1. History

mizuRoute was first developed in the early 2010s, when catchment- and vector-based representations were emerging as an important alternative to regular grid discretizations for large-domain hydrologic modeling. At the time, most large-domain routing models were designed for gridded river networks because they coupled naturally with gridded land surface models. These approaches required the routing network to be regenerated whenever the computational grid changed, and the resulting river networks did not faithfully represent channel geometry, drainage areas, or river lengths, particularly at coarser resolutions. The goal of mizuRoute was to provide a routing model that could operate directly on vector river networks while remaining compatible with the gridded workflows used by existing land surface models.

The origins of mizuRoute trace back to the river routing component of the TopNet hydrologic model (Bandaragoda et al., 2002, Clark et al., 2008), which was developed at the National Institute of Water and Atmospheric Research (NIWA) in New Zealand (now part of Earth Sciences New Zealand). The routing component implemented the kinematic wave-tracking algorithm developed by Derek Goring (Goring, 1994). As part of the TopNet modernization documented by Clark et al. (2008), the original kinematic wave-tracking implementation was rewritten in Fortran 95 to improve modularity and use derived data types to organize river-network and model-state information. This modernized implementation provided the software foundation for the subsequent development of mizuRoute.

The initial development of mizuRoute involved extracting the modernized kinematic wave-tracking (KWT) routing component from TopNet and further refactoring it into a standalone routing model (Mizukami et al., 2016). The principal objective was to generalize the wave-tracking algorithm from a component embedded within a catchment model into a routing system capable of operating over large, vector-based river networks. A second objective was to provide an implementation of the impulse response function (IRF) routing scheme of Lohmann et al. (1996), which was widely used in continental-domain hydrologic prediction studies, including CMIP3- and CMIP5-driven streamflow projection studies in the United States. Supporting both the KWT and IRF routing schemes allowed mizuRoute to to apply the same routing methods used in existing prediction workflows based on the Lohmann IRF approach, while also providing the wave-tracking algorithm inherited from TopNet. The first application of mizuRoute used the U.S. Geological Survey Geospatial Fabric for the contiguous United States, demonstrating the scalability of the approach to continental domains (Mizukami et al., 2016).

As preparations began for the first mizuRoute publication, the routing model required a distinct name. During discussions within our research group, Andy Wood suggested mizuRoute, observing that the surname of the lead developer, Naoki Mizukami, naturally divides into two parts, with the first part, Mizu (水 in kanji), being the Japanese word for “water.” The suggestion was immediately adopted by the group. An additional advantage of the name is that it is concise, memorable, and not an acronym, avoiding the long and often difficult-to-remember names that are common among hydrologic models.

Since its initial release, mizuRoute has undergone several major rounds of development that have substantially expanded its computational capabilities, physical realism, and applicability. To support applications over increasingly large, high-resolution river networks, Martyn Clark redesigned the river-network connectivity algorithms, removing a key computational bottleneck and substantially improving scalability. These improvements enabled mizuRoute to be applied efficiently over continental and global domains.

A second major milestone was the development of a hierarchical river-network decomposition algorithm for hybrid parallel computing (Mizukami et al., 2021). Unlike structured atmospheric and ocean grids, river networks form complex branching trees that cannot be partitioned using conventional domain decomposition methods. The new algorithm decomposed river networks into hydrologically independent tributary domains and nested subdomains, enabling efficient hybrid MPI/OpenMP parallelization. This innovation allowed mizuRoute to scale to global vector river networks, making it practical for large ensemble simulations and coupling with Earth system models.

Subsequent development expanded the scope of mizuRoute beyond river routing to include the explicit simulation of lakes and reservoirs within vector river networks. Between 2020 and 2022, Shervan Gharari (University of Saskatchewan) developed mizuRoute-Lake (Gharari et al., 2024), which provided a flexible framework to represent both natural lakes and managed reservoirs. Rather than relying on a single lake parameterization, mizuRoute-Lake allows different water balance models (including the Döll, HYPE, and Hanasaki formulations) to be selected for individual water bodies within the same simulation. Building on this work, Inne Vanderkelen (Vrije Universiteit Brussel) collaborated with Gharari to investigate the influence of natural lakes and managed reservoirs on continental and global hydrology, demonstrating the importance of representing inland surface waters for large-domain hydrologic and Earth system simulations (Vanderkelen et al., 2022).

More recently, mizuRoute has become an integral component of reproducible, modular hydrologic modeling workflows. mizuRoute provides the river-routing component in standardized workflows developed for SUMMA and large-domain hydrologic prediction (Knoben et al., 2022, Farahani et al., 2025, Tang et al., 2025). These workflows have enabled reproducible applications spanning local catchments to continental and global river networks.

References

Bandaragoda, C., Tarboton, D.G. and Woods, R. (2004). Application of TOPNET in the distributed model intercomparison project. Journal of Hydrology, 298(1-4), pp.178-201. https://doi.org/10.1016/j.jhydrol.2004.03.038

Clark, M.P., Rupp, D.E., Woods, R.A., Zheng, X., Ibbitt, R.P., Slater, A.G., Schmidt, J. and Uddstrom, M.J. (2008). Hydrological data assimilation with the ensemble Kalman filter: Use of streamflow observations to update states in a distributed hydrological model. Advances in water resources, 31(10), pp.1309-1324. https://doi.org/10.1016/j.advwatres.2008.06.005

Farahani, M.A., Wood, A.W., Tang, G. and Mizukami, N. (2025). Calibrating a large-domain land/hydrology process model in the age of AI: the SUMMA CAMELS emulator experiments. Hydrology and Earth System Sciences, 29(18), pp.4515-4537. https://doi.org/10.5194/hess-29-4515-2025

Gharari, S., Vanderkelen, I., Tefs, A., Mizukami, N., Kluzek, E., Stadnyk, T., Lawrence, D. and Clark, M.P. (2024). A flexible framework for simulating the water balance of lakes and reservoirs from local to global scales: mizuRoute‐Lake. Water Resources Research, 60(5), p.e2022WR032400. https://doi.org/10.1029/2022WR032400

Goring, D. G. (1994) Kinematic shocks and monoclinal waves in the Waimakariri, a steep, braided, gravel-bed river, Proceedings of the International Symposium on waves: Physical and numerical modelling, University of British Columbia, Vancouver, Canada, 336–345.

Knoben, W.J.M., Clark, M.P., Bales, J., Bennett, A., Gharari, S., Marsh, C.B., Nijssen, B., Pietroniro, A., Spiteri, R.J., Tang, G., Tarboton, D.G., and Wood, A.W. (2022). Community workflows to advance reproducibility in hydrologic modeling: Separating model‐agnostic and model‐specific configuration steps in applications of large‐domain hydrologic models. Water Resources Research, 58(11), p.e2021WR031753. https://doi.org/10.1029/2021WR031753

Lohmann, D., Nolte-Holube, R. and Raschke, E (1996) A large-scale horizontal routing model to be coupled to land surface parametrization schemes, Tellus A, 48: 708-721 https://doi.org/10.1034/j.1600-0870.1996.t01-3-00009.x

Mizukami, N., Clark, M.P., Sampson, K., Nijssen, B., Mao, Y., McMillan, H., Viger, R.J., Markstrom, S.L., Hay, L.E., Woods, R. Arnold, J.R., and Brekke, L.D. (2016). mizuRoute version 1: A river network routing tool for continental domain water resources applications. Geoscientific Model Development, 9(6), pp.2223-2238. https://doi.org/doi:10.5194/gmd-9-2223-2016

Mizukami, N., Clark, M.P., Gharari, S., Kluzek, E., Pan, M., Lin, P., Beck, H.E. and Yamazaki, D. (2021). A vector‐based river routing model for Earth system models: Parallelization and global applications. Journal of Advances in Modeling Earth Systems, 13(6), p.e2020MS002434. https://doi.org/10.1029/2020MS002434

Tang, G., Clark, M.P., Knoben, W.J., Liu, H., Gharari, S., Arnal, L., Wood, A.W., Newman, A.J., Freer, J. and Papalexiou, S.M. (2025). Uncertainty hotspots in global hydrologic modeling: the impact of precipitation and temperature forcings. Bulletin of the American Meteorological Society, 106(1), pp.E146-E166. https://doi.org/10.1175/BAMS-D-24-0007.1

Vanderkelen, I., Gharari, S., Mizukami, N., Clark, M.P., Lawrence, D.M., Swenson, S., Pokhrel, Y., Hanasaki, N., Van Griensven, A. and Thiery, W., 2022. Evaluating a reservoir parametrization in the vector-based global routing model mizuRoute (v2. 0.1) for Earth system model coupling. Geoscientific Model Development, 15(10), pp.4163-4192. https://doi.org/10.5194/gmd-15-4163-2022

Last updated on July 25th, 2026