Courtesy: UMass News & Media Relations
Contact: Timothy Randhir 413/545-3969

AMHERST, Mass. – Surprise and uncertainty are becoming more common in water resource planning, says University of Massachusetts Amherst water resource researcher Timothy Randhir, and there is a need for new approaches. Now he and colleagues propose dusting off a “surprise potential framework” developed decades ago that they say is nevertheless well suited to decision-making when few facts are available.

“A recent example is the situation where Cape Town, South Africa, is running out of water unexpectedly quickly,” first author Randhir points out. “Another example is Hurricane Katrina, a situation where we had no prior event on that scale to help us plan for it.”

As he explains, the “Shackle, Vickers and Katzner” (SVK) framework was developed in the 1970s and 1980s and is “a totally different approach than is used in most situations today, but it can be very useful in situations where you don’t have a lot of data. This will be more important as we face more and more unexpected events, more and more frequently.”

Randhir defined it and created measurable terms for recording observations and optimizing decisions, while developing the framework to address what the authors call uncertain, “severe, naturally occurring arsenic contamination” in water supplies in Bangladesh. Details appear in an early online issue of the Journal of Hydrology.

Randhir and colleagues write that exceptionally high natural arsenic levels are found in groundwater in Bangladesh. Before the 1970s, people relied mainly on surface water for drinking and daily use, but since then it has become increasingly polluted and a public health threat.

In response, the researchers note, “The government and donor agencies suggested the cost-effective solution of digging shallow tube-wells to provide access to safe water,” and groundwater pumping was able to meet demand so that by 2000, almost 97 percent of people in rural Bangladesh were drinking water from 2.5 to 3 million shallow-tube wells. But in the 1990s, they learned that most well water contained arsenic.

Most of the highly contaminated wells are in shallow aquifers 50 to 150 feet below the surface, while deep aquifers are nearly arsenic-free, the authors point out. However, it is uncertain whether deep aquifers will remain arsenic-free over time, they add.

The UMass Amherst researchers used the mathematical framework with variables related to arsenic contamination in shallow vs. deep wells to calculate the probability of various decision outcomes for water resource managers as an analytical framework for future decision-making and mitigation measures for similar situations of uncertainty.

Overall, they say that “a diversification of the water supply system” emerged as “a robust strategy to avert unintended outcomes of water contamination. Through optimization, they observe that shallow wells are slightly preferable than deep wells and surface water treatment, considering the nature of uncertainty and cost involved.

They suggest that the SVK method can be applied “in a variety of other cases that involve decision making under uncertainty and surprise, a frequent situation in natural resources management,” and it can be “a very useful tool for utility managers who can incorporate subjective beliefs and expert opinions into the decision-making process to develop a robust water supply system.”

Randhir adds, “We primarily recommend that you don’t put all your eggs in one basket, but diversify your approach to water supplies, treatment methods and so on, to keep the drinking water supply as safe as possible.”

This work was supported by a Healey Endowment Research Grant and U.S. Department of Energy Interdisciplinary Research Fellowship at UMass Amherst’s Environmental Institute and by a summer research grant at the Florida International University, Miami.