Research on monitoring and evaluation of changes during the development process of cotton plant using multispectral camera

This study was carried out in cotton plant production areas in order to determine the relationships between ground observation values and spectral bands / indices, to make recommendations in terms of bands and indices in estimating observation values, to investigate the advantages of remote sensing systems and the possibilities of their use in agricultural lands. This study was carried out in the trial land of GAP International Agricultural Research and Training Center Directorate in 2024. 19 cotton genotypes were used as plant material. The trial was carried out according to the randomized block design with 4 replications. Plant height (cm), 100 seed weight (gr), ginning percentage (%), seed cotton yield, fiber length (mm), fiber fineness (mic.), fiber strength (gr/tex), fiber maturity (%), spinning consistency index (SCI), short fiber index (%), fiber yellowness (+b), fiber brightness (Rd) fiber technological properties were taken as ground observations. Within the scope of the study, bands obtained from GEOSAT-2 on 20 June, Jilin-1 KF-01 on 19 July, and SuperView-1 satellites on 9 September and 38 indices obtained from these bands were examined. Correlation analysis was performed between the examined ground observations and the examined band/indices. Statistically significant relationships were determined between 100 seed weight, plant height, seed cotton yield, yarn feasibility index, fiber fineness, fiber maturity, fiber length, short fiber, fiber brightness properties and some spectral bands/indices, but no statistically significant relationships were determined with ginning efficiency, fiber strength and fiber yellowness properties. Generally low level correlations were determined between fiber quality properties and examined spectral bands/indices, and it was concluded that direct estimation of these properties from multispectral bands was limited. However, one weak relationship was observed in fiber length property. Indices obtained from multispectral images can provide valuable information about the morphological development process of cotton plant. Temporal analyses of June, July and September months revealed the traceability of periodic spectral changes and supported the usability of remote sensing techniques in terms of phenological monitoring in cotton. The bands/indices used in the study can be evaluated as auxiliary parameters in studies on the estimation of yield and fiber quality properties. If fiber quality properties show a weak relationship with the existing bands, sensors with more bands or different imaging techniques with high resolution can be used to monitor these parameters. Analyses made with images obtained at different dates are important in terms of monitoring periodic development. It is recommended that such data be evaluated with a time series approach. Since it would be more beneficial to conduct studies with satellite images from a much later period (August - September) than from an early period, it is recommended that later studies be conducted. Repeating the study in different regions and with different varieties will increase the generalizability of the results.