Should We Be Concerned About the Fraction of Groundwater Pumping Potentially Derived from Streamflow Capture in Brazil?

The well-known statement that “all water discharged by wells is balanced by a loss of water somewhere” introduced by Theis (1940) reminds us that groundwater pumping is supplied partly by aquifer storage and partly by capture — water that would otherwise support streamflow, often dominated by streamflow capture. Despite this long-standing conceptual understanding, groundwater and surface water are still managed separately in most countries. Brazil is no exception. Ongoing climate change impacts on surface water availability, combined with rapid expansion of irrigated agriculture, have increased reliance on groundwater across the country. At the national level, groundwater irrigation permits account for approximately 20% of total permitted irrigation volumes, while surface water permits account for the remaining 80%, based on water permits issued by the Brazilian National Water and Sanitation Agency in 2020. When combined with surface-water withdrawals streamflow capture may cause river flows to fall below environmental flow thresholds, posing risks to ecosystems and water users. This raises a central question: should we be concerned about the fraction of groundwater pumping potentially derived from streamflow capture in Brazil? Here, we apply analytical streamflow depletion functions that integrate local ground-based data with national-scale water assessments, balancing model simplicity and accuracy while relying exclusively on publicly available datasets. Groundwater pumping data are obtained from the Groundwater Wells Database for Brazil (GWDBrazil), compiled from the Geological Survey of Brazil, including 169,091 legally registered wells with depths shallower than 100 m likely focus on unconfined aquifers. Long-term annual minimum flows (1971–2000) for ~ 380,000 rivers of 2nd order or higher are derived from HydroATLAS, while hydrogeological parameters are taken from the Global Hydrogeology MaPS (GLHYMPS) dataset. We estimate the fraction of potential streamflow capture from groundwater pumping as a proportion of long-term annual minimum discharge for the period 1970–2025. Our results show that, although only about 3% of the analyzed rivers experience more than 10% of potential streamflow capture, a threshold commonly used in the absence of river-specific environmental flow studies — the capture fraction represents ~ 48% of the groundwater abstracted volume in likely unconfined aquifers analyzed in 2025, i.e., nearly half of groundwater withdrawals are supplied by streamflow capture. In the São Francisco Basin — critical for water-food-energy security in Brazil and globally — approximately 10% of rivers analyzed (1,610 of 15,420) already have more than 10% of their long-term annual minimum flows committed by streamflow capture alone. When surface-water withdrawals — still the dominant source of water use in Brazil — are included, this fraction increases to approximately 42% (6,442 of 15,420 rivers). Nationally, streamflow capture may not represent a widespread concern when considering the legally registered wells, but distinct hotspots of hydrological stress clearly emerge, particularly when surface-water withdrawals are accounted for. These findings highlight the importance of integrated groundwater–surface water management and demonstrate how an accessible framework can support science-based water policy by jointly accounting for surface-water withdrawals and streamflow capture when granting water permits, helping to avoid water overallocation, reduce environmental impacts, and ensure long-term water and food security.