Depending on environmental factors fungal diseases of crops are often distributed heterogeneously in fields. Precision
agriculture in plant protection implies a targeted fungicide application adjusted these field heterogeneities. Therefore an
understanding of the spatial and temporal occurrence of pathogens is elementary. As shown in previous studies, remote
sensing techniques can be used to detect and observe spectral anomalies in the field. In 2008, a sugar beet field site was
observed at different growth stages of the crop using different remote sensing techniques. The experimental field site
consisted of two treatments. One plot was sprayed with a fungicide to avoid fungal infections. In order to obtain sugar
beet plants infected with foliar diseases the other plot was not sprayed. Remote sensing data were acquired from the
high-resolution airborne hyperspectral imaging ROSIS in July 2008 at sugar beet growth stage 39 and from the HyMap
sensor systems in August 2008 at sugar beet growth stage 45, respectively. Additionally hyperspectral signatures of
diseased and non-diseased sugar beet plants were measured with a non-imaging hand held spectroradiometer at growth
stage 49 in September. Ground truth data, in particular disease severity were collected at 50 sampling points in the field.
Changes of reflection rates were related to disease severity increasing with time. Erysiphe betae causing powdery
mildew was the most frequent leaf pathogen. A classification of healthy and diseased sugar beets in the field was
possible by using hyperspectral vegetation indices calculated from canopy reflectance.
A fast and precise sensor-based identification of pathogen infestations in wheat stands is essential for the implementation
of site-specific fungicide applications. Several works have shown possibilities and limitations for the detection of plant
stress using spectral sensor data. Hyperspectral data provide the opportunity to collect spectral reflectance in contiguous
bands over a broad range of the electromagnetic spectrum. Individual phenomena like the light absorption of leaf
pigments can be examined in detail. The precise knowledge of stress-dependent shifting in certain spectral wavelengths
provides great advantages in detecting fungal infections. This study focuses on band selection techniques for
hyperspectral data to identify relevant and redundant information in spectra regarding a detection of plant stress caused
by pathogens. In a laboratory experiment, five 1 sqm boxes with wheat were multitemporarily measured by a ASD
Fieldspec® 3 FR spectroradiometer. Two stands were inoculated with <i>Blumeria graminis</i> - the pathogen causing
powdery mildew - and one stand was used to simulate the effect of water deficiency. Two stands were kept healthy as
control stands. Daily measurements of the spectral reflectance were taken over a 14-day period. Three ASD Pro Lamps
were used to illuminate the plots with constant light. By applying band selection techniques, the three types of different
wheat vitality could be accurately differentiated at certain stages. Hyperspectral data can provide precise information
about pathogen infestations. The reduction of the spectral dimension of sensor data by means of band selection
procedures is an appropriate method to speed up the data supply for precision agriculture.
Plant stresses, in particular fungal diseases, show a high variability in spatial and temporal dimension with respect to
their impact on the host. Recent "Precision Agriculture"-techniques allow for a spatially and temporally adjusted pest
control that might reduce the amount of cost-intensive and ecologically harmful agrochemicals. Conventional stressdetection
techniques such as random monitoring do not meet demands of such optimally placed management actions.
The prerequisite is an accurate sensor-based detection of stress symptoms. The present study focuses on a remotely
sensed detection of the fungal disease powdery mildew (Blumeria graminis) in wheat, Europe's main crop. In a field
experiment, the potential of hyperspectral data for an early detection of stress symptoms was tested. A sophisticated
endmember selection procedure was used and, additionally, a linear spectral mixture model was applied to a pixel
spectrum with known characteristics, in order to derive an endmember representing 100% powdery mildew-infected
wheat. Regression analyses of matched fraction estimates of this endmember and in-field-observed powdery mildew
severities showed promising results (r=0.82 and r<sup>2</sup>=0.67).