A High-Resolution Wheat Phenological Dataset for Disaster Risk Management in China

1. Background Amid escalating climate-driven disasters, reasonably agricultural disaster risk management requires a precise spatiotemporal alignment between climate hazards and crop physiological vulnerability. However, current wheat growth-stage classification schemes are primarily derived from morphological and physiological traits rather than calibrated to the spatiotemporal dynamics of wheat-related disaster risks. Consequently, they often fail to adequately address disaster-sensitive growth stages—key phenological windows that require prioritized monitoring and intervention. Therefore, refining the current wheat growth stage classification framework—by integrating empirically documented disaster occurrence windows across the wheat phenological cycle—into a risk-informed growth stage scheme, and generating corresponding spatially explicit, temporally resolved growth stage datasets, constitutes both a critical scientific priority for advancing precise wheat disaster risk assessment and an operational imperative for evidence-based disaster risk management. A High-Resolution Wheat Phenological Dataset for Disaster Risk Management in China was produced based on a "Physiological-Disaster" coupling framework. This new dataset divides the entire growth cycles of winter and spring wheat in China into three continuous functional phases with distinct risk attributes: Foundation Construction & Adaptation Phase, Sink Capacity Establishment Phase, and Yield Solidification Phase, providing a valuable risk window reference for large-scale wheat disaster risk management. 2. Methodology Guided by the principle of risk homogeneity, traditional discrete phenological stages are reconstructed into continuous vulnerability windows, providing a unified spatiotemporal framework for agricultural disaster risk assessment and management. · Spatial Fusion: The baseline wheat distribution was generated using the 2024 CN_Wheat10 dataset, supplemented with the 2023 China 30 m winter wheat distribution dataset in fragmented wheat-growing regions of southern China (Yunnan, Guizhou, Hunan, and Jiangxi). The merged distribution was subsequently aggregated to a 30 arc-second (~1 km) grid using an Effective Pixel Area Statistical Method, thereby preserving wheat planting area and minimizing aggregation-induced spatial bias. · Spatiotemporal Reconstruction: The framework integrates remote sensing inversions (ChinaCropPhen1km, Version 7), station observations, and physiological logic constraints. Missing nodes were reconstructed using Thiessen polygon spatial interpolation and a physiology-constrained Inverse Deduction Algorithm. · Validation: Chi-square goodness-of-fit tests against historical meteorological disaster records (2010–2020) demonstrated strong "phase-locking" characteristics (P < 0.001). For example, 95% of recorded heat stress events and 97% of hail events occurred within the defined Yield Solidification Phase. 3. Dataset Overview 3.1 Key Features · Temporal Representation: Multi-year climatological mean phenology. · Spatial Resolution: 30 arc-seconds (~1 km at the equator). · Coordinate system: WGS 84 (EPSG:4326) · Crop Types: Winter Wheat and Spring Wheat in China. · Variables: DOY of key phenological stages and durations (days) of functional vulnerability phases. 3.2 Dataset Structure The dataset is provided as a compressed archive (A High-Resolution Wheat Phenological Dataset for Disaster Risk Management in China_v1.0), containing: 1_Onset_Dates/ · Description: Onset dates of key phenological stages. · File Format: GeoTIFF (9 files total). · Naming Convention: [CropType]_Onset_[StageIndex]_[StageName].tif Example: WinterWheat_Onset_02_Overwintering.tif 2_Phase_Durations/ · Description: Continuous durations of specific functional risk phases. · File Format: GeoTIFF (9 files total). · Naming Convention: [CropType]_Duration_[Phase/Total]_[Start2End].tif Example: WinterWheat_Duration_Phase1_Sowing2Overwintering.tif 3_Spatial_Mask/ · Description: Spatial distribution mask extracting valid wheat pixels. · File Format: GeoTIFF (1 file total). · File Name: Wheat_Distribution_Mask_30s.tif (Value 1: Winter Wheat, Value 2: Spring Wheat, NoData: Non-wheat regions). 4_Production_Technical_Annex.docx · Description: Detailed methodology, physiological constraint parameters (e.g., dormancy gap thresholds), validation results, and workflow documentation. 4. Potential Applications · Quantitative Disaster Modeling: Provides high-resolution temporal references for disaster models, enabling dynamic weighting and differentiated vulnerability characterization for specific sensitive stages when quantifying whole-growth-cycle risks. · Agricultural Financial Risk Hedging: Assists in precisely matching risk windows for weather index insurance, reducing the basis risk associated with fixed calendar dates. · Disaster Management Scheduling: Offers a scientific foundation for cross-regional resource allocation based on the revealed "time compression" effects and spatial heterogeneity. Notes (Data Handling & Limitations): · Cross-Year DOY: In the 1_Onset_Dates folder, DOY values for the overwintering onset strictly follow the 1–365 calendar format. When calculating phase durations across the calendar year, users must apply a conditional adjustment (e.g., adding 365 to next-year DOYs) to maintain chronological continuity. · Spatial Smoothing: In regions with sparse meteorological stations, the application of Thiessen polygon interpolation and fixed-parameter deduction may induce a degree of numerical homogenization (spatial smoothing effect) among adjacent pixels. Users are advised to consider these methodological characteristics when conducting large-scale or macro-regional assessments.