Clear-sky irradiance model

A clear-sky model calculates the irradiance a site would receive with no clouds, using latitude, longitude, elevation and date and time, and assuming normal atmospheric conditions. Clear-sky PV power output is the output calculated without clouds or other weather, so it represents the maximum output possible from an installation; the key input is the angle between the sun's rays and the panel normal. Because the model does not depend on measurements and repeats each year for a given location and time, it serves as a baseline for multi-year comparisons, such as spotting irradiance sensor degradation or validating GHI readings. In performance loss rate work, clear-sky filters keep only periods with little cloud by comparing sensor irradiance with modeled clear-sky irradiance, either with a strict multi-criteria method or a simpler clear sky index ratio. Monitoring platforms also use clear-sky direct and diffuse estimates to transpose a pyranometer reading to plane-of-array irradiance and to estimate shading loss. A statistical clear-sky fitting method can build an adaptive clear-sky model from measured power alone and use it to detect clear conditions, fill gaps and estimate degradation. Clear-sky models work less well in cloudy periods, so gap-filling and filtering methods that rely on them perform best when the data actually come from clear-sky periods.

<!-- review notes

• sentence 1: support check: partial - Passages give inputs and normal atmospheric conditions, but 'no clouds' is only implied by the name 'clear-sky', not stated for the model.
• sentence 7: review: partial - The passages say modeled irradiance does not perform well in cloudy periods and that one gap-filling method (DSG1) works best on clear-sky data, but extending this to filtering methods in general is not stated. -->