Establish the cross-site and cross-year generality of the dual-track framework

Determine whether the trade-off between iterative reference-sample curation and classifier capacity generalises to other districts, growing seasons, and crop-label taxonomies beyond the 2022 Rabi wheat-area estimation study in Patiala district, Punjab.

Background

The framework was demonstrated for one district, one growing season, and one wheat label-noise regime. Although reference samples were drawn across Punjab State, deployment and acreage reconciliation were conducted only in Patiala district during the 2022 Rabi season, and the paper notes that the reference samples were not withheld by district.

The conclusion explicitly identifies generalisation of the curation-first versus higher-capacity-classifier trade-off as unresolved. Resolving this problem requires replication across districts, seasons, and crop-label taxonomies, including conditions with different agro-climatic characteristics and possible phenological shifts.

References

Whether this trade-off generalises to other districts, seasons, and crop-label taxonomies is an open empirical question, and we present the dual-track framework -- together with its documented limitations, including the disclosed spatial-autocorrelation caveat in the sampling design (Section~\ref{sec:studyarea}) and the divergent feature-importance rankings reported in Section~\ref{sec:featuresens} -- as a methodological reference for crop-area reconciliation in comparable label-noisy administrative settings, rather than as a definitive or final architecture.

— A Dual-Track Curation-and-Classification Framework for Resolving Ground-Truth Label Noise in Operational Sentinel-2 Wheat Area Estimation  (2609.36975 - Agharia et al., 29 Sep 2026) in Section 7, Conclusion