Abstract
Modern geoscientists have online access to an abundance of different data sets and models, but these resources differ from each other in myriad ways and this heterogeneity works against interoperability as well as reproducibility. The purpose of this paper is to illustrate the main issues and some best practices for addressing the challenge of reproducible science in the context of a relatively simple hydrologic modeling study for a small Arctic watershed near Fairbanks, Alaska. This study requires several different types of input data in addition to several, coupled model components. All data sets, model components and processing scripts (e.g., for preparation of data and figures, and for analysis of model output) are fully documented and made available online at persistent URLs. Similarly, all source codes for the models and scripts are open source, version controlled, and made available online via GitHub. Each model component has a Basic Model Interface to simplify coupling and its own HTML help page that includes a list of all equations and variables used. The set of all model components (TopoFlow) has also been made available as a Python package for easy installation. Three different graphical user interfaces for setting up TopoFlow runs are described, including one that allows model components to run and be coupled as web services.
| Original language | English |
|---|---|
| Pages (from-to) | 377-394 |
| Number of pages | 18 |
| Journal | Earth and Space Science |
| Volume | 4 |
| Issue number | 6 |
| DOIs | |
| State | Published - 2017 |
| Externally published | Yes |
Funding
The Institute of Arctic Biology at the University of Alaska, Fairbanks maintains a Long-Term Ecological Research (LTER) site (funded by NSF) called the Bonanza Creek LTER. The Caribou-Poker Creeks Research Watershed (CPCRW) is one of the study sites for this LTER project where long-term monitoring data are collected and made available online. Clicking on Access Data > Study Sites Catalog in the Data menu of this website brings up a search filter page for the Study Sites Catalog. Typing Caribou in the text box labeled Name, Description, History and clicking on the Submit button generates a listing of available data and a locator map. For this paper, we selected the C2 subbasin within the Caribou Creek watershed. Data for four separate subbasins of Caribou Creek are available, namely C1, C2, C3, and C4. Note that each subbasin name begins with the letter “C” for Caribou. There are four weather stations (or “met stations”) in the vicinity of Caribou Creek, designated as CARSNOW, CPEAK, CRREL, and HR1A. The longitudes and latitudes of these stations are given by We would like to thank Matt Nolan, who initiated the collaboration that led to TopoFlow, Larry Hinzman, who helped to guide and provide funding for its development, as well as the entire Arctic hydrology team at WERC (Water and Environmental Research Center), at the University of Alaska, Fairbanks. S. Peckham is the primary author of TopoFlow, all versions of which are free and open source, as well as the product RiverTools 4.0, which is sold commercially by Rivix, LLC. To avoid any perceived conflict of interest, this paper has described workflows that do not require RiverTools. This work was partially funded by NSF grants ICER 1440332 (GeoSoft), 1440333 (GeoSemantics), and PLR 1503559.
Keywords
- Python
- TopoFlow
- component-based modeling
- hydrologic model
- open source
- reproducibility
Fingerprint
Dive into the research topics of 'Reproducible, component-based modeling with TopoFlow, a spatial hydrologic modeling toolkit'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver