No abstract

Abstract. Landscape epidemiology has made significant strides recently, driven in part by increasing availability of land cover data derived from remotely-sensed imagery. Using an example from a study of land cover effects on hantavirus dynamics at an Atlantic Forest site in eastern Paraguay, we demonstrate how automated classification methods can be used to stratify remotely-sensed land cover for studies of infectious disease dynamics. For this application, it was necessary to develop a scheme that could yield both land cover and land use data from the same classification. Hypothesizing that automated discrimination between classes would be more accurate using an object-based method compared to a per-pixel method, we used a single Landsat Enhanced Thematic Mapper+ (ETM+) image to classify land cover into eight classes using both per-pixel and object-based classification algorithms. Our results show that the objectbased method achieves 84% overall accuracy, compared to only 43% using the per-pixel method. Producer's and user's accuracies for the object-based map were higher for every class compared to the per-pixel classification. The Kappa statistic was also significantly higher for the object-based classification. These results show the importance of using image information from domains beyond the spectral domain, and also illustrate the importance of object-based techniques for remote sensing applications in epidemiological studies.

Quantitative, spatially explicit estimates of canopy nitrogen are essential for understanding the structure and function of natural and managed ecosystems. Methods for extracting nitrogen estimates via hyperspectral remote sensing have been an active area of research. Much of this research has been conducted either in the laboratory, or in relatively uniform canopies such as crops. Efforts to assess the feasibility of the use of hyperspectral analysis in heterogeneous canopies with diverse plant species and canopy structures have been less extensive. In this study, we use in situ and aircraft hyperspectral data to assess several empirical methods for extracting canopy nitrogen from a tallgrass prairie with varying fire and grazing treatments. The remote sensing data were collected four times between May and September in 2011, and were then coupled with the field-measured leaf nitrogen levels for empirical modeling of canopy nitrogen content based on first derivatives, continuum-removed reflectance and ratio-based indices in the 562-600 nm range. Results indicated that the best-performing model type varied between in situ and aircraft data in different months. However, models from the pooled samples over the growing season with acceptable accuracy suggested that these methods are robust with respect to canopy heterogeneity across spatial and temporal scales.

No abstract