We all know that most of the hydrological variables are time-dependent. It changes with time.
Artificial Neural Network on the other hand has been found to perform best for time-dependent problems.
Whether there is any relationship between time-dependent variables and neural network is unknown but it can be assured that neural networks can be used to develop models for the prediction of hydrological variables with better accuracy compared to other hydrological models.
The post below depicts the methodology for the application of neural networks in the development of hydrological models.
This is reposted from my Water and Energy Nexus Blog :
Mrinmoy
My ResearchGate Id : Mrinmoy_Majumder
Home Page: http://www.mrinmoymajumder.com
Author of: Lecture Notes on MCDM