Downsampled Radio Occultation Observations for Data Assimilation Models

Elizabeth George, Sean Elvidge, Matthew Angling

Research output: Contribution to conference (unpublished)Paperpeer-review

Abstract

Comprehensive, global and timely specifications of the Earth’s upper atmosphere (ionosphere and thermosphere) are required to ensure the effective operation, planning and management of a diverse range of systems affected by space weather. AENeAS (the Advanced Ensemble electron density [Ne] Assimilation System) is a pre-operational physics-based data assimilation model of the coupled ionosphere-thermosphere system. It assimilates data from a variety of sources in near real time to produce both nowcasts and forecasts up to 24 hours ahead. Spire Global operates a fleet of over 110 low Earth orbit (LEO) satellites capable of collecting radio occultation (RO) observations of the total electron content (TEC). This data set is particularly useful in providing observations in regions where there are typically few or none from other sources (e.g the oceans). The RO measurements comprise TEC sampled every second producing large amounts of data for assimilation in a typical 15-minute AENeAS assimilation window. Not only can the number of observations cause lengthy run times making real-time forecasting (and even nowcasting) unfeasible, but the high cadence data can also violate the typical Kalman filter assumption of independent observations. The use of Spire TEC data in AENeAS has been investigated as part of the ESA funded HRIDE project. In this study we look at ways of reducing the quantity of data assimilated, whilst minimising information loss. This has been done by investigating the temporal dependence of the data. Using the k lag autocorrelation, a threshold when the correlation decreases by a factor of 1/e has been determined. Downsampling has then been performed by both simply slicing the data and taking the mean over a designated time span. The paper shows that in both cases the computational run time has been reduced with only minimal impact on the analysis results.
Original languageEnglish
Publication statusPublished - 28 Oct 2021
Event17th European Space Weather Week - Technology Innovation Centre, Glasgow, United Kingdom
Duration: 25 Oct 202129 Oct 2021
http://esww17.iopconfs.org/home

Conference

Conference17th European Space Weather Week
Abbreviated titleESWW17
Country/TerritoryUnited Kingdom
CityGlasgow
Period25/10/2129/10/21
Internet address

Fingerprint

Dive into the research topics of 'Downsampled Radio Occultation Observations for Data Assimilation Models'. Together they form a unique fingerprint.

Cite this