In many simulations of energy systems, climate is represented as Typical Meteorological Year (TMY) data. The approach derives hourly meteorological variables from historic records to represent annual climate for a given location. While meeting the need for comparability and indicativeness, the technique suffers from an incapacity to provide information on extreme weather events, like heat waves or passing tropical cyclones, or for specific actual periods of concern.

This paper describes a modified approach whereby representative data may be selected for a targeted purpose. The data is converted to a format appropriate to the simulation under consideration, checked for errors, data omissions “filled” and applied to the model.

Applications may include:
• Model calibration using real time weather data coincident with other empirical measures like solar system output or building energy consumption or temperature (especially if unconditioned).
• Building or system monitoring for underperformance to indicate early restorative action.
• Adjustment of actual output or consumption in a real year to reflect reasonably anticipated outcomes in the actual year.
• Design assessment against challenging weather conditions like a hot dry (El Nino) year or a windy wet (La Nina) year or some shorter duration subset of such periods.
• Forecast future weather in anticipation of climate change (global warming).