Projecting the Potential Effects of Climate Change on a Karstic Aquifer System: Challenges for a Regional Groundwater Management Agency – Invited

The Edwards Aquifer Authority (EAA) is a regional groundwater conservation district in south-central Texas, USA, and is charged with the management of groundwater withdrawals from the San Antonio segment of the Edwards (Balcones Fault Zone) Aquifer (Edwards BFZA). The Edwards BFZA is a prolific limestone karst aquifer system, and its major springs are artesian. The aquifer is the primary source of water for more than two million people in the region and provides springflow to several federally protected species. Ensuring adequate springflow requires maintenance of aquifer system pressure, which is a balance of pumping withdrawals, spring discharges, and addition of recharge at the higher elevations of the recharge zone, where the Edwards Formation is exposed at the surface. Recharge is dominated (~75% or more) by losses of water into discrete karst features within stream beds that cross the recharge zone. The Balcones Fault Zone facilitates the downward movement of water into the subsurface and contributes to the area’s complex topography, which influences precipitation patterns. Further, the Edwards BFZA region has a significant precipitation gradient going from dry to wet from west to east.

The aquifer is managed using a regulatory framework that includes a cap on permitted withdrawals, reductions in permitted withdrawals during drought, pumping forbearance lease programs, and supplemental supplies from an aquifer storage and recovery system that are implemented during significant droughts. These measures provide protection of spring flows specified as part of the Edwards Aquifer Habitat Conservation Plan (EAHCP) and its associated Incidental Take Permit (ITP). The current EAHCP ITP expires in 2028, and its renewal requires assessment of the potential impacts of climate change on spring flows.

HCP-related guidance on methodologies to address climate change is non-specific, so the specific approaches are left to the ITP applicant. Based on the initial ITP process, EAA anticipated a need to provide quantitative estimates of projected spring flows under future climate conditions over the proposed 30-year renewal term. The EAA desired region-oriented downscaling to (i) honor the local geographic effects on the Edward BFZA climate, and (ii) provide ability to effectively differentiate meteorological conditions between stream basins to which recharge is sensitive. Moreover, EAA needed assistance to (i) select among generations of global climate models (i.e., CMIP5 and CMIP6), (ii) adequately cover the range of projected conditions while minimizing the obligation of resources needed to process model results, and (iii) ensure a technically defensible set of climate projections for the region. To address these issues, the EAA partnered with researchers at the South Central Climate Adaptation Science Center who developed a method for climate model selection and produced region-oriented, downscaled global climate model projections.

The region-oriented downscaled climate projections were used as input for estimating future recharge. Current estimates of aquifer recharge utilize stream gauging to measure stream losses and do not directly incorporate a climate or meteorological component. To address this, a new machine learning-assisted recharge model was developed by training a tree-based ensemble model to reproduce historical recharge values using input variables including precipitation and temperature. The resulting model performs well relative to a validation set of historical values and was used with climate model output to produce projected recharge sequences. These recharge sequences were then used in an existing numerical groundwater flow model to produce future projected spring flows and aquifer water levels under various climate scenarios. Results indicate generally lower recharge is expected over the proposed permit period, but no droughts more severe than the current drought of record were identified. The results also suggest the current aquifer management framework may be suitable for use over the next three decades as part of a renewed EAHCP ITP.