You are here

Technical report | Calculation of High Frequency Land Backscatter Coefficients

Abstract

A suitable model of the return ground clutter is required to help assess the performance of over-the-horizon radar. Currently, models for the clutter reflected from the sea exist but there are no models for the backscatter from land. The backscatter coefficient, which characterises the backscattered power, can be determined by considering the difference between observed backscatter ionograms and synthesised ionograms. The synthesised ionograms were generated using a MATLAB ray tracing toolbox, PHaRLAP, and the JORN Real Time Ionospheric Model. Data from the Laverton and Longreach backscatter sounders in September 2015 were analysed and backscatter coefficient results for sea, desert, plateau and hilly terrain in the Northern Territory were determined. It was found that the backscatter coefficient was large for hilly and rough terrains. Conversely, flat, dry deserts produced a lower backscatter coefficient.

Executive Summary

To assess the performance of over-the-horizon radar, a reliable model of the return clutter is necessary. The ionosphere, the various propagation losses and backscatter from the sea are well understood; however, a suitable model for the return backscatter from land is lacking. This is due to the variable nature of the terrain.

Five regions in Northern Australia were analysed and a dataset for the land backscatter coefficients was collated. The data analysis involved comparing the difference between observed backscatter ionograms (plots of the return power as a function of group range and frequency) and synthesised ionograms. The synthesised ionograms were produced using a ray tracing toolbox, PHaRLAP, and the JORN Real Time Ionospheric Model. Range-frequency data cells that contained only one mode of propagation were considered. This was enforced by using filters and manually selecting the desired area on the observed ionogram.

For the purpose of validation of the methodology, sea backscatter was analysed. A sea backscatter coefficient of -22.5 dB was determined, which agrees well with theory for a fully developed sea state. The Simpson Desert was found to have a low backscatter coefficient of -35.0 dB and the Tanami Desert region in the Northern Territory had an aspect sensitive backscatter coefficient of -34.8 dB and -29.4 dB as measured by the Laverton and Longreach sounders respectively. A region in the north of NT near the Daly River had a large backscatter coefficient of -19.7 dB. This region has undulating terrain and we hypothesise that this is the cause of the relatively large value for the backscatter coefficient. The Longreach and Laverton sounders were also used to investigate a region in Central Arnhem in the north of NT. Values of -23.7 dB and -27.5 dB were determined as the backscatter coefficients for this region. We hypothesize that the larger value is due to aspect sensitive backscatter from the Mitchell Ranges in this region.

Key information

Author

Benjamin Slimming and Manuel Cervera

Publication number

DST-Group-TR-3613

Publication type

Technical report

Publish Date

June 2019

Classification

Unclassified - public release

Keywords

High frequency radio wave backscatter, High frequency radio wave propagation, HF backscatter sounder, modelling