A novel low- cost chlorophyll fluorescence Sensor for early detection of environmental pollution

• Ano de defesa: 2017
• Autor(a) principal: Gull, Christopher Johannes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Detectores; Poluição; Fluorescência; Clorofila; Custo; Ciência da Computação
• Link de acesso: http://www.locus.ufv.br/handle/123456789/18061

Pollution seriously affects all living organisms as well as economies directly or indirectly relying on natural growth resources. Monitoring the environment for stresses in plants, caused by pollutants, is necessary in order to anticipate and counteract the adverse effects before they manifest as visible damage. Failure to constantly monitor plants results in decreased crop growth, damage to ecosystems, health-related issues, and, ultimately, economic losses. Especially around affected areas, such as waste deposits, mining activities and factories, but also in and around urban areas, it is important to acknowledge the potential environmental issues that may arise from human activities. Among the consequences we find acid rain, heavy metal contamination, surface ozone, changes in temperature, and drought, contributing to alterations in plant physiology, specifically chlorophyll content and photosynthetic efficiency. Measuring plant efficiency, thus health, using commercial fluorometers, such as PAM (pulse-amplitude modulation) devices, presents a challenge, since cost, complexity and the measurement methods make real-time monitoring a difficult proposition. Although such devices are high-precision instruments, they are merely able to provide ‘snapshots’ of small areas. This makes it difficult to understand the health of plants over large areas and over extended periods of time, frequently resulting in actions taken only after significant changes to plants and productivity. One solution would be for a farmer in an area impacted by pollution to acquire multiple of these devices and to employ a workforce dedicated solely to monitoring plant health, but this is costly and inefficient. Another solution would be to simplify the devices with which to measure, and use a multitude of these. Indeed, in this work, we focus on solving this problem, by reducing costs and complexity, and eliminating the need for human input in the measurement process. We propose a system of low-cost chlorophyll fluorescence sensors able to monitor a large number of individual plants at the same time and wirelessly. These sensors have been designed, prototyped and built from the ground up to provide reasonable accuracy, and capacity for discriminating between plants subjected to stress from non-stressed plants. Where our sensor system, the CFY (chlorophyll fluorescence yield) Sensor, lacks in accuracy, it compensates with a multitude of potential simultaneous measurements from an array of sensors within a network. For this reason, the sensor prototype is inherently designed for wireless sensor networks (WSN). Using two species of plants, Clusia hilariana and Paspalum densum, we have built, tested and verified our methodologies and prototype sensors through a series of experiments. Through these, we observed significant results when employed in an emulated sensor network using one sensor on a large number of plants over extended periods of time. Differentiating the stressed group from the control group was possible, in addition to rapid and well before any visible damage had manifested on leaves. We conclude that it is indeed possible to not only detect plant stress using low-cost methods, but also to do so automatically and in real-time, allowing for early-detection of pollution and providing e.g. a farmer enough time to resolve problems before they become irreversible and costly.