Na. Zabotin et Jw. Wright, Ionospheric irregularity diagnostics from the phase structure functions ofMF/HF radio echoes, RADIO SCI, 36(4), 2001, pp. 757-771
We present a new approach to investigating ionospheric irregularities, usin
g the temporal structure function of totally reflected radio echo phase var
iations. Modern digital ionosondes (e.g., the dynasonde) measure the echo p
hase with very high resolution and precision, at closely spaced antennas, f
requencies, and times. A "stringing" procedure gives continuous and unambig
uous phase variation data for time intervals of any desired length. Quasi-p
eriods of tens of seconds up through several minutes are caused by large-sc
ale movements of the ionospheric plasma, while shorter-period phase variati
ons result from the interaction of the sounding signal with small-scale irr
egularities. The relevant irregularity spatial domain extends from decamete
r radio wavelengths to the first Fresnel scale, a few kilometers. We obtain
a theoretical relation between structure functions of the temporal phase v
ariations and spatial irregularities with a simple model of frozen horizont
al drift. The relation permits solutions of both the direct and inverse pro
blems. Although long-period phase measurements are practicable and essentia
l to exploring larger irregularity scales, they require observing modes ded
icated to multiple fixed-frequency time series, and this undesirably limits
the number of altitudes that can be monitored simultaneously. An alternati
ve "rudimentary structure function" is obtainable from standard dynasonde "
B-mode" ionograms; it offers good altitude and time resolution for irregula
rity studies while permitting other established diagnostics (electron densi
ty profiles, vector velocities, critical frequencies, etc.) with the same d
ata. We show some example analyses by these methods as applied to auroral a
nd magnetic-equatorial dynasonde observations. We find irregularity amplitu
des in the range 0.001 < DeltaN/N < 0.1 (for a nominal scale of 1 km) and s
pectral indices in the range 2 < nu < 4, with evidence of diurnal variation
in both quantities at both locations.