Integrated Phenological Agricultural Drought Index (IPADI): A phenology-only framework based on the relative usefulness of timing and intensity metrics across irrigated and rainfed croplands
Abstract
Agricultural drought is a slow-onset hazard with severe impacts on semi-arid Mediterranean croplands, where crop growth and yield depend on the full seasonal trajectory of vegetation rather than its condition at any single moment. Existing composite drought indices instead combine heterogeneous inputs such as precipitation, temperature, soil moisture, and instantaneous spectral vegetation indices, treating phenology, at best, as one variable among many rather than as a direct record of how drought reshapes the growing season. To address this gap, we introduce the Integrated Phenological Agricultural Drought Index (IPADI), which is constructed entirely from land-surface phenology. IPADI separates timing (PTDI) and intensity (PIDI) into distinct sub-indices so that the contribution of each to drought detection can be tested rather than assumed. Eight phenological metrics were extracted from 10 m Sentinel-2 NDVI time series using TIMESAT across ten agricultural campaigns (2015/2016–2024/2025) over the irrigated and rainfed zones of the Doukkala plain, Morocco. An exploratory analysis revealed that intensity-related metrics carry more weight than timing-related ones, motivating two sub-indices later combined through a weighted sum. Testing eleven weighting combinations against twelve meteorological, hydrological, and vegetation-based reference indices revealed that the intensity-only configuration as the best performer, so that the operational IPADI retained for the Doukkala plain reduces to its intensity sub-index (IPADI = PIDI), indicating that the seasonal intensity aligns more closely with these seasonally averaged references than timing does in this managed setting. The retained index agreed closely with the reference indices and followed the interannual pattern of national cereal production, while its separate computation for each zone helped discriminate climate-driven drought from a management-induced deficit consistent with the documented reduction, then suspension, of Al Massira reservoir releases. These findings show that a compact, phenology-only framework can serve as a high-resolution, transferable tool for agricultural drought monitoring.
Related
- An explainable deep learning framework for sub-seasonal to seasonal agricultural drought prediction over India
- A Continental-Scale Framework for Harmonised Soil Monitoring in African Agricultural Lands: Design, Implementation, and Baseline Field Observations from the Soils4Africa Project
- Nonlinear thresholds in agricultural landscape fragmentation: A GWRF-SHAP diagnostic framework for differentiated rural land management
- What’s in flower? Phenological temperature sensitivities in a restricted alpine flora
- An Artificial Intelligence Integrated Geospatial Decision-Support Framework for Scenario-Based Climate-Adaptive Land-Use Optimization and Environmental Risk Reduction
BibTeX
@article{oussaoui2026integrated,
title = {{Integrated Phenological Agricultural Drought Index (IPADI): A phenology-only framework based on the relative usefulness of timing and intensity metrics across irrigated and rainfed croplands}},
author = {Sabir Oussaoui and Abdessamad Hadri and Abdelghani Boudhar and Ismaguil Hanadé Houmma and El Mahdi El Khalki and Mouad Abbad and Hatim Bechri and Siham Acharki and Christophe Kinnard},
journal = {European Journal of Agronomy},
year = {2026},
month = nov,
doi = {10.1016/j.eja.2026.128388},
url = {https://hal.science/hal-05850821v1},
note = {CC BY},
}
RIS
TY - JOUR
TI - Integrated Phenological Agricultural Drought Index (IPADI): A phenology-only framework based on the relative usefulness of timing and intensity metrics across irrigated and rainfed croplands
AU - Sabir Oussaoui
AU - Abdessamad Hadri
AU - Abdelghani Boudhar
AU - Ismaguil Hanadé Houmma
AU - El Mahdi El Khalki
AU - Mouad Abbad
AU - Hatim Bechri
AU - Siham Acharki
AU - Christophe Kinnard
PY - 2026
DA - 2026/11/03
JO - European Journal of Agronomy
DO - 10.1016/j.eja.2026.128388
UR - https://hal.science/hal-05850821v1
AB - Agricultural drought is a slow-onset hazard with severe impacts on semi-arid Mediterranean croplands, where crop growth and yield depend on the full seasonal trajectory of vegetation rather than its condition at any single moment. Existing composite drought indices instead combine heterogeneous inputs such as precipitation, temperature, soil moisture, and instantaneous spectral vegetation indices, treating phenology, at best, as one variable among many rather than as a direct record of how drought reshapes the growing season. To address this gap, we introduce the Integrated Phenological Agricultural Drought Index (IPADI), which is constructed entirely from land-surface phenology. IPADI separates timing (PTDI) and intensity (PIDI) into distinct sub-indices so that the contribution of each to drought detection can be tested rather than assumed. Eight phenological metrics were extracted from 10 m Sentinel-2 NDVI time series using TIMESAT across ten agricultural campaigns (2015/2016–2024/2025) over the irrigated and rainfed zones of the Doukkala plain, Morocco. An exploratory analysis revealed that intensity-related metrics carry more weight than timing-related ones, motivating two sub-indices later combined through a weighted sum. Testing eleven weighting combinations against twelve meteorological, hydrological, and vegetation-based reference indices revealed that the intensity-only configuration as the best performer, so that the operational IPADI retained for the Doukkala plain reduces to its intensity sub-index (IPADI = PIDI), indicating that the seasonal intensity aligns more closely with these seasonally averaged references than timing does in this managed setting. The retained index agreed closely with the reference indices and followed the interannual pattern of national cereal production, while its separate computation for each zone helped discriminate climate-driven drought from a management-induced deficit consistent with the documented reduction, then suspension, of Al Massira reservoir releases. These findings show that a compact, phenology-only framework can serve as a high-resolution, transferable tool for agricultural drought monitoring.
SN - 1161-0301
SN - 1873-7331
ER -