While the science of weather is an evolving and challenging subject, the effective application of weather data is equally challenging and often where the true value of weather information and decision aids provides significant benefit.

As with any solution, once deployed and effective, there is a desire to maintain and even improve operational savings. As the weather data industry has become more commercialized, one result is a competitive and option-rich environment from which prospective weather data users can select. A key factor in selecting and using vendor data is ensuring operational advantages can be realized.

This paper describes a proof-of-concept project we have recently completed. In this example we have applied data mining techniques to improve operational performance of an industrial system that uses multi-vendor frequent weather data for current and next day decisions. The results from this initial analysis are encouraging. We have found areas where marked improvements appear possible as well as interesting weather vendor specific trends and nuances that can be avoided to use to the customer's advantage.