Kiranmai et al. (2025) Moisture stress assessment in rabi maize through UAV-mounted multispectral sensor

Identification

• Journal: Frontiers in Environmental Science
• Year: 2025
• Date: 2025-10-10
• Authors: Yalamareddy Kiranmai, M. P. Potdar, D. P. Biradar, V. B. Kuligod, Aditya Kamalakar Kanade, Brajendra Parmar, V. S. Malunjkar
• DOI: 10.3389/fenvs.2025.1651422

Research Groups

• Department of Agronomy, University of Agricultural Sciences, Dharwad, Karnataka, India
• Department of Soil Science and Agricultural Chemistry, University of Agricultural Sciences, Dharwad, Karnataka, India
• Department of Soil Science and Agricultural Chemistry, ICAR- Indian Institute of Rice Research, Hyderabad, Telangana, India
• Centre of Excellence on Digital Technologies for Smart and Precision Agriculture, Mahatma Phule Krishi Vidyapeeth, Rahuri, Maharashtra, India

Short Summary

This study evaluated the effectiveness of UAV-mounted multispectral vegetation indices and the Crop Water Stress Index (CWSI) for detecting moisture stress in rabi maize under nine irrigation regimes. It found that reproductive stages are highly vulnerable to water deficit, leading to significant yield reductions, and that these indices strongly correlate with both stress levels and kernel yield.

Objective

• To evaluate the sensitivity of UAV-derived multispectral vegetation indices for detecting water stress in maize.
• To analyze the relationship between these vegetation indices and maize kernel yield.
• To assess the correlation between the Crop Water Stress Index (CWSI), vegetation indices, and kernel yield.

Study Configuration

• Spatial Scale: Field experiment conducted at the Main Agricultural Research Station, University of Agricultural Sciences, Dharwad, Karnataka, India (15°29′N, 74°59′E; 678 m above mean sea level). The experiment involved maize hybrid NK-6240 under nine irrigation regimes with three replications.
• Temporal Scale: Rabi season of 2021–22. UAV multispectral imagery was acquired at the dough stage (R4), while ground-based NDVI and CWSI measurements were taken at tasseling (VT), silking (R1), milky (R3), and dough (R4) stages.

Methodology and Data

• Models used:
  • Normalized Difference Vegetation Index (NDVI)
  • Renormalized Difference Vegetation Index (RDVI)
  • Soil-Adjusted Vegetation Index (SAVI)
  • Optimized Soil-Adjusted Vegetation Index (OSAVI)
  • Transformed Chlorophyll Absorption in Reflectance Index (TCARI)
  • Crop Water Stress Index (CWSI) based on the empirical model by Idso et al. (1981).
  • Statistical analysis: Analysis of Variance (ANOVA), Duncan’s Multiple Range Test (DMRT), Pearson correlation analysis.
• Data sources:
  • UAV-mounted MicaSense RedEdge multispectral sensor (on a DJI Phantom 4 quadcopter) capturing blue (475 nm), green (560 nm), red (668 nm), red-edge (717 nm), and near-infrared (840 nm) bands.
  • Hand-held GreenSeeker sensor for ground-based NDVI.
  • Portable hand-held infrared thermometer for canopy temperature.
  • Hygrometer for field air temperature and relative humidity.
  • Ground-based kernel yield measurements.
  • Local weather station data.

Main Results

• Irrigation limited to the knee-high stage or omitted at tasseling and silking significantly reduced vegetation indices and kernel yield (p ≤ 0.05).
• NDVI, RDVI, SAVI, and OSAVI exhibited strong positive correlations with kernel yield (r = 0.90 to 0.99) and strong negative correlations with CWSI (r = -0.88 to -0.99).
• TCARI increased with stress, showing a strong positive correlation with CWSI (r = 0.96) and a strong negative correlation with yield (r = -0.97), reflecting pigment degradation.
• CWSI reliably reflected water stress, with the highest sensitivity observed during tasseling and silking stages, where values were up to threefold higher under severe stress compared to full irrigation.
• Kernel yield was highest under full irrigation (7040.7 kg ha⁻¹) and significantly reduced by 18%–63% under various stress regimes, with the lowest yields (2628.3–2729.7 kg ha⁻¹) observed when irrigation was severely limited.
• Stage-specific non-water-stressed (LL) baselines for CWSI shifted across phenological stages (slope from -2.82 °C kPa⁻¹ to -1.32 °C kPa⁻¹, intercept from 3.07 °C to 1.15 °C), while the non-transpiring upper limit (UL) remained nearly constant (<0.1 °C).

Contributions

• Integrated UAV multispectral indices (at R4) with stage-specific ground-based CWSI measurements (VT, R1, R3, R4) to provide a comprehensive framework for stress detection and yield prediction in maize.
• Demonstrated the high sensitivity of maize reproductive stages (tasseling, silking) to water deficit using both spectral and thermal indicators.
• Locally calibrated CWSI baselines, confirming that the lower baseline (LL) is the primary driver of variability, which enhances the robustness of stress quantification compared to theoretical models.
• Employed a rigorous experimental design with nine distinct irrigation treatments, creating a strong gradient for evaluating the sensitivity of various indices and their impact on yield.
• Highlighted the strong potential of combining UAV-based multispectral and thermal sensing for guiding precision irrigation and improving water use efficiency in maize production.

Funding

The authors declare that no financial support was received for the research and/or publication of this article.

Citation

@article{Kiranmai2025Moisture,
  author = {Kiranmai, Yalamareddy and Potdar, M. P. and Biradar, D. P. and Kuligod, V. B. and Kanade, Aditya Kamalakar and Parmar, Brajendra and Malunjkar, V. S.},
  title = {Moisture stress assessment in rabi maize through UAV-mounted multispectral sensor},
  journal = {Frontiers in Environmental Science},
  year = {2025},
  doi = {10.3389/fenvs.2025.1651422},
  url = {https://doi.org/10.3389/fenvs.2025.1651422}
}