Skip to content

Data Gaps

It is strongly advised to clean and review your weather data before importing it into PVsyst. If your input data contains missing data or gaps, PVsyst will try to fill the missing values when possible.

Missing sub-hourly values

When importing a sub-hourly file, gaps of less than 1 hour will be interpolated linearly from the 2 closest values. Gaps of more than 1 hour will not be filled.

Illustration for a 20 min time step file:

Diagram showing linear interpolation of sub-hourly gaps shorter than 1 hour, with larger gaps left unfilled

Missing hourly values (1-4 hours)

In case of full hour(s) of data missing (both for hourly and sub-hourly data file imports) the gap filling scheme will depend on the variable considered:

  • GlobHor or GIPMeas (POA) missing: These gaps are not filled during the day as correct irradiance is critical to PVsyst simulation. Missing night values are simply set to 0. Missing day values will be flagged in yellow in the data tables. It is not advised to run simulation with incomplete irradiance data.
  • DiffHor or DNIMeas missing: The diffuse irradiance will be computed from the GlobHor value (if available) using the appropriate diffuse model. For all other variables (temperature, wind, etc), hourly values will be filled by linear interpolation. If the input is a subhourly file, the corresponding subhourly values will be filled by linear interpolation as well.

Illustration for a 20 min time step file:

Diagram showing the hourly gap filling scheme for 1 to 4 hour gaps, with linear interpolation applied to non-irradiance variables

Large data gaps

If more than 4 hours of data is missing:

  • Synthetic temperature will be generated using the site temperature values and GlobHor values. Hourly values are generated and in the subhourly import case, subhourly values are interpolated linearly from hourly values.
  • Monthly values from the site will be used for wind speed, albedo, relative humidity and Linke coefficient. Hourly and subhourly values are identical.
  • Gaps in variables not present in site data (e.g. module temperature) will not be replaced.