Monitoring The Growth of Cotton Plants and Determining Possible Stress Conditions and Developmental Disorders Wıth Images Obtained from Different Remote Sensing Platforms
Project no. 20854 · Status: Kapanmış (Yayın bekleniyor) · 2019
Summary
As well as being an important industrial plant grown as an annual crop, the cotton, is also a strategic agricultural product. It has the largest cultivation area and production value among all industrial crops and requires more input use compared to many other agricultural products. Especially the high cost of input and production in cotton cultivation calls for the entire process, from sowing until harvest, to be monitored meticulously. In Turkey, collection of information regarding cultivated area and yield of many agricultural products, and studies to identify diseases-pests are carried out using traditional methods such as consulting statements made by farmers or survey results. Most of the time, these methods do not yield accurate information. Recently, effective remote sensing technology has started to replace these traditional methods. Remote sensing technology provides high-accuracy results while performing tasks such as monitoring agricultural products, detecting the stress conditions of the plant during the cultivation period and, estimating the yield of plant production. Although there are studies that utilize the Unmanned Aerial Vehicle (UAV) technology in Turkey, studies focusing on plants with high economic value such as cotton have not yet become common. This study aims to examine, using new techniques, the cotton plant, which is widely cultivated in the Aegean region and has an important place in meeting domestic needs as well as in export. Certain tasks were carried out in this context; such as monitoring areas used for cultivation of cotton using data obtained from different remote sensing platforms such as UAV and Earth observation satellites (Sentinel-2), monitoring the plant throughout the entire vegetation period and, extracting its spectral signatures, analyzing multi-band data obtained from UAV and satellite images, and determining the physiological properties of the plant and the conditions that occur for different reasons (water, salinity, alkalinity, etc.) and cause stress on the plant. As part of the study, various vegetation indices were calculated using both terrestrial measurements and spectral data sets obtained from unmanned aerial vehicles. Correlation values between NDVI, one of the vegetation indices, and the nitrogen content of the cotton plant showed a strong statistical relationship ranging from 0.66 to 0.86. Additionally, correlation values between NDVI index data and the phosphorus content of the plant revealed a strong statistical relationship ranging from 0.71 and 0.86. Results showed that there was a significant and strong statistical relationship with potassium, another nutrient element, except for field-1, and similar information was obtained regarding calcium and magnesium. Multispectral images obtained from the unmanned aerial vehicle revealed the spectral characteristics of the cotton plant in different physiological periods. Plant-leaf samples were taken at each stage of the study and laboratory analyzes conducted using these samples revealed the nutrient contents of the cotton plant in various developmental stages. The relationships between all the data obtained in the study were statistically analyzed. Analyses of the data showed that there were statistically significant relationships between the yield values and the reflectance values of the cotton plant. Evaluations revealed that the end of the flowering period is more definitive when estimating the yield of the cotton plant and that referring to this period in studies is useful. Additionally, relationships between the values calculated using the irrigation indices of the data from the UAV and data such as irrigation time and water stress were identified, and the remote sensing data proved to be extremely useful in the generation of irrigation data.