Catchment clustering according to dominant flood generation processes

Background and aim

Clusters of similar catchments are useful in model evaluation, climate impact assessments and to improve hydrological process understanding. While several studies have tried to group catchments into hydrologically meaningful groups using a variety of catchment, climate, and flow indices in different regions of the world (e.g. Jehn et al. 2020; Brunner et al. 2020), a consistent methodology to cluster catchments according to flood generation mechanisms is missing. Therefore, this thesis aims to define clusters of catchments with similar flood generation processes using large-sample datasets in the U.S. (and possibly Europe).

Data and Methods

The process-based cluster analysis relies on large-sample streamflow and climate datasets. Floods are identified using a peak-over-threshold approach for a selection of nearly natural catchments in the U.S. (GAUGES II; Falcone 2011). Each flood event is classified by adapting and using the flood type classification scheme by Stein et al. (2019) and rastered climate data from ERA5-Land (ECMWF 2019). Such classification enables deriving a flood-type distribution for each catchment in the dataset (i.e. relative frequency of each flood type). These flood-type distributions are used to compute a similarity measure, which is subsequently used in a hierarchical clustering algorithm to derive flood-generation clusters, i.e. catchments with similar flood-type distributions. The clusters are mapped and evaluated in terms of catchment and flood characteristics such as magnitude, frequency, ratio of annual peak to flow mean, etc. (i.e. characteristics not used for the clustering). The resulting clusters may later be used to evaluate process-representation in hydrological models.

Challenges

You will improve your data management, visualization, statistical and R-programming skills by working with large-smaple datasets, producing informative maps, performing statistical/machine learning analyses, and adapting existing and developing new R-code.

Supervision

Dr. Manuela Brunner (University of Freiburg, manuela.brunner@hydrology.uni-freiburg.de)

Contact

Manuela Brunner manuela.brunner@hydrology.uni-freiburg.de

Language

English or German

Literature

1. Brunner, M. I., Gilleland, E., Wood, A., Swain, D. L., & Clark, M. (2020). Spatial dependence of floods shaped by spatiotemporal variations in meteorological and land-surface processes. Geophysical Research Letters, 47, e2020GL088000. https://doi.org/10.1029/2020GL088000

2. Brunner, M. I., Newman, A., Melsen, L. A., & Wood, A. (2020). Future streamflow regime changes in the United States: assessment using functional classification. Hydrology and Earth System Sciences, 24, 3951–3966. https://doi.org/10.5194/hess-24-3951-2020

3. ECMWF. (2019). ERA5-Land hourly data from 1981 to present (p. https://cds.climate.copernicus.eu/cdsapp#!/dataset). Copernicus. https://doi.org/10.24381/cds.e2161bac 4. Falcone, J. A. (2011). GAGES-II: Geospatial attributes of gages for evaluating streamflow. Digital spatial data set 2011. GAGES-II. https://water.usgs.gov/GIS/metadata/usgswrd/XML/gagesII_Sept2011.xml

5. Jehn, F. U., Bestian, K., Breuer, L., Kraft, P., & Houska, T. (2020). Using hydrological and climatic catchment clusters to explore drivers of catchment behavior. Hydrology and Earth System Sciences, 24(3), 1081–1100. https://doi.org/10.5194/hess-24-1081-2020

6. Stein, L., Pianosi, F., & Woods, R. (2019). Global flood process classification event-based classification for global study of river flood generating processes. Hydrological Processes, 1–16. https://doi.org/10.1002/hyp.13678

7. Tarasova, L., Merz, R., Kiss, A., Basso, S., Blöschl, G., Merz, B., Viglione, A., Plötner, S., Guse, B., Schumann, A., Fischer, S., Ahrens, B., Anwar, F., Bárdossy, A., Bühler, P., Haberlandt, U., Kreibich, H., Krug, A., Lun, D., … Wietzke, L. (2019). Causative classification of river flood events. Wiley Interdisciplinary Reviews: Water, e1353. https://doi.org/10.1002/wat2.1353