A non-transitory computer readable medium, storing a set of instructions, executable by a computer, to perform a method for predicting the amount of corrosion in a steel used in an electronic computer, using, as an index for evaluating the amount of corrosion when the steel is exposed to the outdoors, a corrosion index Q(tXi, WX) of the steel obtained from a weather observation value WX including a wind speed, a wind direction and an amount of rain for each observation time tXi (where i is an integer of 1 to n) at an observation point X, the method comprising:
- an extremal value estimation step in which, based on the weather observation value WM at which the observation point X is a scheduled use point M of the steel, a corrosion index Q(tMi, WM) of the steel is obtained for each past observation time tMi, extremal value statistical analysis is performed on time-series data of the corrosion index Q(tMi, WM) between observation times tM1 to tMn, and thus a maximal value Qmax(tFi) of the corrosion index within future evaluation periods tF1 to tFn is estimated;
- a corrosion amount estimation step in which a relational expression between a measured corrosion amount obtained in a corrosion test for the steel at a test point Z and the maximum value of a corrosion index Q(tZi, WZ) of the steel obtained for each observation time tZi during a test period of the corrosion test based on a weather observation value WZ at which the observation point X is the test point Z is obtained in advance, the maximal value Qmax(tFi) of the corrosion index estimated in the extremal value estimation step is introduced into the relational expression, and an estimated value of the amount of corrosion in the steel during the evaluation periods is obtained;
- and
- a selecting step of selecting a recommended steel based on the estimated value of the amount of corrosion that was obtained.