Sistemas imageadores acoplados em veículo aéreo remotamente pilotado no monitoramento de estresse hídrico em cultivos de algodão em Sorriso - MT

• Ano de defesa: 2022
• Autor(a) principal: Souto, Roberto Nunes Vianconi
• Orientador(a): Ferreira, Manuel Eduardo
• Banca de defesa: Ferreira, Manuel Eduardo, Alves Júnior, Leomar Rufino, Nascimento, Abadia dos Reis, Luiz, Gislaine Cristina, Jesuz, Cleberson Ribeiro de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Goiás
• Programa de Pós-Graduação: Programa de Pós-graduação em Geografia (IESA)
• Departamento: Instituto de Estudos Socioambientais - IESA (RMG)
• País: Brasil
• Palavras-chave em Português: Sensoriamento remoto; RPA; Estresse hídrico; Cultura agrícola; Algodã
• Palavras-chave em Inglês: Remote sensing; UAV; Water stress; Crop; Cotton
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS::GEOGRAFIA FISICA::HIDROGEOGRAFIA
• Link de acesso: http://repositorio.bc.ufg.br/tede/handle/tede/12981

With global climate change, extreme events are happening more frequently in the legal Amazon region. This affects both ecological and socioeconomic aspects, altering the landscape over vast areas. Increased fires and deforestation in the region, together with irregular hydrological events, exacerbate climate change, especially in the Amazon Basin and in Mato Grosso, the state that has been for many years a leader in deforestation and climate variation in the country. Crops such as cotton are particularly sensitive to these changes and the resulting water stress. Therefore, monitoring this indicator to understand its effects on Mato Grosso territory and environmental dynamics is essential. The use of remote sensing through Unmanned Aerial Vehicles (UAVs) offers promising prospects for agricultural and environmental monitoring. In this context, the present research aimed to evaluate the applicability of sensors attached to a multirotor UAV to detect and monitor water stress in cotton planting. More specifically, we sought to analyze the performance of thermal and multispectral sensor imaging in the detection of water stress in a cotton planting experiment. In addition, the possible interference of crop varieties and covariates such as meteorological, climatic, and edaphic conditions on crop productivity, water stress detection, and productivity itself were evaluated. The methodology adopted was based on intensive field activities to parametrize the state of cotton plantations, considering water stress conditions in an experimental area at the Training and Technological Dissemination Center of the Northern Regional Nucleus of Mato Grosso Cotton Institute, located in the county of Sorriso - MT. The methodological procedures included three main blocks: field surveys, analysis of the spectral behavior of targets in multispectral and thermal images, and statistical analysis. As a result, it was found that thermal and multispectral sensors have the potential to identify areas of water stress. In general, the performance of thermal imaging. However, the multispectral sensor showed a relatively higher discrimination capacity in identifying stress in the early stages of the crop. Compared to the thermal sensor, another positive aspect of multispectral imaging is its strength against momentary weather variations. It is worth mentioning that the model combining monitoring with both types of sensors, thermal and multispectral, showed significantly better results.