Thermosphere Model Evaluation At Low Altitude Using GOCE Densities.pdf

# Thermosphere Model Evaluation at Low Altitude with GOCE Densities

## Abstract
This study assesses the accuracy of three COSPAR International Reference Atmosphere thermosphere models—JB2008, NRLMSISE-00, and DTM2013—at low Earth orbit (LEO) altitudes (120-300 km) using high-resolution GOCE density data. Due to a lack of accurate observations at these altitudes, model performance is not well documented. The GOCE dataset provides unique insights into the thermosphere's behavior at low altitude.

## Introduction
Semi-empirical thermosphere models are crucial for satellite orbit determination and prediction. Three models—NRLMSISE-00, JB2008, and DTM2013—are ISO standards and were selected by COSPAR in 2012. The study compares these models to GOCE density data from the Science Mission (November 2009 - October 2013) and a special accelerometer dataset from the final three weeks of the mission.

## Methodology
- GOCE Data: High-resolution and high-accuracy densities from 275-230 km are obtained from ESA, with an 80 km along-track resolution and a precision of a few percent.
- Model Comparison: The models' performance is evaluated using three time scales: long (annual), medium (monthly), and short (daily). Metrics include mean, RMS, standard deviation of the observed-to-modeled ratio, and correlation.
- Additional Data: Accelerometer data from the final phase of the mission (3 weeks before re-entry) is used to infer densities between 230-170 km, providing further validation.

## Findings
- The GOCE dataset reveals specific weaknesses in each model at low altitude, highlighting areas for improvement in future revisions.
- DTM2013 shows the best agreement with GOCE data, having assimilated part of the GOCE density dataset during its development.
- NRLMSISE-00 has not been updated with recent total density data, impacting its performance.
- JB2008's evaluation is linked to specific index files, which may change over time.

## Conclusion
The study quantifies the performance of CIRA models at low altitudes, crucial for satellite re-entry predictions and mission design. GOCE data provides valuable insights, enabling more accurate thermosphere model development.