This project aims to contribute to the understanding of thermospheric AGWs and AGW-induced TIDs by investigating the statistical distribution of the propagation parameters. Specifically, **the goal is to determine the statistical distribution of TID propagation parameters** over a large altitude range (100-320 km) in the bottom-side ionosphere (thus addressing E4), for two periods of heightened AGW activity, excluding those events for which there is statistical evidence that they were not caused by AGWs (thus addressing E1 and E2). For this, the project will use two Dynasonde datasets taken at Wallops Island, VA, USA, covering the months of June and November 2015, separately for the daytime and nighttime measurements (thus addressing E3). The location was selected due to the demonstrated high quality of measurements taken by the Wallops Island Dynasonde.

The project has **4 main objectives**:

**O1.** Determine the raw height stratified statistical distribution (RHSSD) of TID propagation parameters: vertical and horizontal wavelength, horizontal propagation direction, horizontal phase speed, for the two selected Dynasonde datasets from Wallops Island (relevant in addressing elements of difficulty E3 and E4). This result would take into account all instances of observed TIDs, without any filtering, and it is necessary to establish a baseline for a comparison that is discussed later as part of O4.

**O2.** Determine the degree of agreement between Dynasonde derived wave parameters and the AGW dispersion relation. This would require the development of an automated software tool that would: 1. use Dynasonde derived data to obtain the TID propagation parameters, and 2. use the output of a thermospheric model to test the agreement between the TID propagation parameters and the theoretical Dynasonde dispersion relation (relevant in addressing elements of difficulty E1 and E2). This comparison would essentially determine if an observed TID is caused by an underlying AGW or not.

**O3.** Determine the filtered height stratified statistical distribution (FHSSD) of TID propagation parameters, for the same quantities and datasets as for O1, but this time excluding those events for which at O2 it was determined that they do not satisfy the AGW dispersion relation within the statistical uncertainty (relevant in addressing elements of difficulty E1, E2, E3 and E4).

**O4.** Compare the RHSSD and FHSSD, identify the areas of agreement and those of disagreement. Possible areas of agreement (or disagreement) refer to altitude ranges and/or specific frequency bands for which we might observe similarities (or discrepancies) between the two statistical distributions (relevant in addressing elements of difficulty E1, E2, E3 and E4).

The study of AGWs and TIDs presents several **elements of difficulty**:

**E1.** While most AGWs in the thermosphere will have an associated TID, not all TIDs are caused by AGWs. Specifically, not all ionospheric perturbations will follow a plane wave model. For example, plasma bubbles will exhibit some of the features that are characteristic of AGWs. In such cases, trying to attribute wave-like parameters to a different type of phenomenon will only perturb the analysis and lead to possibly erroneous conclusions.

**E2.** If an AGW packet is being heavily attenuated, then its characteristic wavevector will have a large imaginary component. In most practical situations, this component is very difficult to estimate, since most observations only allow for the wavelength to be determined, and from it, the real part of the wavevector. In such cases, the observed TID parameters again no longer satisfy a plane wave model.

**E3.** While wave activity is known to be ubiquitous, it is also highly variable. This is caused by the high atmospheric variability, both in the thermosphere and in the lower and middle atmosphere. Determining the statistical distribution of wave parameters requires large datasets, ideally much larger than a few days.

**E4.** Wave activity can vary significantly with altitude, due to the impact of filtering mechanisms and the exponentially decreasing neutral density. For example, existing research has shown a shift towards lower frequencies at higher altitudes in the mean wave PSD (*Djuth et al., 2010*; *Negrea and Zabotin, 2016*). Similarly, it is expected that the other wave parameters would vary with altitude (*Negrea et al., 2016*), though research in this direction is in its earlier stages.

**References:**

Djuth, F. T., L. D. Zhang, D. J. Livneh, I. Seker, S. M. Smith, M. P. Sulzer, J. D. Mathews, and R. L. Walterscheid (2010), J. Geophys. Res., 115, A08305, doi:10.1029/2009JA014799.

Negrea, C., and N. A. Zabotin (2016), Radio Sci., 51, doi:10.1002/2015RS005823.

Negrea, C., N. Zabotin, T. Bullett, M. Codrescu, and T. Fuller-Rowell (2016), J. Geophys. Res. Space Physics, 121, doi:10.1002/2015JA021574.