating the forest area obtained through the supervised classification, with the area computed from IVDN, showed homogeneity in the results, varying less than 5% for the two years. Therefore, it is possible to conclude that anthropic activities, without any management plan for planting and harvesting, are having direct effects on the reduction of vegetation in the area, requiring better control in their production and alternatives that do not require burning cane for cutting, where it has direct impacts on the soil.
Avaliação da contaminação por metais pesados em solos urbanos da Região Metropolitana do Recife: bioacessibilidade e correlação com NDVI
(2024-03-07) Mello, Lucas José Souza de; Biondi, Caroline Miranda; Lins, Simone Aparecida da Silva; http://lattes.cnpq.br/7329862411748916; http://lattes.cnpq.br/8326756664758702; http://lattes.cnpq.br/8741487779369891
Urban pollution by heavy metals is a subject of great socio-environmental relevance due to its potential deleterious effects on human and ecological health. Remote sensing, particularly the use of the Normalized Difference Vegetation Index (NDVI), emerges as a promising tool to assess vegetation health and potential impacts of soil contamination. In this regard, the present study aimed to evaluate the total contents and bioaccessibility of heavy metals in urban soil of the Recife Metropolitan Region and its chemical characteristics, obtain the NDVI of the sampled areas, and correlate it with the total metal contents analyzed. The research was conducted in the Recife Metropolitan Region (RMR), where samples of surface soil were collected in distinct urban areas and their points were georeferenced. The samples were analyzed for pH values, Organic Carbon, Soil Cation Exchange Capacity (CTC), and total heavy metal contents using Energy Dispersive X-ray Fluorescence (pXRF). Metal contents were compared with Quality Reference Values, and in vitro assays were performed to determine metal bioaccessibility in the soil. For NDVI estimation, images obtained from the CBERS-4A satellite with 8m spatial resolution were used, and buffers with radii of 55, 110, and 220m were applied to demarcate the area to be analyzed based on the sampled points. The results revealed high levels of heavy metals in urban areas of the RMR, exceeding the Quality Reference Values for the state of Pernambuco, yet the metals exhibit low bioaccessibility. Regarding the size of the analyzed area, there was no significant difference in relation to the values obtained. Additionally, a negative correlation was observed between most soil metal contents and NDVI, meaning that as NDVI increases, metal contents decrease.
Determinação da cobertura vegetal de Olinda-PE: um subsídio a gestão florestal urbana
(2019-12-03) Lopes, Iran Jorge Corrêa; Lima Neto, Everaldo Marques de; Pessoa, Mayara Maria de Lima; http://lattes.cnpq.br/4721886920195910; http://lattes.cnpq.br/6791561445213969; http://lattes.cnpq.br/3433274611248891
Urban sprawl without proper planning create many side effects for its inhabitants, both environmentally and socially. Increasingly, vegetation has been associated to the idea of quality of life, due to the proven benefit it promotes in the urban environment. The objective of this research was to classify the land use of the city of Olinda – PE, as well as to quantify its urban forest. For this, were used remote sensing techniques, with the purpose of generating spatial information that will serve as a basis for urban land use and occupation planning. And so, the urban infrastructure, water, exposed soil, vegetation and cloud classes were identified through the QuantumGIS software with the Maxver supervised classification, with 10 meters resolution SENTINEL – 2 satellite images dated from February of 2018. The urban forest indexes of the city were determined and measured. The county of Olinda presented land use of 6,32% of water, 12,96% exposed soil, 59,86% urban infrastructure, 12,72% of vegetation and 8,14% clouds. The amount of urban forest in the regions was variable, but it was unsatisfactory to promote environmental benefits to the population, except for the rural zone, where is the greater amount of vegetation in the city. It was observed that Olinda is a city with fewer vegetation, compared to some listed in this work. It was possible to realize that the city lacks of creation of green areas and afforestation.
Diagnóstico ambiental por índices de vegetação no Parque Estadual Mata da Pimenteira no período chuvoso e seco
(2022-05-23) Rocha, Alessandro Higor Gomes da; Bezerra, Alan Cezar; http://lattes.cnpq.br/3690303625468223; http://lattes.cnpq.br/1372320248183121
Due to the monitoring needs, as well as the understanding of the vegetation conditions of the Mata da Pimenteira State Park, it was aimed to analyze vegetation indices generated with red and infrared bands, with Sentinel-2 images to verify the land cover in the rainy and dry period from 2016 to 2021. The study site is located in the municipality of Serra Talhada, Pernambuco. The images were processed in Google Earth Engine to obtain a composition of the two periods studied, then, in QGIS software version 3.18.3 (Zurich), the vegetation indices (NDVI and VCI) were determined by raster calculator, a calculation tool available in Qgis that uses as a basis the values of the pixels of the layers. After obtaining the indices, the descriptive statistics of the images were obtained and classified using the r.recode tool, with subsequent counting of the vegetation classes by r.report, so that from this, the thematic map for the analysis and diagnosis of the study area was made. The results indicate higher average NDVI for 2016 and 2017 with 0.7 in the rainy period, and 0.36 in the dry one. The VCI had its highest average in 2016 with 86.04 and lowest in 2018 with 63.63. In the rainy period more than 90% of the area was composed of the high vegetation density with the NDVI and VCI of the very light class. In the dry period, most of the area was of the class "low vegetation density" by NDVI and "severe" by VCI.
Diagnóstico da arborização de vias públicas no entorno dos reservatórios elevados de água no município de Paulista-PE
(2018) Silva, Satyro Barbosa da; Duarte, Simone Mirtes Araújo; Silva, Hernande Pereira da; http://lattes.cnpq.br/1800835100486343; http://lattes.cnpq.br/5876968040869585; http://lattes.cnpq.br/6865576903260120
The benefits that urban tree-planting provides to communities where there are established trees, such as providing shade for pedestrians, physical soil stabilization, reducing the impact of rain, avoiding heat islands and biological deserts, provide scenic beauty and psychological well-being are indisputable, barring or channeling the wind and dampening the sound. However, there are many difficulties encountered in establishing an afforestation project in consolidated urban communities, mainly due to lack of planning, adequate urban furniture, telephony, sanitation and electrical equipment. Trees are sometimes considered as negative points of conflict, being blamed for destroying sidewalks, disrupting electrical wiring, breaking pipes and causing accidents by falling branches or falling over. Based on the principle that the more trees, the better the thermal sensation and the less the need to use treated water in the search for this balance, this work proposes an afforestation project around the five reservoirs administered by Companhia Pernambucana de Saneamento – COMPESA, as a way to benefit communities, not only with sanitation, but also with afforestation. For that, aerial-photogrammetric images of 0.50 x 0.50 m resolution were used, the census of the trees was carried out in the surroundings of the five reservoirs used in the study of the city of Paulista, from which several indices were obtained that allowed to evaluate and elaborate an afforestation plan in the roads that offered the physical conditions to do so. A total of 1,222 individuals were collected, distributed in 19 botanical families and 43 species, in which 86.7% of the species are exotic to the Brazilian flora and 13.3% are native. The most frequent species around the reservoirs were: Ficus benjamina L. (29.7%), Roystonea oleracea (Jacq.) O.F. Cook. (11.3%) and Terminalia catappa L. (10.8%). Based on current standards and similar literature, localities, quantity, adequate distance and species to be planted on the public, road were proposed in order to bring back the well-being that the population needs, totaling 415 trees distributed in 15 species of native origin. The study also shows the need for public intervention through campaigns to raise awareness of the importance of trees and especially in the structuring of roads that lack proper attention.
Dinâmica da cobertura da terra (2016-2023): um estudo no Parque Natural Municipal Mata do Frio e seu entorno, Paulista - PE
(2023-09-18) Lima, Richely da Silva; Lima Neto, Everaldo Marques de; Silva, Emanuel Araújo; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/6791561445213969; http://lattes.cnpq.br/5078677187654553
The Conservation Units (UC) in Brazil are used as tools for increasing environmental preservation and ecosystem protection. However, without proper monitoring and management, they become targets for degradation and environmental crimes, particularly those units located in urban areas, which are subject to anthropogenic pressure. Taking into consideration the reported cases of deforestation within the Municipal Natural Park Mate do Frio in Paulista - PE, this study aimed to analyze the land use and land cover of this conservation unit using remote sensing techniques, specifically the Normalized Difference Vegetation Index (NDVI), to assess changes over a 7-year interval based on satellite images from the Planet Satellite. To achieve this, study area cutouts from the years 2016 and 2023 were utilized, and a 1 km buffer was generated to assess the influence area. The NDVI calculation was applied to the cutout images. and the classes were reclassified based on their values into water, exposed soil, anthropized areas, low vegetation. and dense vegetation. Additionally, images from the studied years were correlated to identify changes in land cover. The study revealed an increase in dense vegetation within the UC, rising from 26.72% to 65.81%, along with a reduction in anthropized areas from 3.33% to 1.89% of the total area. Conversion of anthropogenic areas into low and dense vegetation was observed, accounting for 4.74% and 0.28%, respectively. Despite these positive findings, deforestation of 1.17 ha (5.70% of vegetation area) was noted. Concerning the surrounding area, there was an increase in anthropized areas from 33.17% to 47.12% due to urban expansion, with part of this anthropized area resulting from the deforestation of 67.41 ha of low vegetation. To validate the accuracy of the obtained data, the kappa index was used, showing values above 80% (very good) for the 2016 images and above 90% (excellent) for the 2023 images. The study results indicated that the environmental degradations in the specific UC were not significant but were noticeable, highlighting the need for increased monitoring, environmental education practices with the community, implementation of the unit's management plan, and the delineation of its buffer zone, given the anthropogenic pressure in the surrounding area of the Park.
Dinâmica de uso e cobertura da terra em floresta tropical seca no sertão pernambucano
(2019) Barreto, Thiago Henrique Lagos; Silva, Emanuel Araújo; Salami, Gabriela; http://lattes.cnpq.br/3382724343640625; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/9289202346311385
In view of the increasing degradation that the Caatinga suffers in recent years, mainly in Pernambucano, this work aimed to map the disturbance in dry tropical forest using medium resolution images in the city of Salgueiro-PE. For this, images of LANDSAT 5 satellite were used in the 1998 and 2008 periods, and LANDSAT 8 to 2018. All scenes were georeferenced by the Datum SIRGAS 2000 and the bands used were 5, 4 and 3 for LANDSAT 5 , and the bands 6, 5 and 4 for LANDSAT 8, generating the classification of six classes (forests, agriculture, exposed soil, water bodies, riparian forest and infrastructure) using the QGIS software and the Semi-Automatic Classification Plugin (SCP). The accuracy of the maps was verified by the Kappa coefficient. The three-year maps were crosschecked to quantify remaining forest, forest expansion and deforestation. The Kappa index found for 1998 was 0.72, for 2008 and 2018 it was above 0.8, indicating very good accuracy for 1998 and excellent for 2008 and 2018. In these 20 years, classes that declined in size were forests, exposed soil, bodies of water and infrastructure, being a perimeter of 48.2 km², 84.9 km², 9.4 km² and 16.7 km², respectively. While agriculture and riparian forest increased 81.0 km² and 78.1 km², respectively. The decrease of the exposed soil is due, mainly, to a period of extreme drought in 1998, where there was only rainfall in the first two months. The decrease of the forests is due to several factors, such as population increase, vegetal extraction in the region, an industrial pole of red ceramics and the transformations of the Caatinga in agricultural areas. When observed the changes that occurred in the classes in these 20 years, the remaining forest was 438.1 km², the forest expansion of 181.1 km² and the deforestation of 229.5 km², indicating a good regeneration capacity of the forests and a damaging process of deforestation in this municipality. Therefore, it is concluded that Salgueiro underwent intense anthropic actions harmful to vegetation during 1998 and 2018, provoking deforestation, which provides socioeconomic and environmental problems, thus showing the urgent need for efficient public action.
Dinâmica do risco de incêndios sob efeito do El Niño em paisagem do bioma Caatinga em Petrolina - PE
(2023-02-17) Feitosa, Márcio Faustino; Silva, Emanuel Araújo; Souza, Ioneide Alves de; http://lattes.cnpq.br/0383867840261318; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/7669915736150355
Remote sensing techniques have been used since the 1960s to work on a particular object or specific area. Over time, technologies have gained improvements and new software and high resolution satellites have emerged. The Landsat-8 satellite can capture scenes up to 705 km away from earth, with a percentage of 10% of clouds, these scenes can be processed and studied for a certain purpose, among these was the dynamics of fire risk under the effect of El Niño in landscape of the Caatinga Biome. Therefore, it is intended to study the vulnerability of Caatinga and the use of monitoring technologies. The objective of this work is to evaluate the influence of El Niño on the dynamics of fire risk under the landscape of the Caatinga biome in Petrolina-PE, monitoring the risk of fire in relation to the severity of El Niño. Images of the Lansat-8 satellite in the municipality of Petrolina-PE were obtained from the USGS website. The following criteria were adopted for image selection: Data from 2015 to 2020, counting from August 1st to December 31st, at times when there are few precipitations, few clouds and high temperatures due to the warmer months. The tool in data processing was Qgis software, a free license software, indicated for those seeking high quality in academic work in the area of Remote Sensing. In the interval of these years, six fire risk maps were obtained. Togenerate the forest fire risk maps and adopted the AHP methodology, widely used by several authors. Where we obtained eight variables: hypsometric map, land use and occupation, slope orientation, slope, road system, precipitation, surface temperature and vegetation index of the normalized difference. According to the results, a temporal analysis of fire risks was obtained, proving that the years 2015 to 2018 had a continuous increase, and in 2019 and 2020 there was a fall in fire risk. Between these last years there was an El Niño and a La Niña and 2020 was the year that the pandemic occurred, that is, there were few transport flows on the highways, causing low risk of fires.
Dinâmica espacial do cenário florestal em paisagens do bioma Caatinga no município de Araripina - PE
(2022-05-27) Andrade, Adrielle; Silva, Emanuel Araújo; Melo, Lorena de Moura; http://lattes.cnpq.br/1486808425687522; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/8750022516521279
Dinâmica espaço-temporal de índices de vegetação obtidos por VANT e Sentinel-2/MSI: análise do desenvolvimento da cultura de soja irrigada em pivô central
(2024-02-22) Silva, Mateus Dias Cezar da; Pandorfi, Héliton; http://lattes.cnpq.br/7981297368478991; http://lattes.cnpq.br/0572428299964865
Os índices de vegetação desempenham um papel preponderante no monitoramento do desenvolvimento de culturas agrícolas. Esse estudo buscou avaliar a dinâmica espaço-temporal de índices de vegetação em uma área de cultivo de soja irrigada por pivô central na Fazenda Agro Centro-Oeste localizada no município de São Luís de Montes Belos no estado de Goiás, utilizando imagens obtidas através de VANT e Sentinel-2/MSI, validando a aplicabilidade das imagens do satélite por meio de correlações com as imagens. Foram avaliados os índices de Vegetação Ajustado às Condições do Solo (SAVI), o Índice de Área Foliar (IAF, m2.m-2) e o Índice de Refletância Fotoquímica Modificado (MPRI) bem como a caracterização das chuvas na região de estudo. Todos dados foram submetidos à análise estatística descritiva para obtenção da média, desvio padrão e coeficiente de variação (CV, %) e posteriormente as análises geoestatísticas, a fim de dar apoio na caracterização da dinâmica dos índices de vegetação. No período de avaliação entre os anos de 2018 e 2019, observou-se que ao longo dos estádios de desenvolvimento da soja os índices tiveram comportamento homogêneo e crescente, apresentando um coeficiente de variação (CV) alto (CV=24%) no período inicial e final, mas após estabilização não se tem uma maior variabilidade do CV. Portanto, conclui-se que as análises dos índices de vegetação podem desempenhar um papel fundamental no acompanhamento e no gerenciamento da cultura da soja, bem como efetivar a aplicabilidade das imagens de satélite e correlacionar com as imagens de VANT, auxiliando os agricultores a tomar decisões informadas para otimizar a produção e economizar recursos.
Estimativa de variáveis dendrométricas a partir do sensor LiDAR no IPHONE 13 PRO
(2024-02-29) Santana, Larissa Maria Lopes; Silva, Emanuel Araújo; http://lattes.cnpq.br/2765651276275384; http://lattes.cnpq.br/0582963735033837
Nos últimos anos, o uso de tecnologias de sensoriamento remoto, como o sensor LiDAR, tornou-se fundamental na análise de variáveis dendrométricas em áreas florestais, e o LiDAR no iPhone 13 Pro representa uma inovação, oferecendo uma maneira acessível e eficiente de obter dados precisos sobre a estrutura florestal. Este trabalho visa avaliar a eficácia do sensor LiDAR do iPhone 13 Pro na estimativa de variáveis dendrométricas, como diâmetro e altura das árvores, em um povoamento de Mogno Africano, comparando medições com métodos convencionais para verificar a precisão dos dados obtidos pelo LiDAR. A tecnologia LiDAR tem avançado significativamente, permitindo a obtenção de dados tridimensionais detalhados sobre a estrutura das florestas, sendo amplamente utilizada na engenharia florestal para inventário, planejamento de manejo e modelagem de crescimento de árvores; aplicativos que incorporam LiDAR, como ForestScanner e Arboreal Tree, têm mostrado potencial em fornecer estimativas precisas de variáveis dendrométricas. O estudo foi realizado na Estação Experimental de Cana-de-açúcar do Carpina, em Pernambuco, onde dados de diâmetro e altura foram coletados manualmente e por meio dos aplicativos ForestScanner e Arboreal Tree, utilizando o sensor LiDAR do iPhone 13 Pro, e as medições foram comparadas para avaliar a precisão dos dados obtidos pelos aplicativos em relação aos métodos convencionais. Os resultados mostraram uma forte correlação entre as medições convencionais e as obtidas pelos aplicativos, com o ForestScanner apresentando um coeficiente de determinação (R²) de 0,852 e um RMSE de 2,24 cm para o diâmetro, enquanto o Arboreal Tree mostrou um R² de 0,9501 e um RMSE de 1,44 cm; para a altura, o Arboreal Tree apresentou um R² de 0,7857 e um RMSE de 1,31 m, indicando que ambos os aplicativos fornecem estimativas precisas e podem ser usados como alternativas eficientes aos métodos convencionais de medição. A utilização do sensor LiDAR no iPhone 13 Pro, combinada com os aplicativos ForestScanner e Arboreal Tree, mostrou-se eficaz na estimativa de variáveis dendrométricas, destacando a importância da tecnologia LiDAR em dispositivos móveis como uma ferramenta inovadora e prática para a coleta de dados florestais, oferecendo precisão, economia de tempo e custos na obtenção de informações sobre a estrutura florestal.
Histórico das mudanças nas classes de uso e cobertura do solo nas fazendas da Eucatex no estado de São Paulo
(2022-09-27) Silva, Yasmim Victória de Araújo e; Berger, Rute; Marques, Luísa Pereira; http://lattes.cnpq.br/1603075418219366; http://lattes.cnpq.br/5395827385005105; http://lattes.cnpq.br/0643141145421813
The consumption of forest-based products has increased in recent decades and forestry has been considered a strategic segment to collaborate and encourage the increase in the production of wood products. Until 2019, forest plantations represented about 9.8 million hectares in Brazil. The first land use and land cover classification system with remote sensing data aimed to identify the different categories of land classes. Land cover changes can be related to conversions, which are the complete replacement of one type of cover with another. The objective of this work was to measure the conversion of areas in forest management farms planted with Eucalyptus sp. of Eucatex Florestal, which are the scope of forest certification (FSC-FM), following the natural forest in an interval of up to 27 years (1994 – 2021). The study areas correspond to 51 Eucatex Florestal farms in the regions of Botucatu, Sorocaba and Bauru, in the state of São Paulo, distributed in 18 municipalities. In this study, Eucalyptus is not planted in conjunction with natural forest, but at the stands for commercial purposes, and areas with native species are separated by setbacks or trails. Imagery from satellite Landsat 5, 7, and 8 was used, depending on availability for the requested date. The images were downloaded from the Earth Explorer website and the maximum likelihood method was applied. Of the 51 farms analyzed, 43 had an increase in their natural forest areas and only eight had some type of vegetation loss. The Morrinhos Radar farm showed the greatest increase in the “natural forest” class since 1994, totaling 455.13 ha converted. The Santa Filomena farm had the greatest area loss, with 39.82 ha of its natural forest areas lost, and the other seven farms lost less than 10 ha. There was a increase in the natural forest cover spontaneously, without the application of forest restoration techniques, only with the isolation of the areas. The farms lost more area of natural forest before being acquired by the company, whereas after the implantation of the Eucalyptus stands, there was an increase in the natural areas. The culture of Eucalyptus sp. did not interfere with the regeneration of natural forest areas.
Identificação operacional de áreas com potencial de regeneração e/ou recuperação vegetal nas mesorregiões Sertão e São Francisco pernambucano
(2023-07-17) Gouveia, José Rafael Ferreira de; Nascimento, Cristina Rodrigues; http://lattes.cnpq.br/9289129949520610; http://lattes.cnpq.br/5471553264542605
Fire has been used for many years in Brazil, serving for various purposes. However, if handled wrongly, it can cause fires with immense damage to the environment. The Sertão and São Francisco Pernambucano mesoregions are susceptible to the occurrence of fires, since the predominant biome is the Caatinga, semi-arid climate and low rainfall. This article aims to characterize and quantify hot spots and operationally identify fires in the mentioned mesoregions in the period from 2014 to 2020, in the driest months of the year, as well as the power of plant regeneration. The images from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor on board the TERRA platform, products MOD14A1, MOD09GQ, MOD13Q1 and MOD09Q1 were used in order to characterize the areas affected by hot spots, analyze the Normalized Difference Vegetation Index (NDVI) and implement a script on the Google Earth Engine (GEE) platform for the operational identification of areas with potential for plant regeneration. Results show an increasing behavior in the number of hot spots, with a reduction in the year 2020. The year 2019 had the highest number of regenerated areas, with 37. The script proved to be effective with minimum hits of 56%, being in mostly greater than 75%. In addition, the maximum errors were 25% of omission in October 2020 and 43.75% of commission in September 2016. In this sense, the techniques employed were able to detect the regions affected by the fires, as well as their potential for plant regeneration.
Índice de salinidade do solo por sensoriamento remoto em bacia hidrográfica no submédio São Francisco
(2023) Dias, Maria Caroline da Silva; Lopes, Pabrício Marcos Oliveira; http://lattes.cnpq.br/0703321303981408; http://lattes.cnpq.br/0183881695413526
Influência das áreas verdes urbanas sobre a temperatura de superfície utilizando sensoriamento remoto
(2023-07-07) Siqueira, Ítalo Fernandes Pessôa; Alba, Elisiane; Oliveira, Géssyca Fernanda de Sena; http://lattes.cnpq.br/8717407990656771; http://lattes.cnpq.br/1465154212352591; http://lattes.cnpq.br/6866008330390945
Inteligência artificial na classificação de uso e cobertura da terra no semiárido de Pernambuco
(2020-11-03) Almeida, Gabriela Costa de; Silva, Emanuel Araújo; Moreira, Giselle Lemos; http://lattes.cnpq.br/6171199372079024; http://lattes.cnpq.br/2765651276275384
The Brazilian Tropical Dry Forest, known as Caatinga, is located in Brazil's northeastern region and has severe climatic characteristics, with dry weather and poorly distributed rainfall. Those climatic characteristics make Remote sensing analysis difficult due to its large vegetation differences between the dry and rainy periods. In order to help the remote sensing analysis in this biome, this work aims to test different Artificial Intelligence algorithms through supervised classification and to identify land use and land cover patterns in the city of Petrolina, in Pernambuco. Three algorithms were tested: Random Forest, Artificial Neural Networks, and K-Nearest Neighbors using QGIS and RStudio software based on Landsat 8 images from the dry period. Twenty samples from the classes were selected: Water, Agriculture, Urban Area, Forest, and Exposed Soil, and these samples served as a basis for training the algorithms for the classification of images. Occupancy data and precision quality assessment were obtained using Mapping Accuracy and Kappa Index, respectively: 0.9878706 and 0.9653555 for Random Forest; 0.9199973 and 0.9454833 for Artificial Neural Networks, 0.9873741 and 0.9598640 for K-Nearest Neighbors, all being considered excellent. These values were higher than those found in the most commonly used algorithms, as in the Maximum Likelihood algorithm. It was observed that the use of artificial intelligence algorithms could generate better results in the classification of land use in semiarid regions.