Abstract: Research on oceanic geothermal heat flow helps us understand distribution of marine thermal resources, regional heat loss, hydrothermal circulation, and thermal evolution of the lithosphere. Traditional oceanic plate cooling models show biases when predicting observed heat flow in both young and old oceanic crust, and also lack sufficient quantitative description on regional thermal anomalies or hydrothermal circulation near mid-ocean ridges. In this study, we collected and organized a large amount of geological and geophysical data related to geothermal heat flow, as used a power-law model based on fractal density to describe the general trend of oceanic heat flow. Then, we applied a similarity-based method to predict global heat flow at higher spatial resolution. By combining the predicted heat flow with the power-law model, we can more fully and quantitatively identify regions with anomalous oceanic heat flow. Our results show that the power-law model performs better than traditional plate cooling models and fits observed data more accurately. Areas where the predicted heat flow is lower than the power-law model may indicate the influence of hydrothermal circulation at mid-ocean ridges. Moreover, anomalous heat flow regions show varying degrees of anisotropy. These findings challenge previous interpretations that attributed discrepancies between traditional cooling models and observations solely to hydrothermal circulation. This study improves our ability to describe oceanic heat flow patterns more precisely. It provides a useful reference for future heat flow prediction and analysis, and offers new insights and support for a deeper understanding on anomalous oceanic heat flow and hydrothermal circulation.