The use of tree-ring
data, as opposed to recorded water levels, was the basis for determining
(1) long-term probabilities of exceeding critical hydroclimatic
thresholds, and (2) worst-case scenarios – what is possible
in terms of the severity of drought.