Biodiversity datasets with high spatial resolution are critical prerequisites for river protection and management decision-making. However, traditional biomonitoring tools mainly focus on local point-estimates, which calls for new approaches for predicting biodiversity at fine spatial scales. Here, we developed a novel approach by combining environment DNA (eDNA) with remote sensing (RS) technologies to identify the species distribution with high spatial resolution. Our data showed that eDNA-based species richness of aquatic insects had significant positive correlations with RS-based vegetation index, for example, the Green normalized difference vegetation index (GNDVI), Normalized difference red-edge 2 (NDRE2) and Anthocyanin reflectance index 2 (ARI2) were closely related to species richness (Pearson’ r >0.5). Using the gradient Boosting of regression (GBDT) and Random Forest (RF), our models could predict precisely species richness (R2=0.80) and distribution (>70% accuracy). In general, this study provides a new approach to achieve high spatial resolution prediction of species richness and distribution, which supports decision-making on aquatic biodiversity protection under climate changes and human impacts.