The rapid evolution of Earth observation satellite missions and computational methods made satellite remote sensing a foundation of state-of-the-art crop yield prediction. Therefore, the aim of this review is to analyze dominant drivers of crop yield prediction research based on satellite remote sensing, including dominant sensor types, satellite missions, crops, and specific research topics, as well as to identify present issues and research gaps. This review summarizes the bibliometric analysis of satellite-based crop yield prediction publications during 2000–2025, including 1174 articles that were indexed in the Web of Science Core Collection. Annual publication and citation trends, geographic patterns of research publications, prevalent satellite missions and sensor types, predominant crops used in research and trends in research themes were analyzed in the study. Findings show that there has been a consistent expansion of the study topic regarding publication count, with multispectral data, especially that of Sentinel-2, Landsat, and MODIS missions, being utilized in most of the literature in the field, while radar-based approaches are becoming increasingly important, providing complementary data to multispectral imagery. The review indicates a methodological shift in the models of simple regressions to machine learning, deep learning, and multi-sensor data fusion frameworks that use dense satellite imagery time series.
Plašćak et al. (Thu,) studied this question.