We characterize the scaling laws of many-to-one dense wireless sensor networks and derive conditions governing the observability of random fields by such networks. We further extend our results to the multimodal case where the sensors observe multiple random processes simultaneously. Quite interestingly, our results show that an unbounded number of spatially bandlimited Gaussian processes can be observed simultaneously by a dense multimodal wireless sensor network.