Hageltom et al. (2026) Adapting Noah-MP-Crop for sorghum modeling in a semi-arid environment: a case study of Gedaref region in East Sahel

Identification

• Journal: Agricultural and Forest Meteorology
• Year: 2026
• Date: 2026-10-17
• Authors: Yasir Hageltom, Joël Arnault, Nadir Ahmed Elagib, Patrick Laux, Harald Kunstmann
• DOI: 10.1016/j.agrformet.2026.111533

Research Groups

• Institute of Meteorology and Climate Research - Atmospheric Environmental Research (IMKIFU), Campus Alpin, Karlsruhe Institute of Technology (KIT), Garmisch-Partenkirchen, Germany
• Institute of Geography and Center for Climate Resilience, University of Augsburg, Augsburg, Germany
• Institute of Geography, Faculty of Mathematics and Natural Sciences, University of Cologne, Cologne, Germany

Short Summary

This study adapted the Noah-MP-Crop model for rainfed sorghum in Sudan's semi-arid Gedaref region by developing sorghum-specific parameters and integrating satellite-derived dynamic sowing dates. The adapted model, particularly with dynamic sowing, significantly improved simulations of sorghum canopy development and yield compared to a fixed sowing date approach.

Objective

• To develop sorghum-specific physiological and phenological parameters for the Noah-MP-Crop model using a stepwise calibration approach.
• To introduce a dynamic sowing-date scheme informed by remote-sensing observations to better reflect rainfall-driven management practices in semi-arid regions.
• To assess the model’s skill in reproducing sorghum growth, canopy development (Leaf Area Index), and yield under rainfed conditions.

Study Configuration

• Spatial Scale: Gedaref State, Sudan (eastern Sahel, Blue Nile Basin, approximately 33–37 °E, 12–15 °N), simulated on a 4 km grid.
• Temporal Scale:
  • Spin-up: 2012–2013
  • Calibration: 2014–2018
  • Validation: 2019–2021 (surface energy fluxes limited to 2019–2020)
  • Sowing date evaluation: 2022 growing season
  • Long-term simulation: 2000–2024

Methodology and Data

• Models used: Noah-MP-Crop (an extension of the Noah-MP land surface model), configured within the WRF-Hydro framework.
• Data sources:
  • Atmospheric Forcing: Hybrid dataset combining ERA5-Land reanalysis (hourly, 0.1° spatial resolution for wind speed, 2-meter air temperature, 2-meter specific humidity, surface pressure, downward shortwave and longwave radiation) and Integrated Multi-satellite Retrievals for GPM (IMERG) precipitation product (half-hourly, 0.1° spatial resolution).
  • Land Cover: Copernicus Global Land Service (CGLS) Dynamic Land Cover product (100 m spatial resolution), reclassified to IGBP 17-class scheme.
  • Soil Properties: FAO/UNESCO Digital Soil Map of the World (30-arc-second, ~1 km spatial resolution).
  • Sorghum Distribution: Spatial Production Allocation Model (SPAM) 2020 dataset (10 km spatial resolution).
  • Sowing Dates: Farmer-reported observations (80 locations, 2022 season) for evaluation; MODIS Surface Reflectance (MOD09GQ Collection 6.1, 250 m spatial resolution) for Normalized Difference Vegetation Index (NDVI)-based dynamic sowing date estimation.
  • Model Evaluation Data:
    • Surface Energy Fluxes: FLUXCOM MLM-ERA5 ensemble product (daily, 0.1° spatial resolution) for latent heat flux (LE) and sensible heat flux (H).
    • Vegetation Dynamics: Global Land Surface Satellite (GLASS) LAI product version 6 (8-day, 500 m spatial resolution).
    • Sorghum Yield: Mechanized Farming Corporation of Gedaref State records (5 locations, 2014–2021, excluding 2017).
  • Calibration: Stepwise approach using Latin Hypercube Sampling (LHS) with a multi-objective function based on residual Kling–Gupta Efficiency (KGE) scores for LAI, grain yield, LE, and H.
  • Sensitivity Analysis: Partial Rank Correlation Coefficient (PRCC) using LHS parameter ensembles.

Main Results

• Dynamic Sowing Dates: The satellite-based method estimated sowing dates with a mean bias of -2 days and a Mean Absolute Error (MAE) of 12.8 days against field observations. It successfully captured spatial gradients and interannual variability in sowing timing.
• LAI Simulation: Dynamic sowing significantly improved LAI simulation during validation, increasing Nash-Sutcliffe Efficiency (NSE) from 0.27 to 0.57 and reducing MAE from 0.21 to 0.16 compared to fixed sowing.
• Energy Fluxes: Both dynamic and fixed sowing configurations reproduced seasonal patterns of LE and H well, with minor differences. LE simulation showed high performance (validation NSE ~0.75), while H performance was moderate (validation NSE ~0.44).
• Yield Simulation: Dynamic sowing reduced yield MAE by 27% during calibration and 55% during validation relative to the fixed sowing date approach. Performance varied across sites, with some showing strong agreement (e.g., R² = 0.99, MAE = 43 kg ha⁻¹ at Alhawata during calibration). Long-term simulations showed spatial yield patterns consistent with rainfall gradients.
• Parameter Sensitivity: LAI and yield were highly sensitive to phenological parameters (GDDS1-GDDS5), photosynthetic efficiency (QE25), and biomass allocation to grain (GRAINPTS5, GRAINPTS6). Latent heat flux was less sensitive to crop parameters.

Contributions

• First adaptation and evaluation of the Noah-MP-Crop model for rainfed sorghum in a semi-arid African dryland environment (Gedaref region, Sudan).
• Development and integration of sorghum-specific phenological and physiological parameters into Noah-MP-Crop.
• Implementation and validation of a novel dynamic sowing-date scheme derived from satellite NDVI, demonstrating its critical role in improving crop growth and yield simulations in rainfall-dependent agricultural systems.
• Application of a robust stepwise, multi-objective calibration and sensitivity analysis approach, constrained by diverse satellite and field observations, to address parameter uncertainty.
• Provides a refined modeling framework for assessing climate variability impacts on sorghum productivity and supporting food security in vulnerable semi-arid regions.

Funding

• German Federal Ministry for Research, Technology, and Space (Bundesministerium für Forschung, Technologie und Raumfahrt, BMFTR), under grant no 02WGR1643A-C through the SPS-Blue Nile project.
• Open access funding enabled and organized by Projekt DEAL.

Citation

@article{Hageltom2026Adapting,
  author = {Hageltom, Yasir and Arnault, Joël and Elagib, Nadir Ahmed and Laux, Patrick and Kunstmann, Harald},
  title = {Adapting Noah-MP-Crop for sorghum modeling in a semi-arid environment: a case study of Gedaref region in East Sahel},
  journal = {Agricultural and Forest Meteorology},
  year = {2026},
  doi = {10.1016/j.agrformet.2026.111533},
  url = {https://doi.org/10.1016/j.agrformet.2026.111533}
}