Multitemporal data mining: From biomass monitoring to nuclear proliferation detection

We are living in an era of unprecedented population growth and migration, expanding urban and agriculture lands, depleting forests and portable water resources, and natural hazards and climate changes that are changing the face of the planet Earth. Multitemporal remote sensing observations provide a powerful means to monitor the Earth to identify and characterize these changes in near-real time. Data mining is proven to be highly useful in analyzing multi-resolution, multi-spectral, multisensor, and multi-temporal remote sensing data. In this paper we describe the state-of-the-art data mining approaches with applications in biomass and critical infrastructure monitoring.