Improved wheat area maps from satellite data reduce official record errors
A Dual-Track Curation-and-Classification Framework for Resolving Ground-Truth Label Noise in Operational Sentinel-2 Wheat Area Estimation
Computer Vision and Pattern Recognition
Summary
Estimating how much land is used to grow wheat often has disagreements between government records and satellite images. The authors developed a method that carefully checks and cleans up the labels used to train a computer to spot wheat fields from satellite data. Their approach focused on important time periods for wheat growth and used a type of machine learning called XGBoost, which worked better than some deep learning methods with the limited data available. This method produced wheat area estimates that matched official numbers more closely than previous map products did.
What this means in practice
- •For agricultural monitoring teams: Provide more accurate wheat area estimates by correcting label noise in administrative data using curated satellite classification methods.
- •For land use analysts: Detect and reconcile discrepancies between satellite-based crop maps and official land records to improve spatial accuracy of agricultural statistics.
Tested on one dataset.
Authors
Kasimali Agharia, Ujjwal Kumar Gupta
Abstract
Operational estimation of wheat-cultivated area is persistently constrained by discordance between administrative record-keeping and remotely sensed classification products. We address this administrative reference discordance for the 2022 Rabi season in Patiala district, Punjab, India, using a thirteen-timestep Sentinel-2 NDVI time series. A curated 849-sample reference dataset, developed through an iterative rule-based bootstrapping procedure, underpins both a feature sensitivity analysis and an operational classifier. Feature sensitivity independently assessed via Cohen's d and gradient-boosted information gain converges on the February-to-March grain-fill window as most discriminative. Four classifiers (1D-CNN, LSTM, hybrid CNN-LSTM, and XGBoost) were benchmarked on an identical 679/170 sample split. XGBoost achieved the highest overall accuracy (78.82%) against deep-learning baselines (64-66%), consistent with tree-based ensembles' favourable parameter-to-sample ratio in low-sample regimes. At full-population deployment across 36.25 million valid district pixels, the operational classifier attained 86.31% precision and 71.05% recall. The predicted wheat extent deviated by only +2.99% from the official tabular target, whereas the government's spatial reference mask exhibited a +25.11% positive area bias against the identical target. This asymmetry indicates that a classifier trained on an auditor-curated reference set reconciles more closely with the official tabular area than the spatial product conventionally used to validate it. We present this dual-track curation-and-classification framework as a methodological reference for crop-area reconciliation in label-noisy administrative settings.