Assessing the impact of middle atmosphere observations on day-to-day variability in lower thermospheric winds using WACCM-X

Highlights

• Atmospheric specifications in the 50-100 km altitude region significantly impact whole atmosphere simulations.

• Thermospheric tides are affected by middle atmosphere specifications via non-linear wave-wave interactions.

• Middle atmospheric observations affect barotropic and baroclinic instabilities that give rise to traveling planetary waves.

Abstract

Recent studies have shown that day-to-day variability in thermospheric winds (100–300 km altitude) driven by meteorological variability from below affects ionospheric E and lower F regions, highlighting the need for accurate, continuous specification of day-to-day variability throughout the entire atmosphere for geospace weather prediction systems. To better understand the nature of forcing from below on the coupled thermosphere/ionosphere system, this study uses the Specified Dynamics Whole Atmosphere Community Climate Model eXtended (SD-WACCM-X) to quantify how the meteorology of the underlying atmosphere impacts the thermosphere. For this study, global meteorological specifications are produced by a high-altitude version of the Navy Global Environmental Model (NAVGEM-HA), which assimilates standard meteorological observations from the surface through the lower atmosphere, and satellite-based observations of temperature and constituents in middle atmosphere (MA) region 10–90 km altitude. Two SD-WACCM-X simulations for the January–February 2013 period are performed using NAVGEM-HA specifications produced with and without assimilation of MA observations. Results show that the availability of MA observations strongly constrains the modeled spectrum of planetary scale waves (zonal wavenumbers 1–3) in the thermosphere. Specifically, the amplitudes of the solar non-migrating DE3 tide and westward quasi-two day wave (Q2DW) are nearly twice as large in SD-WACCM-X simulations without MA observations compared to simulations with MA observations. Model diagnostics show that these differences are related to non-linear wave-wave interactions impacting the DE3 mode and to sources of baroclinic/barotropic instability near the summer mesospheric easterly jet impacting the Q2DW. This study highlights the importance of MA observations for constraining whole atmosphere models needed for next-generation space weather prediction capabilities.

Introduction

The development of whole atmosphere models (ground to exobase) in the last decade along with the availability of stratospheric and mesospheric atmospheric data specifications has brought the geospace community closer to understanding the short-term variability (times days) in the lower thermosphere: when the Sun is quiescent, a large fraction of the day-to-day thermospheric variability is driven by lower atmosphere weather, which is then mapped into ionospheric behavior because of the wind-dynamo coupling. For example, migrating solar tides generated by the absorption of solar radiation in the lower atmosphere (Chapman and Lindzen, 1970) are associated with a modulation of the daily variability of vertical ion drifts (Millward et al., 2001; Fang et al., 2013). In particular, the migrating semidiurnal solar tide has been associated with a shift of vertical ion drifts, and consequently of the peak electron density, to earlier local times during days immediately following a sudden stratospheric warming (e.g., Goncharenko et al., 2010; McDonald et al., 2015; Fuller Rowell et al., 2017). Non-migrating tides associated with the zonal structure of the ionosphere (Forbes et al., 2008) have been found to be very sensitive to the background meteorological conditions of the lower atmosphere (McDonald et al., 2018). Traveling planetary waves are large-scale oscillations of the type of Rossby or Rossby-gravity that are generated by instabilities of the background flow or by any other type of forcing that are close to the spectral location of the traditional normal modes (see, Sassi et al., 2019). Together, migrating tides, nonmigrating tides, and free traveling planetary waves are major drivers of day-to-day variability in the coupled ionosphere-thermosphere system related to meteorological variability in the lower atmosphere. This study uses a whole atmosphere model driven by atmospheric specifications below 90 km to describe how specification of lower atmospheric meteorological variability impacts the thermosphere through these different processes.

Whole atmosphere modeling has benefited from increased computing capability that has allowed the inclusion of more sophisticated representation of atmospheric physics, refined horizontal resolution and deeper vertical domain (e.g. Liu et al., 2010, 2018). Equally important toward the development of such whole atmosphere modeling capabilities has been the availability of satellite-based observations of the middle atmosphere (10–100 km altitude), enabling the extension of forecast/assimilation systems to the lower thermosphere (Eckermann et al., 2009, 2018; Wang et al., 2011; McCormack et al., 2017; Pedatella et al., 2019). The development of these systems provides new research tools that can be used to explore fundamental questions in theoretical understanding of whole atmosphere interactions (Liu, 2016; Sassi et al., 2019, for reviews).

As models extend to higher altitudes, the quality of upper atmospheric predictions will depend critically on global satellite-based observations in the data-poor middle atmosphere region, through which tides and planetary waves propagate. Previous studies have documented that the day-to-day variability of the zonal mean jet in the upper mesosphere and lower thermosphere (UMLT) is prominently affected by and depends on the altitude where observations are available to constrain the atmospheric flow (e.g. Sassi et al., 2018). We also have theoretical evidence that the lower atmospheric meteorology impacts thermospheric mass densities (Sassi et al., 2016) and tidal variability (Pedatella and Liu, 2012, 2013), with repercussion on ionospheric morphology (Pedatella and Liu, 2013; McDonald et al., 2015, 2018). At the same time, the quality of atmospheric specifications produced by modern numerical weather prediction (NWP) systems extending to the UMLT has benefited from observations therein to reduce analysis errors (Ren et al., 2011), and the observation-minus-forecast (O-F) differences (Hoppel et al., 2013).

In light of these facts, the ability to accurately model, and ultimately predict, thermospheric “weather” linked to forcing from below via planetary waves and tides depends critically on the availability of middle atmosphere observations. The concern is that many of the research satellites providing key middle atmospheric observations of temperature and constituents have far exceeded their originally planned lifetimes, and in many cases no replacements will be readily available, creating a possible future gap in observational capabilities. To the best of our knowledge, the fundamental question of how our ability to model thermospheric variability due to solar tides and traveling planetary waves would be impacted by loss of MA observations has not yet been fully addressed. The goal of this study is to quantify such impacts using a whole atmosphere model with specified dynamics provided by a state-of-the-art data assimilation system extending through the altitude region from 0 to 100 km. We focus on the case of the January–February 2013 SSW, which previous studies have shown to be a dramatic example of lower atmospheric variability impacting the coupled thermosphere/ionosphere system.