A global systematic review of the remote sensing vegetation indices

Abstract

Vegetation indices (VIs), with the advantages of being easy to understand, simple form, and robust, have emerged as a pivotal and widespread tool for monitoring and assessing vegetation health and dynamics. Decades of research have produced numerous VIs, broadening their use and impact across various fields, but possibly overwhelming users with too many options. This study conducted a bibliometric review of VI-related literature in the web of science (WOS) database since 1986, examining current trends and issues in data sources, geographic areas, eco-functional areas, applications, and technical methods. It also analyzed the correlation among 86 VIs from global satellite data and assessed the sensitivity of 16 VIs to different parameters using radiative transfer model simulations at leaf and canopy scales. This review revealed that (1) VI research accelerated since 1986, particularly after 2012, largely due to the availability of earth-observing satellite data and new VIs. (2) The central concern of VI is its sensitivity to vegetation parameters, with recent interest in complex terrain effects. (3) VI is difficult to distinguish structural and spectral information. Optimization of soil-adjusted vegetation indices (OSAVI) has the highest sensitivity to leaf area index (LAI), and Sentinel-2 red edge position (S2REP) has the highest sensitivity to chlorophyll among the 16 selected VIs. Overall, VI performance depends on band selection and formula, with an ideal VI balancing sensitivity to vegetation and interference resistance. VI Selection should be tailored to user needs, focusing on relevant vegetation parameters and the study area's conditions.