Explanation of fitted gridded data

Gridded data sets are a level-4 ISR data product for local ISR measurements. They are derived from basic derived parameter data and currently they are available primarily for Millstone Hill. Gridded data combine all high-elevation pointing directions and all experiment waveforms into a single data product with a fixed time sequence (typically 15 minutes) and a fixed set of altitudes (with increasing altitude spacing at higher altitudes). Gridded experiment files contain both basic parameters (electron density, electron temperature, ion temperature) and derived parameters (for example, neutral temperature and, in the future, vector velocity and neutral winds).

Each gridded dataset corresponds to a single UT day. In some cases the basic derived data from which the gridded data is derived may come from the basic derived parameters in more than one experiment. Each gridded dataset is treated as a separate Madrigal experiment, which includes both the gridded Madrigal file and a number of summary plots. The gridded basic parameters are computed by fitting a least-squares bicubic spline to all basic derived parameter data beginning three hours before and ending three hours after the target day. The additional data at the beginning and end of the day reduces end effects in the fit and ensures near continuity of two successive gridded data sets.

The gridded basic parameters are a convenient starting point for computing additional derived parameters. For example, neutral temperature and atomic oxygen can be estimated from a simplified version of the energy equation for the ions. Integrated electron content, foF2 and hmax are determined from the gridded electron density. Vector velocities and the electric field can be computed if suitable look directions are available in the basic derived parameter data and these will be added to the gridded parameter file in the near future. Neutral winds will also be added in the near future.

Plots of the gridded data are also available for gridded file data. They include color contour plots of the data, color plots of the data, fit and data-fit, altitude profiles at selected times and time series plots at selected altitudes.

The gridded datasets should be used with a certain amount of care. The spline fits smooth the data, which reduces statistical errors but can also hide real variations in the data. The plots that compare the data to the fit are helpful in assessing the quality of the fits. Also note that while random errors in the data are reduced they are not entirely eliminated. However unlike in the basic derived parameter data nearby gridded data points are correlated. As a result any statistical errors show up as smooth wiggles in the data and care should be taken not to assign undo significance to these. Again, the plots can be useful in determining whether small variations are real.