Hydrologists Tamlin Pavelsky of UNC-Chapel Hill and George Allen of Texas A&M University used a combination of satellite imagery and field measurements coupled with statistical modeling to calculate worldwide river and stream surface measurements (related: Global Dataset of River Widths Developed from Landsat Imagery). The end result is the Global River Widths from Landsat (GRWL) Database which was developed from GIS data collected from 3,693 gauge stations and 7,376 Landsat TM, ETM+ and OLI scenes. These researchers calculated the total river and stream surface area (RSSA) from these data sources at a volume that is about 44% higher than previously estimates based on extrapolations of small sample sizes. “We added up all the rivers and streams around the world [and] got an area larger than the size of Texas. The estimate was significantly larger than the previous best estimate, particularly in the Arctic,” noted George Allen, an associate professor of geography.
This database will be used to help assess carbon emissions. “As we try to mitigate the effects of climate change, it’s really important that we clearly understand where the carbon that we are emitting goes, and that requires us to accurately quantify the global carbon cycle,” said Tamlin Pavelsky, senior author and associate professor of global hydrology in the College of Arts and Sciences at UNC-Chapel Hill. “Our new calculation helps scientists better assess how much carbon dioxide is moving from rivers and streams into the atmosphere each year.”
This data will also be used by the Surface Water and Ocean Topography satellite (SWOT), launching in 2021. Part of the SWOT mission will be to measure continental water surfaces, including the width, height, and slope of rivers and the surface area and elevations of lakes. The GRWL GIS dataset will be used to facilitate analysis of the raw SWOT data.
The R code to calculate River and Stream Surface Area (RSSA) is freely available for download from Github.
Download the data: Global River Widths from Landsat (GRWL) Database