Prediction of the Severity of Atmospheric Corrosion by Discriminant Analysis of Local Meteorological Factors

Abstract

Discriminant function analysis has been used to enable the corrosiveness of an environment to be classified into one of several broad categories, knowing the values of specified atmospheric factors. Of various combinations of meteorological factors that were tried, the following combination proved to be the most discriminating: average of daily maximum temperature in °c, average daily maximum relative humidity, number of rainy days in a month, total rainfall in mm in a month, and average pollution (SO₂ + NaCl in mg/cm² per day over a month). These criteria were used to classify stations at Delhi, Bombay and Tezpur, giving results that agreed well with those obtainedin exposure trials.