• coolseason
• fertilization
• Greenseeker
• irrigation
• NDVI
• spectral reflectance
• warmseason

Like all modern agriculture sectors, turfgrass production and management is headed towards cost reduction, resource optimization and reduction of the environmental impact. In recent years, within the European Union several legislative, monitoring and coordinating actions have been undertaken to encourage sustainable use of resources, reduction in the use of chemicals and improvement of the urban environment. In this respect, two concepts that are strictly related to most of the aspects above are: “precision agriculture” and “precision conservation” and more specifically “precision turfgrass management.” Optical sensing has become a crucial part of precision turfgrass management and spectral reflectance in particular has been an active area of research for many years. However, while turfgrass status evaluation by proximity-sensed spectral reflectance appears to be an established and reliable practice, much more could be achieved in terms of monitoring of large turfgrass areas through remote sensing, and in particular through satellite imagery.
This thesis reports the results of four trials attempting:
a) to evaluate the spectral signatures of several turfgrass species\cultivars, for future use in satellite monitoring. This experimental study focused on 20 turfgrass species\cultivars, including perennial ryegrasses, tall fescues, kentucky bluegrasses, bermudagrasses (ecotypes, seeded and vegetatively propagated cultivars) and zoysiagrasses. Various agronomical and biological parameters were studied (quality, color, dry matter, chlorophyll, carotenoids, nitrogen content) and turfgrass spectral reflectance for all entries was gathered. Results showed that, within the same species, selected vegetation indices are often able to discriminate between different cultivars that have been established and maintained with identical agronomical practices. Evaluation of the spectral reflectance of plants using field spectroradiometry provides the possibility to identify different species\ cultivars, especially through the use of hyperspectral proximity and remote sensing;
b) to calculate on these 20 species and cultivars the most interesting vegetation indices by simulating the available wavelengths deriving from World View 2 satellite imagery. Results showed that within the same species selected vegetation indices are often able to discriminate between different varieties that have been established and maintained with identical agronomical practices;
c) to evaluate the proximity sensed spectral reflectance on Festuca arundinacea turf with 9 water replenishment levels (Linear Gradient Irrigation System) and 2 nitrogen conditions. ET0 was estimated using the Hargreaves and Samani method. The following parameters were determined: turf quality, drought tolerance, pest problems, temperature of the surface, clippings weight and relative nitrogen content, turf growth and soil moisture. Spectral reflectance data were acquired using a LICOR 1800 spectroradiometer. Pearson correlation coefficients were studied among all parameters and vegetation indices. Nitrogen fertilization influenced significantly turf quality, clippings weight, nitrogen content and turf growth. Water replenishment influenced significantly all parameters except nitrogen content. Among all parameters the highest correlation coefficient was registered relating drought tolerance with turf quality (r = 0.88) and with surface temperature (r = - 0.88). Among vegetation indices results showed that Water Index (WI) and Normalized Difference Water Index (NDWI), are better able to discriminate between different levels of water replenishment. Comparing WI with NDWI, the correlation coefficients were higher for Water Index in all the parameters, in particular the highest WI value was registered for drought tolerance (r = 0.91). This preliminary research demonstrates that spectral remote sensing can be a useful diagnostic tool to detect water stress in turfgrasses;
d) to compare N status in different turfgrasses, from remote multi-spectral data acquired by GeoEye-1 satellite and by two ground-based instruments. The study focused on creating a nitrogen concentration gradient on 3 warm-season turfgrasses (Cynodon dactylon x transvaalensis ‘Patriot’, Paspalum vaginatum ‘Salam’, Zoysia matrella ‘Zeon’) and 2 cool-season (Festuca arundinacea ‘Grande’, Lolium perenne ‘Regal 5’). The linear gradient ranged from 0 to 342 kg ha-1 of N for the warm-season and from 0 to 190 kg ha-1 of N for the cool-season turfgrasses. Proximity and remote reflectance measurements were acquired and used to determine Normalized Difference Vegetation Index (NDVI). Results showed that the N status is highly correlated with the spectral reflectance. Our results prove that NDVI measured with the ground-based instruments are highly correlated with data from satellite. The correlation coefficients between the satellite and the other sensors ranged from 0.90 to 0.99 for the warm-season and from 0.83 to 0.97 for the cool-season species. 'Patriot' had a clippings N concentration ranging from 1,20 % to 4.1 %, thus resulting the most reactive species to N fertilization. GeoEye-1 satellite can adequately assess the N status of different turfgrass species, and its spatial variability within a field depending on the N rates applied on the surfaces. In future information obtained from satellite could allow target management depending on the real need of the turf.