Selection of test variable for minimal time detection of basin response to natural or induced changes

Abstract

In an age of increasing interference of man in nature it is important to detect rapidly the effects of purposeful management in a given sector as well as the inadvertent effects in a different sector. Due to the variability of nature and man's incomplete knowledge of its state it is difficult to assess whether an apparent change is the result of nature's caprice or of man's deliberate action. To answer this delicate question classical statistical tests have been utilized by the scientists and technicians. Often the physical variables have been reduced to mere statistical variables. While the statistical information about the variable has been somewhat utilized the physical information about it has often been neglected. As a result the tests used were far from being optimal. Detection tests which utilize both the physical and statistical information can be designed and they will have a superior power to the standard tests. The test variable to be used is a linear combination of several physical and random variables with weight factors determined by a minimization procedure. The minimization is restricted by several equality constraints of a physical origin. The test has proven to be more powerful than the classical (unconstrained) simpler tests. When applied to the Colorado River Basin Pilot Project area the power of the test, expressed in years needed for detection, is increased by a factor of two.