Causal Effect of Impervious Cover on Annual Flood Magnitude for the United States. Issue 5 (5th March 2020)

• Record Type: Journal Article
• Authors: Blum, Annalise G.; Ferraro, Paul J.; Archfield, Stacey A.; Ryberg, Karen R.
• Is Part Of: Geophysical research letters. Volume 47:Issue 5(2020)
• Page Start: no
• Page End: no
• Publication Date: 2020-03-05
• Subjects: floods -- urbanization -- causal -- inference -- panel regression; Geophysics -- Periodicals; Planets -- Periodicals; Lunar geology -- Periodicals; 550
• DOI: 10.1029/2019GL086480
• Languages: English
• ISSNs: 0094-8276

Abstract

Abstract: Despite consensus that impervious surfaces increase flooding, the magnitude of the increase remains uncertain. This uncertainty largely stems from the challenge of isolating the effect of changes in impervious cover separate from other factors that also affect flooding. To control for these factors, prior study designs rely on either temporal or spatial variation in impervious cover. We leverage both temporal and spatial variation in a panel data regression design to isolate the effect of impervious cover on floods. With 39 years of data from 280 U.S. streamgages, we estimate that a one percentage point increase in impervious basin cover causes a 3.3% increase in annual flood magnitude (95%CI: 1.9%, 4.7%) on average. Using 2, 109 streamgages, some of which have upstream regulation and/or overlapping basins, we estimate a larger effect: 4.6% (CI: 3.5%, 5.6%). The approach introduced here can be extended to estimate the causal effects of other drivers of hydrologic change. Key Points: We estimate that annual floods increase by 3.3%, on average, for each percentage point increase in impervious basin cover This is the first study to apply a panel regression design to estimate the causal effect of impervious cover on floods Our approach demonstrates how to leverage temporal and spatial variation to isolate a causal effect, separate from other drivers of change