Fusion of aerial gamma-ray survey and remote sensing data for a deeper understanding of radionuclide fate after radiological incidents: examples from the Fukushima Dai-Ichi response

摘要

We consider fusion of heterogeneous data, consisting of multispectral imagery collected from the WorldView-2 satellite and aerial gamma-ray spectra collected during the NNSA Aerial Measuring System response to the Fukushima Dai-ichi Nuclear Power Plant crisis. These data are analyzed using the non-linear dimension reduction technique of diffusion maps. The data is then used for classification of regions in the scene. A strong spatial coherence is observed in the gamma-ray data that is closely coupled to the underlying terrain classification obtained from the multispectral data, indicating correlation between these a priori uncorrelated measurements.