Influence of measuring angle, nitrogen fertilization, and variety on spectral reflectance of winter oilseed rape canopies

A field experiment with winter oilseed rape was conducted near Göttingen (northern Germany) in the growing season 1998–99. Twelve varieties were compared at two nitrogen (N) application rates (0, 206 kg N ha–1) regarding shoot dry matter production, shoot N content, shoot N uptake (beginning of shooting, beginning of flowering), and seed yield. Canopy reflectance was measured one week before the beginning of shooting and one week before flowering at different wavelengths between 550 and 940 nm using a spectral reflectance sensor, type FAL II. The vegetation indices red edge inflection point (RIP), soil adjusted vegetation index (SAVI), and normalized vegetation index (NDVI) introduced in literature as indicators of growth parameters were calculated from the reflectance values and compared with growth parameters. The results showed that the measuring angle (–45° to 45°) had a significant influence on the calculated vegetation indices. Therefore, the measuring angle should be kept constant during the measurements. Nitrogen fertilization led to an increase of all vegetation indices. The results revealed significant differences between the tested varieties, indicating problems regarding an accurate estimation of the N status of the crop. Consistent differences in the vegetation indices appeared between varieties with extremely different growth parameters. However, no general relationship between growth parameters and vegetation indices of the 12 varieties could be found. Therefore, variety‐specific calibrations are necessary. The indices RIP and SAVI resulted in similar statements and seem to be good parameters regarding N effects and variety differences, whereas NDVI could resolve variety differences in the fertilized treatments only imperfectly. Einfluss von Messwinkel, Stickstoffdüngung und Sorte auf die spektrale Reflexion von Winterrapsbeständen In der Vegetationsperiode 1998–99 wurde bei Göttingen ein Feldversuch mit Winterraps durchgeführt. Zwölf Sorten wurden in zwei N‐Stufen (0, 206 kg N ha–1) hinsichtlich Sprosstrockenmasse, N‐Gehalt im Spross und N‐Aufnahme in den Spross (Schossbeginn und Blühbeginn) sowie Samenertrag verglichen. Die Lichtreflexionswerte der Bestände wurden eine Woche vor Schossbeginn sowie eine Woche vor Blühbeginn für den Wellenlängenbereich von 550–940 nm mit Hilfe eines spektraloptischen Sensors, Typ FAL II, ermittelt. Aus den Reflexionswerten verschiedener Wellenlängen wurden anschließend drei aus der Literatur als Indikatoren