Woreket et al. (2026) Remote sensing for estimating crop water productivity: a systematic review of concepts and methods

Identification

• Journal: Modeling Earth Systems and Environment
• Year: 2026
• Date: 2026-03-03
• Authors: Wubalem Woreket, Solomon Seyoum, Marloes Mul, Gebeyehu Abebe
• DOI: 10.1007/s40808-026-02755-2

Research Groups

• Space Science and Geospatial Institute, Addis Ababa, Ethiopia
• IHE Delft Institute for Water Education, Delft, Netherlands
• Debre Berhan University, Debre Birhan, Ethiopia

Short Summary

This systematic review synthesizes 93 studies (2020-2025) to critically examine remote sensing concepts and methods for estimating Crop Water Productivity (CWP) by analyzing approaches for crop yield and actual evapotranspiration (ETa), aiming to provide a consolidated reference for advancing CWP assessment.

Objective

• Clarify the conceptual foundations and significance of Crop Water Productivity (CWP).
• Summarize major remote sensing approaches used to estimate actual evapotranspiration (ETa) and crop yield, the two core components of CWP.
• Assess existing validation strategies for remote sensing-based CWP estimation.
• Propose future research directions to improve the reliability and scalability of CWP estimation.

Study Configuration

• Spatial Scale: Global, with a pronounced concentration of reviewed studies from China, India, and the USA; methods discussed are applicable from field to continental scales.
• Temporal Scale: Review period of studies: 2020–2025.

Methodology and Data

• Models used:
  • Crop Yield Estimation: Empirical models (statistical, machine learning, deep learning like CNNs, RNNs), Light Use Efficiency (LUE) models (GPP, NPP, AGB), Process-based crop growth models with Data Assimilation (e.g., AquaCrop, WOFOST, DSSAT, DAISY, APSIM).
  • Actual Evapotranspiration (ETa) Estimation: Surface Energy Balance (SEB) models (e.g., SEBAL, SEBS, SSEBop, S-SEBI, METRIC, STIC, TSEB, ALEXI, ETLook, HTEM, TTME), Semi-Empirical Formula (SEF) methods (e.g., Priestley-Taylor, PT-JPL, Ts-VI), Statistical Regression (SR) approaches (e.g., machine learning algorithms).
• Data sources:
  • Remote Sensing: Optical sensors (e.g., Sentinel-2, Landsat series, MODIS), Thermal infrared sensors (e.g., Landsat TIRS, MODIS thermal bands, VIIRS, Sentinel-3), Radar sensors (e.g., Sentinel-1), Unmanned Aerial Vehicle (UAV) data.
  • Meteorological Data: Ground-based weather stations, global precipitation datasets (e.g., CHIRPS), reanalysis data (e.g., ERA-5).
  • Validation Data: Eddy Covariance (EC) systems, lysimeters, Bowen ratio, scintillometer method (for ETa); Harvest/crop-cutting data, farm records/official statistics, plot-based sample surveys (for yield).

Main Results

• Crop Water Productivity (CWP) is a crucial metric for water use efficiency, derived from crop yield and actual evapotranspiration (ETa).
• Remote sensing is indispensable for CWP estimation, offering spatially explicit, multi-temporal, and cost-effective observations.
• For crop yield estimation, empirical models (59% of reviewed studies) are most common due to simplicity and scalability, but their accuracy is variable and calibration-dependent. Light Use Efficiency (LUE) models (22.6%) offer moderate accuracy with more biophysical realism, while process-based crop growth models with data assimilation (14%) provide the highest accuracy and physiological realism but are complex and data-intensive.
• For ETa estimation, Surface Energy Balance (SEB) models (e.g., SEBAL, METRIC) consistently achieve high accuracy but are complex and require thermal imagery. Semi-Empirical Formula (SEF) methods offer a balance of accuracy and computational efficiency, while Statistical Regression (SR) approaches are simple and scalable but generally less accurate and generalizable.
• A "trilemma" exists in CWP estimation, highlighting trade-offs between accuracy, complexity, and scalability, with no single method optimizing all three.
• Validation remains a critical bottleneck, relying on resource-intensive ground-truthing (e.g., Eddy Covariance, lysimeters for ETa; harvest/crop-cutting for yield), which is often sparse.
• Major crop types investigated include wheat, sugarcane, rice, maize, soybean, and cotton.

Contributions

• Provides a clear conceptual synthesis of CWP, emphasizing its relevance to sustainable agriculture and water resource management.
• Offers a comprehensive evaluation of remote sensing approaches for estimating ETa and crop yield, detailing the parameters, assumptions, and methodological requirements of each technique.
• Proposes a structured set of future research directions, including multisource data integration, advanced image preprocessing, integration of active/radar sensors, and improved ground-based validation protocols.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Citation

@article{Woreket2026Remote,
  author = {Woreket, Wubalem and Seyoum, Solomon and Mul, Marloes and Abebe, Gebeyehu},
  title = {Remote sensing for estimating crop water productivity: a systematic review of concepts and methods},
  journal = {Modeling Earth Systems and Environment},
  year = {2026},
  doi = {10.1007/s40808-026-02755-2},
  url = {https://doi.org/10.1007/s40808-026-02755-2}
}