C2TSD is a novel framework that integrates temporal disentanglement and contrastive learning for superior spatiotemporal imputation.
