ABSTRACT

As an assessment method in the protection of the environment, a stochastic approach is used to analyze the monthly precipitation data in Alabama. Monthly mean precipitation data from 12 stations in discrete equidistant form were analyzed. The computed precipitation trends for the decade 1986-1995 showed 10 stations with positive trends ranging from +0.62 mm to +4.06 mm per year, while the other two stations showed negative trends. For the three decades studied only two stations, Mobile and Huntsville, show consistently positive trends, predicting more rainwater to recharge the local aquifers and more incidences of floods in these areas. For the periodicity, the power spectrum showed that the periodic components tend to be insignificant in Alabama, with yearly oscillations contributing less than 30% to the periodicity. The auto-correlogram showed that the precipitation of each month does not depend on that of the previous month, indicating a variable monthly precipitation behavior in Alabama.