OCAGraph: An effective observation capability association model for Earth observation sensor planning

Abstract

• Association relations among Earth observation (EO) sensors are modeled as a graph. • N -ary association relations among EO sensors are characterized for the first time. • Time-varying association relations are captured by multiple space–time snapshots. • Collaborative observation planning can be achieved comprehensively and optimally. The optimal collaborative observation planning of Earth observation (EO) sensors is essential for disaster monitoring. This goal requires comprehensively characterizing the observation capability associations among these sensors. Previous studies only solved the static binary-associated observation capability for the optimized collaborative planning of two-sensor combinations, while the dynamic N -ary-associated observation capability that supports the optimized collaborative planning of N -sensor combination remained to be addressed. This study proposes a space–time composite observation capability association graph (OCAG) model, where its constituent sensors and association relations are modeled as nodes and edges, respectively, to address this gap. The proposed model divides an observation scenario into several states, with each state utilizing an OCAG to characterize the association relations, especially the N -ary association relations, among the EO sensors. The N -ary association relations can function as optimized observation plans to satisfy various observation requirements. A flood observation planning experiment was conducted in Hubei, China. Experimental results indicate that all association relations, especially the N -ary ones, can be solved at multiple space–time snapshots, enabling sensor planners to comprehensively and optimally collaborate EO sensors for flood observation.