Loading...
Publication
Avaliação de sensores aplicada à refletância das folhas para a fertilização sustentável em olival
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.88 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Com a intensificação das culturas, nomeadamente do olival, surgiram preocupações
quanto ao consumo de água e ao impacto ambiental causado pelo uso de produtos
fitofarmacêuticos e fertilizantes. Assente nesta preocupação, na União Europeia a
estratégia Farm to Fork tem como algumas das principais metas até 2030 reduzir 50% a
utilização dos pesticidas químicos e diminuir fatores que causem pelo menos 50% das
perdas de nutrientes do solo. Com foco na eficiência de produção e sustentabilidade,
esta dissertação tem como objetivo principal, através da deteção proximal, validar
medições de refletância das folhas de oliveira (Olea europaea L.), obtidas com um sensor
multiespectral de baixo-custo e um espectrómetro de alta resolução, para o diagnóstico
nutricional em macronutrientes principais do olival. O ensaio decorreu no olival em sebe
do Departamento de Olivicultura do INIAV, I.P., em Elvas no ano de 2022. Para esse
efeito foi delineado um ensaio com fertilização diferenciada por fertirrega, durante o
ciclo vegetativo, em três cultivares de oliveira, ‘Azeiteira’, ‘Arbequina’ e ‘Koroneiki’. O
material vegetal foi recolhido na maturação dos frutos, somando um total de 120
amostras e um total de 14 400 folhas, onde posteriormente se procedeu às medições
de refletância. Após a recolha de dados com os sensores, as folhas foram submetidas a
análises químicas clássicas para quantificação de nutrientes. A correlação entre os dados
recolhidos pelos sensores com o teor de macronutrientes principais mostrou ser
bastante promissora com o espectrómetro de alta resolução onde se obtiveram os
melhores resultados para o Fósforo (R2: 0,67 – modelo KNeighborsRegressor) e
Potássio (R2: 0,60 – modelo SVR).
With the intensification of agriculture, notably olive cultivation, there have arisen concerns regarding water consumption and the environmental impact caused using phytopharmaceutical products and fertilizers. Against this backdrop, one of the European Union's key objectives through its Farm to Fork strategy, aiming for 2030, is to cut the use of chemical pesticides by 50% and to reduce the factors causing at least 50% of soil nutrient loss. Focused on the efficiency and sustainability of production, the principal aim of this dissertation is to validate olive leaf (Olea europaea L.) reflectance measures through proximal detection, obtained with a low-cost multispectral sensor and a high-resolution spectrometer, for nutritional assessment and diagnosis in the main macronutrients of the olive grove. The trial was conducted in the hedge olive grove of the INIAV's Department of Olive Culture in Elvas, in the year 2022. For this purpose, a trial with differentiated fertilization by fertigation was designed during the vegetative cycle on three olive tree cultivars, 'Azeiteira', 'Arbequina', and 'Koroneiki'. The plant material was collected when the fruits matured, totaling 120 samples and 14 400 leaves, where reflectance measurements were subsequently taken. Following data collection with the sensors, the leaves underwent classical chemical analyses for nutrient quantification. The correlation between the data collected by the sensors and the content of the main macronutrients proved very promising with the high-resolution spectrometer, where the best results were obtained for Phosphorus (R2: 0,67 – KNeighborsRegressor model) and Potassium (R2: 0,60 – SVR model). Conversely, the low-cost sensor presented some difficulty in corroborating the results observed in the laboratory analyses, necessitating further tests for its validation.
With the intensification of agriculture, notably olive cultivation, there have arisen concerns regarding water consumption and the environmental impact caused using phytopharmaceutical products and fertilizers. Against this backdrop, one of the European Union's key objectives through its Farm to Fork strategy, aiming for 2030, is to cut the use of chemical pesticides by 50% and to reduce the factors causing at least 50% of soil nutrient loss. Focused on the efficiency and sustainability of production, the principal aim of this dissertation is to validate olive leaf (Olea europaea L.) reflectance measures through proximal detection, obtained with a low-cost multispectral sensor and a high-resolution spectrometer, for nutritional assessment and diagnosis in the main macronutrients of the olive grove. The trial was conducted in the hedge olive grove of the INIAV's Department of Olive Culture in Elvas, in the year 2022. For this purpose, a trial with differentiated fertilization by fertigation was designed during the vegetative cycle on three olive tree cultivars, 'Azeiteira', 'Arbequina', and 'Koroneiki'. The plant material was collected when the fruits matured, totaling 120 samples and 14 400 leaves, where reflectance measurements were subsequently taken. Following data collection with the sensors, the leaves underwent classical chemical analyses for nutrient quantification. The correlation between the data collected by the sensors and the content of the main macronutrients proved very promising with the high-resolution spectrometer, where the best results were obtained for Phosphorus (R2: 0,67 – KNeighborsRegressor model) and Potassium (R2: 0,60 – SVR model). Conversely, the low-cost sensor presented some difficulty in corroborating the results observed in the laboratory analyses, necessitating further tests for its validation.
Description
Keywords
deteção proximal diagnóstico nutricional eficiência de produção Olea europaea L.. efficiency of production nutritional diagnosis proximal detection