Enhancements to Mini-RF X/C-band Data Quality through Cross-Channel Calibration and Reprocessing Strategies

Since 2009, the Mini-RF synthetic aperture radar (SAR) onboard NASA’s Lunar Reconnaissance Orbiter (LRO) has been collecting both S-band (12.6 cm) or X/C-band (4.2 cm) observations to provide near-global coverage of the Moon, including large portions of permanently-shadowed regions. Mini-RF currently operates in a bistatic configuration after failure of the transmitter in December 2010. Incident signals of circular polarization are transmitted from ground stations on Earth and received by Mini-RF in their H- and V- orthogonal linear polarizations, thus preserving the hybrid polarimetric nature of the radar system. This architecture enables the generation of Stokes parameters, which encode information used to infer surface and near-surface properties. In particular, these data can be used to characterize wavelength-scale surface roughness, regolith density, composition, as well as identify areas of buried water ice deposits. X/C-band coverage includes a significant fraction of the south polar region, making this dataset uniquely capable to inform future exploration and landing site assessment for the Artemis and Commercial Lunar Payload Services (CLPS) programs.

Accurate derivation of the Stokes parameters relies on well-calibrated and isolated H- and V- receive channels. Initial post-launch calibration efforts indicated H- and V- gain imbalances that varied significantly from test-to-test. Moreover, the quality of processed X/C-band observations included artifacts that were not present in S-band data. In spite of these issues, a significant fraction of the south polar region was still observed with X/C-band, with the intent to reprocess the data when the issues contributing to poor data quality are better understood.

In this work, we report on test campaigns aimed to further investigate this observed gain imbalance primarily affecting X/C-band observations. Recent evidence indicates this imbalance is caused by the presence of cross-channel leakage of received signals within the antenna. To correct for this leakage, test data collected from ground stations are used to develop a model to obtain complex correction coefficients. The model is linear, which implies that the impact on the signals can be removed. We present an example application of these coefficients to a collection of X/C-band bistatic observations of Mare Imbrium, demonstrating significant improvement in data quality.

In addition to cross-channel signal leakage, monostatic X/C-band data quality suffered from issues related to the utilization of a commercially-purchased radar processor. Recently, the Mini-RF team has manually reprocessed a small number of "test-case" X/C-band monostatic observations utilizing the in-house bistatic radar processor. Results indicate significant improvements in data quality are achievable. We find that the combination of applying the cross-channel leakage correction to archived monostatic X/C-band data and reprocessing it with a modified version of the current Mini-RF bistatic processing algorithm represents an opportunity to greatly enhance the quality and usability of the data. We anticipate that a fully calibrated and reprocessed X/C-band dataset can provide new insights into lunar regolith processes, acting at smaller scales and to shallower depths relative to complementary S-band observations. This knowledge will augment our understanding of lunar conditions critical to support future human exploration of the Moon.