Surface soil moisture information is needed for monitoring and modeling surface processes at various spatial scales. While many reflectance based soil moisture quantification models have been developed and validated in laboratories, only few were applied from remote sensing platforms and thoroughly validated in the field. This paper addresses the issues of a) quantifying surface soil moisture with very high resolution spectral measurements from remote sensors in a landscape with sandy substrates and low vegetation cover as well as b) comprehensively validating these results in the field. For this purpose, the recently developed Normalized Soil Moisture Index (NSMI) has been analyzed for its applicability to airborne hyperspectral remote sensing data. Three HyMap scenes from 2004 and 2005 were collected from a lignite mining area in southern Brandenburg, Germany. An NSMI model was calibrated (R2=0.92) and surface soil moisture maps were calculated based on this model. An in-situ surface soil moisture map based on a combination of Frequency Domain Reflectometry (FDR) and gravimetric data allowed for validating each image pixel (R2=0.82). In addition, a qualitative multitemporal comparison between two consecutive NSMI datasets from 2004 was performed and validated, showing an increase in estimated surface soil moisture corresponding with field measurements and precipitation data. The study shows that the NSMI is appropriate for modeling surface soil moisture from high spectral-resolution remote sensing data. The index leads to valid estimations of soil moisture values below field capacity in an area with sandy substrates and low vegetation cover (NDVI < 0.3). Further studies will analyze the validity of the NSMI for surface soil moisture estimation from spaceborne hyperspectral sensors like the Environmental Mapping and Analysis Program (EnMap) in different landscapes.
Classifying remotely sensed images from urban environments is challenging. Urban land cover classes are spectrally heterogeneous and materials from different classes have similar spectral properties. Image segmentation has become a common preprocessing step that helped to overcome such problems. However, little attention has been paid to impacts of segmentation on the data's spectral information content. Here, urban hyperspectral data is spectrally classified using support vector machines (SVM). By training a SVM on pixel information and applying it to the image before segmentation and after segmentation at different levels, the classification framework is maintained and the influence of the spectral generalization during image segmentation hence directly investigated. In addition, a straightforward multi-level approach was performed, which combines information from different levels into one final map. A stratified accuracy assessment by urban structure types is applied. The classification of the unsegmented data achieves an overall accuracy of 88.7%. Accuracy of the segment-based classification is lower and decreases with increasing segment size. Highest accuracies for the different urban structure types are achieved at varying segmentation levels. The accuracy of the multi-level approach is similar to that of unsegmented data but comprises the positive effects of more homogeneous segment-based classifications at different levels in one map.
Data of Landsat-7 Enhanced Thematic Mapper (ETM+) and Terra Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) have been examined to reveal differences and similarities concerning urban change analysis. A mosaic from two ASTER scenes and a Landsat ETM+ scene were compared on the basis of invariant targets. A sensitivity analysis was conducted to determine the influence of geometric and spectral differences between these two sensors. It could be shown that - despite comparable sensor concepts - dissimilarities in geometric and spectral properties lead to measurable variations in statistical and thematic indicators. However, it is believed that notwithstanding these differences, a multi-temporal change analysis based on Landsat-TM, ETM+ and Terra-ASTER is a feasible approach to monitor urban change, if divergences are correctly understood and the related uncertainties can be quantified.