A quasi three-dimensional model of groundwater flow has been developed for the Mesilla Bolson and Mesilla Valley stream-aquifer system in southcentral New Mexico. The quasi three-dimensional model solves the two-dimensional horizontal flow equations in the Santa Fe Group and flood-plain allivium aquifers, while simulating the steady state vertical leakage across the aquitards that separate the two aquifers. The model can account for groundwater-surface water interaction through stream infiltration, canal losses, drain discharge, and evapotranspiration.

A calibrated version of the quasi three-dimensional model was developed through steady state and transient analysis of the Mesilla Bolson and Mesilla Valley stream-aquifer system. In addition to quantitative estimates of the mountain front recharge, improved estimates of aquifer properties and confining bed characteristics were obtained through steady state calibration. Transient simulations during the 18-year period of 1966 to 1983 showed that groundwater-surface water exchange processes in the Mesilla Valley dominate the current head distributions in both aquifers from year to year. The mean annual volume of applied irrigation water during the period 1966-1983 was estimated as 240,300 acre-feet, while the average annual seepage loss from surface waterways for the same period was simulated as 116,200 acre-feet.

Predictive runs with the quasi three-dimensional model suggest that with continued and increased pumping for 100 years in the vicinity of Las Cruces only, the piezometric head levels in the Santa Fe Group within the cone of depression may be as much as 60 feet lower than existing levels. With the proposed El Paso wells on the West Mesa, piezometric head levels in the vicinity of the proposed well field may be as much as 200 to 400 feet lower than existing levels after 100 years of pumping, depending on recharge conditions in the Mesilla Valley and the behavior of storativity in the Santa Fe Group. The effects of proposed El Paso pumping on the West Mesa will be propagated to the Mesilla Valley; increased pumping will reduce drain flows and increase downward moving leakage.

Several limitations of the quasi three-dimensional model have been identified during the course of this modeling investigation. Consequently, the results of the predictive simulations should be used with caution. It is recommended that predictive simulation results be used primarily as qualitative (rather than as quantitative) indexes of the Mesilla Bolson’s response to future stresses on the groundwater system.