Polarimetric synthetic aperture radar (PolSAR) can observe the ground day and night and in all weather, making it valuable for land-cover mapping, change detection, disaster monitoring, ocean surveillance, and crop assessment. The quadrature-polarimetric (QP) mode collects rich scattering information, but this information comes with trade-offs, including narrower swath coverage and range-ambiguity challenges. Hybrid-polarimetric (HP) mode offers a practical alternative by transmitting circularly polarized waves and receiving horizontal and vertical channels simultaneously. It can widen coverage, improve revisit opportunities, and reduce data volume. Based on these challenges, deeper research is needed to determine when HP mode can reliably support real-world remote-sensing applications.
Researchers from the National Key Laboratory of Microwave Imaging, Aerospace Information Research Institute, Chinese Academy of Sciences; the University of Chinese Academy of Sciences; The University of Tokyo; the China Center for Resources Satellite Data and Application; and Henan University reported (DOI: 10.34133/remotesensing.1025) the findings in Journal of Remote Sensing on 22 April 2026. The study evaluates the target polarimetric representation capability (TPRC) of HP mode using real LuTan-1 data.
To ensure a fair comparison, the team developed a general refined Wishart pseudomixture model (GRWpMM), a unified classification framework applicable to multiple polarimetric modes. The method incorporates training set presegmentation (TSP), which divides complex land-cover categories into physically more consistent subgroups, and a reliability-based training strategy designed to improve robustness under reduced polarimetric information. The researchers tested the approach on three L-band datasets: the classic Flevoland dataset, and real LuTan-1 observations from Hami and Xinxiang. The LuTan-1 datasets were labeled using multispectral imagery and in situ detection, covering cropland, bare soil, urban areas, greenhouses, forest, water, buildings, and other land covers. Results showed that HP and QP modes achieved similar TPRC for targets without specific orientation properties, with average producer's accuracies of 97.62% and 99.26%, respectively. For targets with clear orientation properties, however, accuracy dropped to 83.32% for HP, compared with 94.07% for QP.
The authors said HP mode should not be viewed simply as a lower-information substitute for QP mode. Instead, they said, it has a clear operating window: it can support efficient, wide-area observation when targets have relatively stable or non-oriented scattering behavior, but it requires caution for structures with strong directionality. They said the study's value lies in showing not only where HP performs well, but also why it struggles—especially when steep slopes or oriented dihedral-like structures create ambiguity in the radar's polarimetric representation.
The findings provide practical guidance for future Earth-observation and lunar radar missions that must balance coverage, revisit frequency, data volume, and target information. HP mode appears well suited for large-scale monitoring of crops, grasslands, sea ice, and lunar surfaces, where orientation effects are weaker and broad, frequent coverage is valuable. For urban structures, greenhouses, and other directionally organized targets, the study suggests that HP can still be useful, but should be supported by texture, spatial, or other complementary features. As compact polarimetric modes become more widely adopted, LuTan-1 offers an important benchmark for deciding when HP can stand alone and when fuller polarimetric information remains necessary.
###
References
DOI
10.34133/remotesensing.1025
Original Source URL
https://spj.science.org/doi/10.34133/remotesensing.1025
Funding information
This work was financially supported by the Excellent Young Scientists Fund, China (Grant No. 62422121).
About Journal of Remote Sensing
Journal of Remote Sensing, an online-only Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.