A study of radar aspect sensitivity in the lower atmosphere

Abstract

The goal of this thesis is related to atmospheric temperature measurements using in situ techniques in tandem with a direct numerical simulation to better understand the zenith angle dependence of VHF (30-300 MHz) radar backscatter from the atmosphere. We begin our study with a high-resolution balloon-borne in situ temperature measurement made over Wichita, KS, in 1995. Very steep vertical temperature gradients were found at the edges of vertical potential steps, regions of near zero vertical potential temperature gradient. We use wavelet analysis to isolate the organized components of the signal and, after subtraction from the original signal, the residual signal is found to have the characteristics of isotropic turbulence. This confirms our hypothesis that the measured temperature profile is a superposition of coherent structures and a background isotropic turbulence. From a radar perspective, we show that this wavelet analysis allows us to predict the radar backscatter as a function of zenith angle from a high- resolution one-dimensional temperature measurement. Unfortunately, radar measurements were not available at this point. We next explore the cause of aspect sensitivity directly via a multi-instrument investigation of the lower atmosphere over the Jicamarca Radio Observatory (JRO) near Lima, Peru. The joint analysis of radar backscatter and in situ measurements of the temperature structure shows that a combination of Fresnel scattering and turbulence is the most likely explanation for aspect sensitive echoes. Furthermore, the strong backscatter seems to originate from vertical potential temperature steps; such as those observed over Wichita, KS. Finally, we show that the measured potential temperature steps and the structures seen in a direct numerical simulation (DNS) of a Kelvin-Helmholtz instability (KHI) are remarkably similar. Not only do we find good agreement between the observation and the simulation; the similarity is also seen in the wavelet spectrum, which is the behavior of the wavelet coefficient as a function of scale size. We extend the results from experimental observations and numerical simulation by predicting the characteristic radar backscatter and show that it is consistent with observations.