Soil coring at a grass vegetation site. Soil cores are being used to investigate how vegetation type relates to recharge through the desert floor.

F. Phillips, M. Walvoord (NMT)

The groundwater recharge rate is a critical component of the water balance from the viewpoint of sustainability of water resources. Recharge rates in arid regions are known to be much higher at high elevations than at low elevations, but the magnitude of recharge rates at low elevation has long been a matter of controversy. Even very low recharge over desert floor areas could be quite significant for water use. For example, an average recharge of 1 mm yr-1 over the desert floor area of the Rio Grande basin could supply the domestic needs of a population twice the current one of New Mexico. This project aims to understand vegetation control on the hydrodynamics of desert vadose zones and ultimately groundwater recharge rates.

Activities and Results

Activities during the current year centered primarily on finalizing simulations and writing up results for publication. Considerable effort was invested in modifying the FEHM code to properly simulate the transport of stable isotope tracers. There was also a major effort to perform extensive sensitivity analyses. Michelle Walvoord wrote and defended her Ph.D. dissertation, with much of this work in press or in revision.

We have relied on vadose-zone profiles of chloride concentration and water potential as our main lines of evidence, and have interpreted these using the multi-phase (liquid and vapor), variable temperature model FEHM. Our results have shown a remarkably uniform pattern of water movement in vadose zones of desert floor areas. These are characterized by upward water potentials in the upper ~50 m and inventories of chloride equivalent to >10 kyr of atmospheric deposition. The pervasive occurrence of such profiles in desert lowlands implies that these areas have not been sources of diffuse groundwater recharge since the end of the last glacial period, 10 to 15 kyr ago. Based on intercomparison of vadose-zone profiles under differing vegetation communities we have concluded that the main control on vadose-zone hydrodynamics (and thus recharge) is type of vegetation. Desert shrub (and to a lesser extent desert grassland) creates root-zone water potentials so low that no water can move below the root zone and, in fact, water is extracted from the deep vadose zone.

Figure 2. Chrloride profiles beneath different vegetation types and the implication for groundwater recharge..

Our results have three major implications. The first is that where we observe characteristic vadose-zone profiles consisting of very negative water potentials (~<-4 MPa), combined with large soil chloride inventories, we can infer that the vadose zone is locked into a long-term drying transient that precludes any diffuse groundwater recharge. This inference is of significance for both water-balance studies in the SAHRA project and for water resources investigations in arid regions in general. A second implication is that when these conditions are met, total hydraulic gradients in the top ~50 m of the vadose zone are upward, and thus contaminants cannot be transported downward to the water table. This implication is of considerable significance for nuclear waste disposal programs, and other types of waste disposal as well. Finally, preliminary data from our Trans-Pecos investigations suggest that the key controlling factor on vadose-zone hydrodynamics is vegetation, and that areas in which hydraulic gradients are at least episodically downward can be identified by mapping vegetation type. If confirmed, this finding has important implications for both quantifying the distribution and amount of groundwater recharge, and for ecohydrological controls on the positions of ecotones.

Plans

Our investigations so far have resulted in two major findings: 1) a conceptual model that indicates that during the last 10,000 years desert floor environments have been areas of upward hydraulic gradients, dominated by vapor transport, and thus are not sites of active recharge; and 2) water fluxes below the root zone are mainly controlled by vegetation, and measurements across certain ecotones indicate that recharge is happening beneath vegetation communities other than desert scrub and desert grassland. The critical aspect of this second finding is that it implies a vegetational, rather than direct climatic control on recharge. The first finding is certainly relevant for assessment of water balances in arid drainage basins, but further exploration has implications for paleohydrology and waste disposal issues. In contrast, the second finding can be pursued to develop a tool that will permit soil/groundwater balances to be estimated on the basin scale, based largely on remote sensing data.

At present, the ecohydrological linkage between vegetation and fluxes below the root zone is based on a very limited data set. Our first priority will be to test and quantify the ecohydrological hypothesis by the collection of more data. We will drill shallow soil augerings and measure matric potential and chloride concentration along climatic transects that cross ecotones. These will avoid areas of steep topography, so climatic gradients will be gradual. In contrast, ecotones are abrupt. Abrupt changes in the pattern of matric potential and chloride across the ecotones will support the ecohydrological control hypothesis, whereas gradual transitions will support hydroclimatic control. In either case, the results can be directly applied to quantification of water balance and soil-water partitioning that are on-going projects within TA2. Ultimately, our results should enable the basin-scale model that will integrate SAHRA's efforts to be parameterized for the interconnection of shallow and deep vadose zone processes.