For a variety of reasons, meteorological station records are generally not complete. Missing data can affect how useful a station dataset is for calculating, for example, monthly averages and climate normals (30-year averages). There are standards related to the amount of missing data allowed in calculations, but rigid adherence to these can sometimes result in a lot of data being discounted, and this can be a problem, particularly in areas where data are sparse.
We provide four missing data options for the analysis of observed station data. These permit you to control how much missing data is acceptable in your custom analysis:
- 5%
- 10%
- 15%
- WMO parameters
For options 1-3, the percentage of missing data allowed is calculated for each period of the station record (month, year or season, depending on your selection). This means that for the annual calculations the missing values could be spread throughout the year, or occur all at once. For example, if you select the 5% missing data option, this means that as many as 18 missing values would be allowed in a year – they could be spread throughout the year or there could be a whole month missing.
For option 4, the WMO (World Meteorological Organization) parameters, the criteria are applied for each month regardless of whether monthly, seasonal or annual calculations are selected, which makes the calculations for this option slightly slower than those for options 1-3. A given month is considered missing or invalid if either of the following two criteria are met:
- Observations are missing for 11 or more days during the month;
- Observations are missing for a period of 5 or more consecutive days during the month
Presence of a single invalid month within an annual or seasonal calculation will result in a no-data (NaN) value in the output