Hybrid imaging and modeling approach boosts accuracy of rice water stress detection
en-GBde-DEes-ESfr-FR

Hybrid imaging and modeling approach boosts accuracy of rice water stress detection

25/08/2025 TranSpread

This approach delivers improved accuracy for key indicators, enabling more reliable drought assessment and potentially supporting smarter irrigation management.

Water stress is a persistent challenge for rice production, particularly under increasingly erratic climate conditions. Accurately assessing drought impacts in rice paddies is difficult due to complex microclimates and fluctuating irrigation patterns. Rice fields managed with alternating wet and dry irrigation conserve water but create spatially and temporally variable soil moisture conditions. Traditional water stress evaluation often relies on single indicators, such as leaf water content or canopy temperature, which can overlook subtle physiological and structural differences across plant parts. While ground-based spectral measurements and vegetation indices have been widely used to monitor crop status, they can struggle with complexity and noise. Physical models such as PROSAIL simulate canopy light interactions, but alone may not capture the full phenotypic response to water stress. Conversely, data-driven approaches like machine learning can adapt to specific datasets but require large volumes of measured data. Hybrid approaches, combining physical modeling and machine learning, offer a promising path forward for accurately identifying and quantifying drought impacts.

A study (DOI: 10.1016/j.plaphe.2025.100016) published in Plant Phenomics on 28 February 2025 by Yuanyuan Zha ’s team, Wuhan University, demonstrates that integrating radiative transfer modeling with multidimensional imaging significantly improves the accuracy and sensitivity of rice water stress detection under complex field conditions.

Researchers conducted two years of experiments in rice fields with long-term and short-term water-deficit treatments, simulating real-world irrigation practices. Four comprehensive canopy-related indicators were examined, with particular focus on canopy chlorophyll content (CCC) and canopy equivalent water (CEW). These were retrieved through a hybrid method that combined PROSAIL radiative transfer model outputs with traits extracted from hyperspectral and front-view images. The process began with generating a synthetic dataset from PROSAIL to pre-train a machine learning model, followed by fine-tuning using measured data. By integrating spectral reflectance from hyperspectral imaging with phenotypic features such as leaf angle distribution from front-view images, the model achieved higher retrieval accuracy for CCC (R = 0.7920) and CEW (R = 0.8250) than conventional data-driven or physical inversion methods. Long-term drought primarily affected canopy structure and pigmentation, reducing leaf area index (LAI) during vegetative stages but sometimes increasing pigment concentration during reproductive stages. Short-term drought-rehydration cycles did not significantly alter leaf size or water content but slowed pigment decline, with CCC and CEW proving sensitive markers of these functional changes. The red-edge spectral region (742–841 nm) was identified as optimal for assessing drought impacts. Image-derived leaf angle parameters improved PROSAIL simulation accuracy by over 40%, while combining top- and front-view imaging yielded complementary insights into canopy traits.

The new method enables more robust detection of both immediate and delayed drought effects, offering practical value for precision irrigation scheduling and water-saving management in rice cultivation. CCC and CEW, in particular, integrate multiple dimensions of plant physiology—leaf area, pigmentation, and organ water content—providing a holistic view of crop status. With portable imaging devices and automated analysis pipelines, this approach could be adapted for on-farm use, allowing farmers and agronomists to monitor water stress more efficiently and intervene before yield losses occur.

###

References

DOI

10.1016/j.plaphe.2025.100016

Original Source URL

https://doi.org/10.1016/j.plaphe.2025.100016

Funding information

The research was funded by the National Natural Science Foundation of China, China (52279042), the National Key Research and Development program of China, China (2021YFC3201204), and the Key Research and Development Program in Guangxi, China (AB23026021).

About Plant Phenomics

Science Partner Journal Plant Phenomics is an online-only Open Access journal published in affiliation with the State Key Laboratory of Crop Genetics & Germplasm Enhancement, Nanjing Agricultural University (NAU) and distributed by the American Association for the Advancement of Science (AAAS). Like all partners participating in the Science Partner Journal program, Plant Phenomics is editorially independent from the Science family of journals. Editorial decisions and scientific activities pursued by the journal's Editorial Board are made independently, based on scientific merit and adhering to the highest standards for accurate and ethical promotion of science. These decisions and activities are in no way influenced by the financial support of NAU, NAU administration, or any other institutions and sponsors. The Editorial Board is solely responsible for all content published in the journal. To learn more about the Science Partner Journal program, visit the SPJ program homepage.

Title of original paper: A hybrid method for water stress evaluation of rice with the radiative transfer model and multidimensional imaging
Authors: Yufan Zhang a, Xiuliang Jin b, Liangsheng Shi a, Yu Wang a, Han Qiao a, Yuanyuan Zha a
Journal: Plant Phenomics
Original Source URL: https://doi.org/10.1016/j.plaphe.2025.100016
DOI: 10.1016/j.plaphe.2025.100016
Latest article publication date: 28 February 2025
Subject of research: Not applicable
COI statement: The authors declare that they have no competing interests.
Attached files
  • Figure 4. Pipeline of comparing different trait retrieval methods.a. Machine learning-based method using a CNN model b. PROSAIL model retrieval method using a look-up-table. c. Hybrid method.
25/08/2025 TranSpread
Regions: North America, United States, Asia, China
Keywords: Applied science, Engineering

Disclaimer: AlphaGalileo is not responsible for the accuracy of content posted to AlphaGalileo by contributing institutions or for the use of any information through the AlphaGalileo system.

Testimonials

For well over a decade, in my capacity as a researcher, broadcaster, and producer, I have relied heavily on Alphagalileo.
All of my work trips have been planned around stories that I've found on this site.
The under embargo section allows us to plan ahead and the news releases enable us to find key experts.
Going through the tailored daily updates is the best way to start the day. It's such a critical service for me and many of my colleagues.
Koula Bouloukos, Senior manager, Editorial & Production Underknown
We have used AlphaGalileo since its foundation but frankly we need it more than ever now to ensure our research news is heard across Europe, Asia and North America. As one of the UK’s leading research universities we want to continue to work with other outstanding researchers in Europe. AlphaGalileo helps us to continue to bring our research story to them and the rest of the world.
Peter Dunn, Director of Press and Media Relations at the University of Warwick
AlphaGalileo has helped us more than double our reach at SciDev.Net. The service has enabled our journalists around the world to reach the mainstream media with articles about the impact of science on people in low- and middle-income countries, leading to big increases in the number of SciDev.Net articles that have been republished.
Ben Deighton, SciDevNet

We Work Closely With...


  • e
  • The Research Council of Norway
  • SciDevNet
  • Swiss National Science Foundation
  • iesResearch
Copyright 2025 by AlphaGalileo Terms Of Use Privacy Statement