Noroviruses (NoVs) are the leading cause of acute gastroenteritis (AGE) outbreaks. Since 2014, novel genetic variants of NoV have been continuously identified and have caused a sharp increase in the number of AGE outbreaks. The specific geographical distribution and expanding genetic diversity of NoV has posed a challenge to conventional surveillance. Here, we describe the long-term dynamic correlation between NoV distribution in sewage and in the local population through the molecular surveillance of NoV in Guangdong, 2013-2018. The relative viral loads of the GI and GII genotypes in sewage were calculated through RT-PCR. A high-throughput sequencing method and operational taxonomic unit (OTU) clustering pipeline were developed to illustrate the abundances of different genotypes and genetic variants in sewage. Our results showed that the NoV viral loads and the emergence of new variants in sewage were closely associated with NoV outbreak risks in the population. Compared with the outbreaks surveillance, the dominance of the newly emerged variants, GII.P17-GII.17 and GII.P16-GII.2, could be detected one or two months ahead in sewage of a hub city. In addition, the dynamics of pre-epidemic variants, which were rarely detected in clinics, could be captured through sewage surveillance, thus improving our understanding of the origin and evolution of these novel epidemic variants. Our data highlight that sewage surveillance could provide nearly real-time and high-throughput data on NoV circulation in the community. With the advances in sequencing techniques, the sewage surveillance system could also be extended to other related infectious diseases. All rights reserved, Elsevier.