Processing COSMIC/FORMOSAT-3 data for use in near-real time assimilation

Data assimilation techniques for space weather are finding increasing success in ionospheric remote sensing due to the growing abundance of data from ground and space-based GPS receivers and new UV remote sensing satellites. The six-satellite COSMIC/FORMOSAT-3 constellation, launched in April 2006, now provides unprecedented global coverage from 1500-2000 radio occultation measurements daily. The Global Assimilative Ionosphere Model (GAIM) developed at NASA's Jet Propulsion Laboratory (JPL) is being adapted to process COSMIC data in near real-time (NRT). Calibrated measurements of ionospheric total electron content (TEC) from COSMIC suitable for input into GAIM are currently made available with latencies between 30 and 120 minutes. Ground GPS receiver networks continue to provide near real-time TEC data from two worldwide networks consisting of ~75 sites capable of 5-minute latency, and ~125 additional hourly sites, operated by JPL and others. GAIM combines ground TEC data with vertically resolved occultation data from COSMIC to produce three-dimensional, global electron density distributions that reproduce the hour-to-hour ionospheric “weather” much more accurately than has been possible before the COSMIC era.

We will describe the status of a new processing system for COSMIC being developed to enable NRT data input to GAIM. The system produces TEC measurements every 10 seconds from raw GPS phase and range data acquired between the COSMIC receiver and a GPS satellite. Critical steps in the processing are: leveling the differential phase data to pseudorange data, and estimating the interfrequency biases for the fore and aft receiver/antenna pairs on each COSMIC satellite. COSMIC satellites tend to exhibit large multipath contributions in the range data, leading to poorly determined phase level when not accounted for. The new processing system utilizes a leveling method with weights derived from in-situ multipath estimates, to limit the effect of large multipath on the leveling. Additionally, the new system utilizes an on-the-fly interfrequency bias estimation technique for each COSMIC receiver/antenna pair. This technique uses a moving 10-day window to smooth variability of the daily bias estimates. This COSMIC data processing system is an essential building block needed to implement the near real-time assimilation of COSMIC data into GAIM, providing improved ionospheric weather nowcasting and forecasting.