Assessing the performance of thermospheric modelling with data assimilation throughout solar cycles 23 and 24

TL;DR

This study evaluates the effectiveness of data assimilation in thermospheric models, TIEGCM and DTM, across solar cycles 23 and 24, highlighting their performance and potential for operational space weather forecasting.

Contribution

It compares physical and semi-empirical thermospheric models with satellite data, demonstrating data assimilation improves model accuracy and supports operational forecasting.

Findings

Both models underestimate densities at solar maximum.

DTM performs better at solar minimum.

Data assimilation yields an average ~4% improvement.

Abstract

Data assimilation procedures have been developed for thermospheric models using satellite density measurements as part of the EU Framework Package 7 ATMOP Project. Two models were studied; one a general circulation model, TIEGCM, and the other a semi-empirical drag temperature model, DTM. Results of runs using data assimilation with these models were compared with independent density observations from CHAMP and GRACE satellites throughout solar cycles 23 and 24. Time periods of 60 days were examined at solar minimum and maximum, including the 2003 Hallowe'en storms. The differences between the physical and the semi-empirical models have been characterised. Results indicate that both models tend to show similar behaviour; underestimating densities at solar maximum, and overestimating them at solar minimum. DTM performed better at solar minimum, with both models less accurate at solar maximum. A mean improvement of ~4% was found using data assimilation with TIEGCM. With further improvements, the use of general circulation models in operational space weather forecasting (in addition to empirical methods currently used) is plausible. Future work will allow near-real-time assimilation of thermospheric data for improved forecasting.