Sand mining far outpaces natural supply in a large alluvial river

Abstract. The world's large rivers are facing reduced sediment loads due to anthropogenic activities such as hydropower development and sediment extraction. Globally, estimates of sand extraction from large river systems are lacking, in part due to the pervasive and distributed nature of extraction processes. For the Mekong River, the widely assumed estimate of basin-wide sand extraction is 50 Mt per year. This figure is based on 2013 estimates and is likely to be outdated. Here, we demonstrate the ability of high-resolution satellite imagery to map, monitor, and estimate volumes of sand extraction on the Lower Mekong River in Cambodia. We use monthly composite images from PlanetScope imagery (5 m resolution) to estimate sand extraction volumes over the period 2016–2020 through tracking sand barges. We show that rates of extraction have increased on a yearly basis from 24 Mt (17 to 32 Mt) in 2016 to 59 Mt (41 to 75 Mt) in 2020 at a rate of ∼8 Mt yr−1 (6 to 10 Mt yr−1), where values in parentheses relate to lower and upper error bounds, respectively. Our revised estimates for 2020 (59 Mt) are nearly 2 times greater than previous best estimates for sand extraction for Cambodia (32 Mt) and greater than current best estimates for the entire Mekong Basin (50 Mt). We show that over the 5-year period, only 2 months have seen positive (supply exceeds extraction) sand budgets under mean scenarios (5 months under the scenarios with the greatest natural sand supply). We demonstrate that this net negative sand budget is driving major reach-wide bed incision with a median rate of −0.26 m a−1 over the period 2013 to 2019. The use of satellite imagery to monitor sand mining activities provides a low-cost means to generate up-to-date, robust estimates of sand extraction in the world's large rivers that are needed to underpin sustainable management plans of the global sand commons.