VIsToLAI: A modular open-source platform for estimating the leaf area index from remote sensing-derived vegetation indices

VIsToLAI, a unique open-source software framework developed in Python, stands out for its ability to estimate the leaf area index (LAI) using time series data derived from various remote sensing vegetation indices (VIs). The framework integrates both empirical regression models and machine learning (ML) approaches, offering a flexible and scalable workflow for LAI estimation. Through case studies on four major staple crops—rice, barley, wheat, and maize—this study demonstrates the framework's ability to accurately estimate LAI across diverse crop types and environmental conditions. Results show that machine learning models, particularly extra trees and gradient boosting, outperform traditional empirical models in terms of accuracy and robustness, especially under heterogeneous data conditions. VIsToLAI’s modular architecture enables the easy incorporation of new indices and algorithms, as well as seamless integration into existing remote sensing workflows. The software provides a valuable tool for bridging remote sensing data with agricultural modeling, supporting precision agriculture and large-scale monitoring initiatives.