WATER ABOVE THE MOUNTAIN FRONT – ASSESSING MOUNTAIN-BLOCK RECHARGE IN SEMIARID REGIONS By

Mountains provide as much as 90-100% of the freshwater to surrounding basins in arid and semiarid regions because of their distinctive and complex topography, and the consequent effects on precipitation (P) and evapotranspiration (ET). One of the primary objectives of this dissertation is to estimate mountain-block recharge (MBR), an important component of the mountain contribution to groundwater replenishment of surrounding basins, as well as its response to regional climate variability in a semiarid mountain environment of the southwestern U.S. Two major limitations, a lack in understanding hydrologic processes and sparse observation networks, hinder predictive mountain-block hydrologic modeling, and reliable estimation of MBR. Developing approaches to address these limitations is another objective of this dissertation. A geostatistical algorithm (Auto-Searched Orographic and Atmospheric effects De-trended Kriging, ASOADeK) is first developed for mapping mountain precipitation using sparse gauge data. ASOADeK constructs monthly precipitation maps comparable to PRISM products, and with higher spatial resolution. ASOADeK is also useful for studying regional climatic settings. In arid and semiarid regions, the dominant water flux out of the mountain block is ET. A Topographyand Vegetation-based surface energy partitioning model for ET modeling (TVET) is developed to include the effects of vegetation and topography on mountain-hillslope energy partitioning. The TVET model can be used to map daily potential evaporation and potential transpiration over mountain terrains. The model is also useful for ecohydrologic studies, for example, evaporation and transpiration partitioning of sparsely vegetated ecosystems. To understand the factors influencing distributed MBR, both generic and specific two-dimensional hydrologic simulations at the hillslope scale were conducted, using the variably saturated hydrologic modeling code, HYDRUS-2D. The results show that the controlling factors for distributed MBR include bedrock permeability, atmospheric forcing (precipitation, and potential evapotranspiration, or PET), vegetation coverage, and soil cover. Among these, bedrock characteristics are the primary control, affecting both the amount and patterns of mountain-block recharge. For bedrock with permeability above a certain threshold (10 ~ 10 m, equivalent to saturated hydraulic conductivity of 10 ~ 10 m/sec), local climate conditions (regional climate setting + local orographic modification + elevation-and-slope-aspect effects on P and PET), which determine the water availability at the soil-bedrock interface, are the most important controlling factors. Vegetation strongly affects distributed mountain-block recharge by modifying surface energy balance and soil hydraulic properties. Root-zone soil thickness has a significant influence, especially for the matrix-flow-dominant bedrock (e.g., non-welded tuff). A change of vegetation cover in mountains can lead to a significant change in basin-scale groundwater balance. These results provide criteria for classifying hydrologically similar response units (HRU) in mountain blocks. A framework for the HRU-based approach for quantifying mountain-block recharge is provided. This framework and related sensitivity studies suggest that future efforts should focus on better characterization of mountain bedrock hydraulic properties and better quantification of high-resolution (both temporally and spatially) mountain precipitation estimates. A simple point-simulation-based approach is applied to map potential mountainblock recharge in two mountain ranges, northern New Mexico. Assuming uniform bedrock, soil cover and vegetation coverage, the long-term mean downward water flux across the soil-bedrock interface (upper-bound estimate of distributed MBR) can be statistically associated with long-term mean local climate forcing (i.e., mean PET and P). Similarly, the actual ET flux can be related to mean local climate conditions. With these correlations derived from simulations with recharge-optimal bedrock and soil conditions, maps of upper-bound distributed MBR and water yield (or upper-bound of total MBR, i.e., the difference between precipitation and actual ET) are constructed for two mountain ranges with distinctive bedrocks, the southern part of the Sangre de Cristo Mountains and the Jemez Mountains, both in northern New Mexico. The results show that distributed MBR is restricted to the higher elevations in the Sangre de Cristo Mountains, while it is more widely distributed in the Jemez Mountains. The area-weighted average upperbound distributed MBR is about 35% of the water yield in the Sangre de Cristo Mountains, and 50% in the Jemez Mountain. The results also suggest that previous total MBR estimates (70 mm/yr) for the Sangre de Cristo Mountains are reasonable if the bulk bedrock permeability is close to 1×10 m. For the Jemez Mountains, the results give a total MBR between 70~120 mm/yr, about one half of previous estimates for the San Juan Mountains. To understand the response of MBR to climate variability, the teleconnections of seasonal precipitation in the mountains in northern New Mexico with PDO (Pacific Decadal Oscillation) and ENSO (El Niño-Southern Oscillation) are investigated. The results suggest a strong correlation between winter and spring precipitation and ENSO and PDO cycles. The summer precipitation, which is dominated by the North American Monsoon, does not have clear correlation with ENSO and PDO cycles. For winter and spring precipitation, PDO effects are more dominant than ENSO effects. Low PDO effects are strongly dampened by El Niño, and slightly enhanced by La Niña. ENSO modulation of high PDO effects is not as strong as for low PDO effects. The high PDO effect on winter precipitation is enhanced by El Niño, but not much affected by La Niña. PDO and ENSO effects on winter precipitation are modified by topography, with larger anomalies at higher elevations for wetter winters, and larger anomalies at lower elevations for drier winters. The effects of PDO and ENSO effects on distributed MBR are examined by the recharge-climate index functions, which are derived from generic hydrologic simulations. The results suggest that ENSO and PDO associated climate variability can typically lead to a 10~20% change in distributed MBR for the two tested mountainous ranges. Because of its multi-decadal period, PDO effects on MBR may influence groundwater resources in surrounding basins.