Quoted from: https://ral.ucar.edu/projects/wrf_hydro/overview 

Scientists and society need a way to understand and predict how the complex components of the water cycle interact with the complexities of the landscape in order to provide data and information to address issues relating to water availability, water quality, hazards and impacts both in the short term and long term and across scales.

The Weather Research and Forecasting Model Hydrological modeling system (WRF-Hydro) was developed as a community-based, open source, model coupling framework designed to link multi-scale process models of the atmosphere and terrestrial hydrology to provide:

• An extensible multi-scale & multi-physics land-atmosphere modeling capability for conservative, coupled and uncoupled assimilation & prediction of major water cycle components such as: precipitation, soil moisture, snow pack, ground water, streamflow, and inundation

• Accurate and reliable streamflow prediction across scales (from 0-order headwater catchments to continental river basins and from minutes to seasons)

• A research modeling testbed for evaluating and improving physical process and coupling representations

WRF-Hydro model output can supply forecasters and decision makers with locations and timing of rapid river stage increase as well as the duration of high waters and inundation while accounting for landscape dynamics essential to flood risks such as land cover change as well as the control effects of infrastructure such as dams and reservoirs.

The WRF-Hydro modeling system was originally designed as a model coupling framework to facilitate easier coupling between the Weather Research and Forecasting model and components of terrestrial hydrological models.  WRF-Hydro is both a stand-alone hydrological modeling architecture as well as a coupling architecture for coupling of hydrological models with atmospheric models.  WRF-Hydro is fully-parallelized to enable its usage on clusters and high performance computing systems alike.

Like the WRF model it does not attempt to prescribe a particular or singular suite of physics but, instead, is designed to be extensible to new hydrological parameterizations.  Although it was originally designed to be used within the WRF model, it has evolved over time to possess many additional attributes as follows:

• Multi-scale functionality to permit modeling of atmospheric, land surface and hydrological processes on different spatial grids
• Modularized component model coupling interfaces for many typical terrestrial hydrological processes such as surface runoff, channel flow, lake/reservoir flow, sub-surface flow, land-atmosphere exchanges
• Parallel code development for application on commodity cluster and higher performance computing systems
• Stand-alone capabilities for hydrological prediction and research uncoupled to atmospheric models
• Efficient coupling architecture so that it can be embedded within (or coupled to) other types of Earth system models such as the NCAR Community Earth System Model (CESM) or the NASA Land Information System (LIS)
• Utilization of many standard data formats for efficient job construction and evaluation
• Pre-/post-processing workflows

The architecture is intended to significantly simplify the often laborious task of integrating, or coupling, existing and emerging hydrological models into the WRF modeling framework.  In doing so, an extensible, portable and scalable environment for hypothesis testing, sensitivity analysis, data assimilation and environmental prediction has emerged. 

The WRF-Hydro system has adopted a ‘community-based’ development processes with an open and participatory working group environment.  NCAR in collaboration with other NSF and university entities are developing a support structure for WRF-Hydro in the way of model documentation, public, online code repositories, test cases and many pre- and post-processing utilities.