Remote sensing products validation activity and observation network in China

Well design and coordinated implementation of validation activity is necessary to evaluate the accuracy of remote sensing product. However, validation is not a straightforward task and remain many challenges. A generally recognized difficult issue is the inconsistence between sparse observations and remote sensing pixels, strong spatial and temporal variations of surface variables, and the intrinsic heterogeneity of land surfaces. Thus, to develop, design and conduct reasonable validation schemes and activities to acquire ground truth at pixel scale over heterogeneous land surfaces is urgently needed. This contains, from the perspective of measurement, integrating various ground observations collected at multi-scale, in order to validate different types of RSPs from site to network, especially for those land surface variables with strong spatial-temporal variations. To this end, a dedicated validation initiative has been launched in China since 2011. The main scientific objectives and research contents are to develop mathematical approaches for spatial sampling optimization to acquire the ground truth at pixel scale over heterogeneous land surfaces, to form a series of recognized and practicable technical specifications to guide validation of various RSPs, and to establish a prototype of national validation network for long term operation. Specific validation activities, such as HiWATER, were conducted from site to network, through multi-scale observations collected from multi-platform and multi-source sensors, to experimentally examine those proposed methodologies and guidelines. Following the experience of these validation exercises, we are coordinating a Chinese validation network to use standardized and recognized technical specifications in implementing future validation attempts, aiming to extend validation exercises from point scale to regional scale and to national scale across different zones.