Ablila et al. (2025) Development and Validation of a New Remotely Sensed Combined Drought Anomaly Index (CDAI) for Monitoring Agriculture Drought Over Morocco

Identification

Journal: Earth Systems and Environment
Year: 2025
Date: 2025-12-01
Authors: Youness Ablila, El Houssaine Bouras, Abdelhakim Amazirh, Saïd Khabba, Riad Balaghi, Maria‐José Escorihuela, Abdelghani Chehbouni, Salah Er‐Raki
DOI: 10.1007/s41748-025-00948-w

Research Groups

AgroBiotech Center, Faculty of Sciences and Technics, Cadi Ayyad University, Marrakech, Morocco
Escola Tècnica Superior d’Enginyeria Agroalimentària i Forestal i de Veterinària, Universitat de Lleida, Spain
CRSA, Centre for Remote Sensing Applications, Mohammed VI Polytechnic University, Benguerir, Morocco
LMFE, Faculty of Sciences Semlalia, Cadi Ayyad University, Marrakech, Morocco
National Institute for Agronomic Research (INRA), Rabat, Morocco
isardSAT, Barcelona, Spain
CESBIO, Centre d’Etudes Spatiales de la BIOsphère, CNES/CNRS/IRD/UPS, Toulouse, France

Short Summary

This study developed a new Combined Drought Anomaly Index (CDAI) for monitoring agricultural drought in Morocco, integrating multiple remote sensing-based indices using Principal Component Analysis. The CDAI was validated against cereal yield and in-situ precipitation data, demonstrating strong correlations and superior performance in early drought detection compared to existing indices.

Objective

To develop a new Combined Drought Anomaly Index (CDAI) that integrates multiple aspects of agricultural drought conditions (vegetation health, land surface temperature, precipitation, and evapotranspiration) for a more robust assessment across rainfed cereal areas in Morocco.
To create a robust and reliable index capable of detecting drought impacts several months in advance and across different agro-climatic zones to improve agricultural drought management strategies.

Study Configuration

Spatial Scale: Rainfed cereal-producing areas across 33 Moroccan provinces. Data aggregated to approximately 5.56 kilometers spatial resolution.
Temporal Scale: Monthly data from 2000 to 2022 (22-year period).

Methodology and Data

Models used:
- Principal Component Analysis (PCA) for CDAI development.
- Penman-Monteith equation (used in MODIS ET product).
- Linear regression for detrending cereal yield data.
- Pearson correlation for validation.
Data sources:
- Remote Sensing:
  - MODIS MOD13Q1.061 (Normalized Difference Vegetation Index - NDVI)
  - MODIS MOD16A2 (Evapotranspiration - ET)
  - MODIS MOD11A1 V6.1 (Land Surface Temperature - LST)
  - CHIRPS (Precipitation)
  - ESA CCI Land Cover maps ("Cropland, rainfed" class)
- Observation/In-situ:
  - Rainfed cereal yield data (durum wheat, bread wheat, barley) from the Ministry of Agriculture of Morocco for 33 provinces.
  - In-situ precipitation data from 21 meteorological stations for Standardized Precipitation Index (SPI-3, SPI-6) calculation.

Main Results

The CDAI showed strong correlations with Detrended Standardized Cereal Yield (DSCY), with approximately 76% of provinces exhibiting correlation coefficients (R) ranging from 0.60 to 0.89. The overall correlation reached 0.87 (p < 0.01).
CDAI exhibited strong agreement with in-situ Standardized Precipitation Index (SPI-3 and SPI-6), with R values reaching 0.90 in December and January (for SPI-3) and 0.89 in February (for SPI-6) (p < 0.01).
CDAI outperformed commonly used indices (Vegetation Condition Index - VCI, Temperature Condition Index - TCI, and Vegetation Health Index - VHI) in both accuracy and stability during the early agricultural drought detection period (January to March).
Analysis of the CDAI indicated that Morocco experienced drought conditions during 32% of the study period (2000–2022), with 5.25% of the time categorized as exceptional drought.
The 2015–2016 agricultural season, particularly December 2015, was identified as the driest period, affecting 63.6% of the provinces and with 57% of the area experiencing exceptional drought. The 2008–2009 season was the wettest.

Contributions

Development of a new, open-source, remote sensing-based Combined Drought Anomaly Index (CDAI) specifically tailored for agricultural drought monitoring in Morocco.
Integration of multiple complementary remote sensing variables (NDVI, LST, ET, Precipitation) using Principal Component Analysis to provide a comprehensive and robust drought assessment.
Demonstrated superior performance in early agricultural drought detection (up to four months in advance) compared to existing single-source (VCI, TCI) and composite (VHI) indices.
Extensive validation using independent in-situ cereal yield data and Standardized Precipitation Index (SPI) across 33 provinces, confirming its reliability.
Provides a consistent and stable drought monitoring tool across different regions and months, overcoming the variability limitations of other indices.
Enables high-resolution (approximately 5 kilometers) and long-term (2000–2022) drought detection, offering valuable insights for early warning systems and agricultural risk management, especially in areas lacking in-situ observations.

Funding

Joint International Laboratory TREMA
PRIMA-IDEWA project
PRIMA-BIOMEnext project
Yield Gap project (agreement between OCP Foundation and UM6P)
GrowCast project (funded by OCP through UMRP program)
European Commission Horizon 2020 Programme for Research and Innovation (H2020) Marie Sklodowska-Curie Research and Innovation Staff Exchange (RISE) action (ACCWA, grant agreement no: 823965)

Citation

@article{Ablila2025Development,
  author = {Ablila, Youness and Bouras, El Houssaine and Amazirh, Abdelhakim and Khabba, Saïd and Balaghi, Riad and Escorihuela, Maria‐José and Chehbouni, Abdelghani and Er‐Raki, Salah},
  title = {Development and Validation of a New Remotely Sensed Combined Drought Anomaly Index (CDAI) for Monitoring Agriculture Drought Over Morocco},
  journal = {Earth Systems and Environment},
  year = {2025},
  doi = {10.1007/s41748-025-00948-w},
  url = {https://doi.org/10.1007/s41748-025-00948-w}
}

Original Source: https://doi.org/10.1007/s41748-025-00948-w