Jorge F. Talavera

The inventory of shallow groundwater plus surface water is more a less a constant number at a certain location, time and pluvial station. It may be called the "geohydrological inventory" and it is nearly constant - even with the existence of multiple mechanisms of water exchange between the ground and surface. A material balance will show that the net amount received from precipitation will be moving around the two: ground and surface, but remain constant. Mechanisms such as precipitation, evaporation, evapo-transpiration, soil characteristics relative to infiltration and hydraulic conductivity, geomorphology, hydraulic gradients and vegetation, among others, determine the rates of recharge and discharge of the aquifer. Aquifer net inventory gains, for example, will cause higher hydraulic levels (or differentials) which in turn will tend to augment the natural discharge from underground to the surface at another point within the aquifer, yielding therefore, a final equilibrium state, albeit different from the original one in terms of heights and flows. All these natural disturbances, which are aleatory in nature, may be observed during a period of time, and be correlated to external factors such as precipitation, local temperature, and state of the local ecosystems such as vegetation. Most likely it will yield a cyclical pattern that under idealized cases of pristine systems -- and the absence of "macro trends" such as global warming -- will not necessarily show a particular trend or tendency.

External sources and sinks to this geohydrological inventory are clearly the atmosphere and the ocean. The atmosphere is the main source by giving up precipitation and it is the sink for evaporation or evapotranspiration. The ocean will be the main sink as it is the continuous receptor of water from the continents due to its lowest altitude. Groundwater eventually joins surface water discharging to the ocean except for a small amount -- equivalent to 2% of the precipitation -- which will percolate very deep into the land and be highly immobilized. This water will ultimately also reach the ocean but will take a period of time several orders of magnitude higher than what the shallow water takes, and therefore may be considered external to the geohydrological inventory which is fairly dynamic in nature.

Groundwater is tapped extensively throughout the world for agriculture use and human consumption. Similar to natural discharge, human intervention displaces the equilibrium of the surface-ground exchange to a different point such as an extra discharge would. The difference with natural disturbances is that the human intervention is not aleatory but instead a permanent and pervasive extraction of groundwater and will therefore cause to build a time trend into the system. Net extraction for domestic use will for example diminish the hydraulic level and therefore, baseflow, which variables in turn will influence vegetation and most sensitively, riparian vegetation. Human net extraction and use therefore, will mean a net loss of geohydrological inventory brought about by precipitation and diligently kept stored by nature to deal with all these natural aleatory disturbances. Thus knowing that human extractions will have predictable effects, the evaluation of the effects turns to be only a matter of quantification related basically to the magnitude of the extraction and the capacity of the system being extracted upon to absorb the disturbances and still regain a new equilibrium state, albeit a poorer one, or yet to instead deteriorate indefinitely affecting severely other uses of water and the ecosystem. It is imperative to know therefore the technical and social criteria to elucidate what can be considered a safe or sustainable yield.

On one extreme the most classic and utilitarian meaning of "safe yield" was a withdrawal rate that will equate statistically with the recharge amount of groundwater. This criteria has been discredited by facts and experience because it treats groundwater as an static entity instead of what it really is: a moving part of the total geohydrological inventory that permanently discharges naturally to the surface, as well as tries to maintain optimum levels that will sustain vegetation.

On the other extreme the most restrictive and environmentally friendly meaning of "safe or sustainable yield" is a zero net withdrawal rate or such a minimal rate that will equate to the deep percolation rate and therefore equate to a magnitude that is already lost to the geohydrological inventory and does not affect it.

Between the two extremes, there are several criteria that range between 10% and 40% of the average recharge as plausible rates of yield or utilization. These might be better called "reasonable compromise yields" than safe yields. While there is no general answer that will satisfy all cases, the approximation to the optimal solution on each particular case will depend on the degree of detail of study of the system and a benefit-cost tradeoff of social expectations and environmental consequences. Specifically, groundwater systems are encased in fairly complex geological structures and observations quite frequently will require decades to establish a clear trend and or to make clear and patent the detailed consequences to the environment.

Finally, if a general rule need be proposed, it may be suggested that the initial rate of net utilization should not be greater than 20% of the recharge value which fraction is about average of many successfully managed systems reported in the literature. This usage rate should be evaluated subsequently for its impact on the base flow and the whole watershed ecosystem. Fine tuning to an optimal value will require several years of feedback information.