The importance of source data in river network connectivity modeling: A review

River network connectivity (RC) describes the hydrologic exchange of water, nutrients, sediments, and pollutants between the river channel and other “sites” via heterogenous flowpaths along the river corridor. As water moves downstream it carries these constituents, creating a stream‐to‐ocean continuum of connectivity that regulates global water, carbon, and nutrient cycling. River network connectivity models have developed over many decades, culminating in recent years with network‐scale RC models that explicitly simulate the transport and exchange of water and elements from headwaters to coasts, sometimes requiring models to contain tens of millions of river reaches. These advances provide transformative insights into the aggregate effects of RC on water and material transport across scales from local to global. Yet, recent reviews have pointed to several challenges that need to be overcome to continue advancing network‐scale RC modeling. In service of these goals, I summarize recent network‐scale RC maps and models to identify similarities and differences across the large‐scale RC modeling landscape. Although our computational and upscaling abilities have significantly improved and have revealed new insights, current models are still limited by the quantity, quality, resolution, and lack of standardization of the available in situ databases and source data maps necessary for the modeling. This suggests that we can extend recent advances if we keep improving these source datasets, while continuously revisiting our physics and theory to explain those new data. In doing so, we will continue to expand the role of network‐scale RC models in informing water quality modeling and management into the future.