Purpose <p>Sediment contamination by heavy metals, and nutrients has been evident in the riverine ecosystems, while their source apportionment, and spatial distribution are complicated to understand. This study aimed to identify the sources, and spatial heterogeneity of heavy metals and nutrients through the adoption of integration approaches that could analyze their patterns, improve understanding of dynamics, and provide insights to enhance decision-making in conserving rivers.</p> Materials and methods <p>Surface sediment samples (<i>n</i> = 184<i>)</i> were collected and analyzed from the eight different rivers in the Great Chao Phraya River network, Thailand. Principal Component Analysis (PCA) was employed in identifying the sources, while its integration with Inverse Distance Weighting (IDW), and Geographically Weighted Regression (GWR) were executed in exploring the spatial heterogeneity of contaminants. A large pool of explanatory variables, including natural and anthropogenic factors were utilized in the geospatial regression using a multiple GWR model.</p> Results and discussion <p>The PCA revealed that Cr, Ni, and Cu (partly) were potentially sourced from natural/lithogenic origin; Cu, Cd, Zn, and Pb from anthropogenic sources; Hg and As from mining and industries; and the nutrients (TC, TN and TP) from domestic activities and agriculture. An integration approach of PCA with IDW showed that the loadings of the sediment-associated heavy metals, and nutrients increased towards the downstream (0 ~ 4), which aligns with the degree of anthropogenic influences. The local explanatory power of GWR was highest for PC4 (&lt; 0.4), and PC1 (&lt; 0.37), followed by PC2 (&lt; 0.21) and PC3 (&lt; 0.2). Clear spatial patterns of higher local R² values in the lower Chao Phraya were detected, except for PC4 (Hg, As). Although comparatively lower local R² values were noted, the random distribution of the local residuals underscored the good performance of the GWR model. The linear relationships of the local coefficients along the urbanization gradient demonstrate that urban land use is a dominant factor impacting sediment quality.</p> Conclusions <p>This study illustrated that the integration of multivariate analysis, and geospatial approaches provides meaningful insights into understanding the sources, spatial dynamics of sediment contaminants, and the influencing factors. Moreover, the higher performance of the GWR model with increasing levels of urbanization possibly implied the better suitability of using the GWR model in developed watersheds. Future studies should test the different integration approaches, which could offer valuable insights into understanding the spatial characteristics of sediment quality.</p>