*4.4. Climate Change Scenarios*

Under climate change, the frequency, intensity as well as the incidence of extreme rainfall events are subject to change [48]. Meanwhile, temperature, surface runoff, and land use will also be influenced by the changing climate. The shift in extreme rainfall characteristics can influence the soil erosion directly by changing the erosive capacity [49]. The variance of soil moisture, temperature, and land use can affect soil erosion via affecting soil properties such as erodibility. In addition, these factors will interact in a nonlinear way to regulate soil erosion and related sediment transport processes. One of the best ways to project the rainfall-runoff and soil erosion dynamics under climate change is via numerical models that can address the complicated hydro-sedimentary dynamics. Such studies have been conducted by several scientists [35,44,45].

Over GCEW, the decrease in the cultivated land is proved to reduce the fine sediment source. In addition, such land use change can also decrease the peak flow and runoff volume, resulting in the reduction of sediment supply and transport in channels [50]. In this study, the current sediment model is built on WRF-Hydro in the expectation that it will be used to nowcast/hindcast the streamflow and soil erosion during rainfall events over GCEW. In addition, unlike the WRF-Hydro, WRF-Hydro-SED, for now, only supports simulation in serial mode. Thus, long-term simulation considering climate change scenarios is not feasible to by carried out using the current model. Future model development by parallelizing the model code, introducing a morphological evolution algorithm, and considering the rainfall and land use evolution, as well as the complex interrelation between those factors under climate change scenarios, are expected to alleviate such limitations.

#### **5. Conclusions**

In this study, by adapting the sediment algorithm from CASC2D-Sed, we introduced a sediment module into the WRF-Hydro platform, allowing for the development of a fully distributed, process-based soil erosion and sediment transport model (WRF-Hydro-Sed). The model's performance was evaluated via a comparison with the observed streamflow and sediment concentration data at the Goodwin Creek Experimental Watershed during rainfall events.

WRF-Hydro-Sed is able to generate satisfactory results of streamflow and sediment yield during rainfall events. The streamflow can be calibrated successfully based on a single rainfall event with the adjustment of a few hydro-parameters including refkdt (the parameter that controls runoff–infiltration partition) and channel geometries. With the single event calibrated hydro-parameters, the model can also perform satisfactorily in simulating the hydrograph during a validation event. Based on calibrated hydro-parameters, sediment concentration, sediment flux, as well as sediment yield can also be calibrated successfully at watershed scale by adjusting sediment parameters related to land use and soil category. Satisfactory results are also generated for a validation event using calibrated sediment parameters. The model's performance in simulating sediment yield is better than sediment concentration and flux.

The model's performance in streamflow simulation is sensitive to forcing data. The original NLDAS-2, given its 1/8 degree coarse resolution, may not be an optimal choice to provide rainfall forcing for simulation over a relatively small watershed like the Goodwin Creek under local storm events. High resolution meteorological forcing data is recommended for application of the WRF-Hydro-Sed on a small watershed.

Calibrated hydro-parameters based on a single event can be applied to different rainfall events to reproduce the hydrograph. While it might not be practical to have a set of parameters that can be suitable for any rainfall event, an intensive calibration based on multiple events can improve the model's performance to a certain degree, but with extensive computational efforts. In this case, intensive calibration over a long time scale might not be an optimal strategy if computational cost is a major concern and if the model performance based on a single event calibration is acceptable. With the calibrated sediment parameters based on a single event, the sediment yield over different events can be simulated within the same magnitude observed. Moreover, the model shows promising

potential in simulating annual soil erosion on a watershed scale. While simulated sediment yield is considered acceptable for 71% of the events (12 out of 17), substantial bias can be found during certain events mainly due to the bias transferred from the streamflow simulation. Future development of the model by including the bank and channel erosion algorithm is expected to further improve model performance.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4441/12/6/1840/s1, the equations used to calculate correlation coefficient, RMSE, NSE and KGE.

**Author Contributions:** Conceptualization, Z.G.X. and D.Y.; methodology, D.Y.; software, D.Y. and W.Y.; validation, D.Y.; formal analysis, D.Y. and M.M.; investigation, D.Y. and M.M.; resources, Z.G.X. and D.J.G.; data curation, D.Y.; writing—original draft preparation, D.Y.; writing—review and editing, D.Y., Z.G.X. and A.R.; visualization, D.Y. and Z.G.X.; supervision, Z.G.X.; project administration, Z.G.X. and D.J.G.; funding acquisition, Z.G.X. and D.J.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project was funded by The Water Institute of the Gulf under project "Project Louisiana Rivers' Sediment Flux to the Coastal Ocean using a Coupled Atmospheric-Hydrological Model" (award number RCEGR260003-01-00). This project was paid for (in part) with federal funding from the Department of the Treasury through the Louisiana Coastal Protection and Restoration Authority's Center of Excellence Research Grants Program under the Resources and Ecosystems Sustainability, Tourist Opportunities, and Revived Economies of the Gulf Coast States Act of 2012 (RESTORE Act). The statements, findings, conclusions, and recommendations are those of the author(s) and do not necessarily reflect the views of the Department of the Treasury, CPRA or The Water Institute of the Gulf. Dongxiao Yin is also supported by the Economic Development Assistantship program at LSU. Support from NCAR's Advanced Study Program's Graduate Student (GVP) Fellowship is appreciated. This research is also partially funded by USGS The Cooperative Ecosystem Studies Units (award# G20AC00099).

**Acknowledgments:** Louisiana Optical Network Initiative (LONI), and Cheyenne (doi: 10.5065/D6RX99HX) provided by NCAR's Computational and Information Systems Laboratory, sponsored by the National Science Foundation are acknowledged for high-performance computing support. We acknowledge the NCAR WRF-Hydro developing team for help with model setup. Glenn V. Wilson and Jacob Ferguson are thanked for providing us with the data of GCEW.

**Conflicts of Interest:** The authors declare no conflict of interest.
