What is the IPGP method for creating Quasi-Definitive Data ?
Quasi-definitive data are calculated every month, shortly after the end of the month. The processing is very similar to that for producing definitive data, except that the time interval for the baseline calculation is shorter (a few months instead of one year or more) and the data preprocessing (despiking, etc.) is less thorough.
The IPGP method can be applied even if baselines have irregular variations provided they meet the INTERMAGNET standards (Peltier and Chulliat, 2010).
The processing is made of the following three steps, repeated after each month M:
- Variation data of month M are pre-processed in order to remove errors and spikes, correct for jumps (if any) and replace gaps by data from a back-up magnetometer (if available). This step is usually completed within a few working days after the end of month M, although for some observatories it can take longer.
- A temporary baseline is calculated from 1st December of the previous year to the last day of month M, for each component of the field. Absolute data are visually inspected and outliers are removed prior to the baseline calculation. If needed, absolute data prior to 1st December of the previous year are included in order to extend the time interval of the temporary baseline. (This happens if, for example, M is at the beginning of the year and/or too few absolute data are available at the beginning of the interval.) IPGP baselines are calculated using cubic smoothing splines with a constant smoothing parameter (csaps function of the Matlab software, which relies on an algorithm described in de Boor, 1978).
- Baseline-corrected 1s and 1min data are calculated from 1st January of the current year to the last day of month M. A final quality control is performed at this step, by visual inspection of the time series of components and scalar differences. Once validated, the baseline-corrected data for month M are reformatted within the IAGA quasi-definitive data format and delivered to INTERMAGNET.
References:
- Peltier, A. and A. Chulliat, On the feasibility of promptly producing quasi-definitive magnetic observatory data, Earth Planets Space, 62(2), e5-e8, doi:10.5047/eps.2010.02.002, 2010.
- De Boor, C., A Practical Guide to Splines, Springer-Verlag, New York, 1978.