SATELLITE REMOTE-SENSING OFFERS SCALABLE GLOBAL ESTIMATES OF RIVER SUSPENDED SEDIMENT TRANSPORT

Remote sensing of river water offers potential insight into river processes operating on broader spatial and temporal scales than can be easily monitored with in situ methods. Though remote-sensing techniques have been used in various river monitoring applications for decades, river-to-river variability has to this point limited successful application beyond individual case studies. We constructed a record of > 130,000 in situ measurements of suspended sediment concentration (SSC) paired with contemporaneous Landsat satellite images of the same location to develop globally applicable calibrations for detecting SSC on rivers > 90 m wide. Calibration data from a wide range of physiographic settings on four continents allowed for the automated grouping of rivers that have similar typical river color. Because this river color grouping is made absent any in situ data, it can be applied to any river location worldwide, allowing for SSC estimates with an average uncertainty of 73%, an important improvement over a single “global” calibration. Rivers transporting abundant sand as suspended load and/or with high fractional particulate organic carbon (POC) nonetheless present challenges for remote sensing. We quantify these effects and account for these and other river-specific challenges to SSC estimation by correcting calibrations for individual river systems. Critically for researchers interested in using these techniques, SSC estimate uncertainties can be reduced substantially with only minimal site-specific in situ data; even as few as 5–10 in situ­ sample–satellite pairs improve average uncertainty to 49%. We illustrate the opportunities offered by these techniques for understanding large-scale patterns and trends in fluvial suspended sediment. Because the Landsat satellite record extends from the mid-1980s to the present for most locations worldwide, these methods can be combined with field and other remote sensing data to quantify rapid changes occurring in watersheds across the globe.