Data Processing: Interpolation in Time

The pressure and temperature data were interpolated linearly to the synoptic times 0000 GMT, 0300 GMT, ..., 2100 GMT. This interpolation introduced additional error into the pressure and temperature data. For the typical interpolation interval of about two hours, the added errors in pressure were +/-0.01 mb. In the extreme and rare case of interpolation over 24 hours, the added error was about =/-2 mb. In the event of a gap in the raw data exceeding two days, no interpolations were made.

The procedure for interpolating the position data to the synoptic times used Kalman smoothing techniques similar to those described in Thorndike and Cheung (1977). Typical estimation errors were +/-300 m for position and +/-1 cm sec^-1 for velocity. When the data were sparse the estimation errors were substantially greater.