In messamat/globalIRmap: What the Package Does (One Line, Title Case)

Global prevalence of non-perennial rivers and streams

Authors

Mathis Loïc Messager ([e-mail])
Bernhard Lehner ([e-mail])
Charlotte Cockburn
Nicolas Lamouroux
Hervé Pella
Ton Snelder
Klement Tockner
Tim Trautmann
Caitlin Watt
Thibault Datry ([e-mail])

Contents

This repository contains the research compendium for the article: Messager, M. L., Lehner, B., Cockburn, C., Lamouroux, N., Pella, H., Snelder, T., Tockner, K., Trautmann, T., Watt, C. & Datry, T. (2021). Global prevalence of non-perennial rivers and streams. Nature. https://doi.org/10.1038/s41586-021-03565-5 \cr

Abstract

In this study, we developed a statistical Random Forest model to produce the first reach-scale estimate of the global distribution of non-perennial rivers and streams. For this purpose, we linked quality-checked observed streamflow data from 5,615 gauging stations (on 4,428 perennial and 1,187 non-perennial reaches) with 113 candidate environmental predictors available globally. Predictors included variables describing climate, physiography, land cover, soil, geology, and groundwater as well as estimates of long-term naturalised (i.e., without anthropogenic water use in the form of abstractions or impoundments) mean monthly and mean annual flow (MAF), derived from a global hydrological model (WaterGAP 2.2; Müller Schmied et al. 2014). Following model training and validation, we predicted the probability of flow intermittence for all river reaches in the RiverATLAS database (Linke et al. 2019), a digital representation of the global river network at high spatial resolution.

Accessing study data

The data repository for this study includes two datasets:

• A geometric network of the global river system where each river segment is associated with:
• 113 hydro-environmental predictors used in model development and predictions, and
• the probability and class of flow intermittence predicted by the model.
• The point locations of the 5,516 gauging stations used in model training/testing, where each station is associated with a line segment representing a reach in the river network, and a set of metadata.

How to use this compendium

The main purpose of this compendium is to provide guidance for reproducing the analysis in the manuscript.

The License tab details the terms of use of the code associated with this study.\cr

The Data tab focuses on our treatment of the two main sources of data for this project:

• a digital representation of the global river network (RiverATLAS)
• streamflow gauging station records at 5615 locations globally

The Workflow tab explains the requirements and analytical steps and provides guidelines to reproduce/understand the analysis for this study.\cr

All of the source code that generated the datasets, statistical results and figures contained in the manuscript is on three GitHub repositories: 1. globalIRmap_HydroATLAS_py: Python code used in computing new global river network hydro-environmental attributes.
2. globalIRmap_py: Python code used in processing all spatial data aside from global river network hydro-environmental attributes.
3. globalIRmap: R code used in statistical analysis, reporting, and figure production.

Acknowledgements

• Funding for this study was provided in part by the Natural Sciences and Engineering Research Council of Canada (B.L., C.C., C.W., M.L.M., NSERC Discovery grants RGPIN/341992-2013 and RGPIN/04541-2019); McGill University (M.L.M., Tomlinson Fellowship), Montreal, Quebec, Canada; H2O’Lyon Doctoral School (M.L.M., Doctoral Fellowship, ANR-17-EURE-0018), Lyon, France; T.D., N.L., H.P. and T.T. were supported by the DRYvER project (http://www.dryver.eu/), which has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 869226.
• We thank T. Elrick and the Geographic Information Centre at McGill University for providing us with high-performance computing resources.
• We thank the Global Runoff Data Centre (GRDC) for providing us with global streamflow gauging data.
• The structure of this compendium is based on the work of Patrick Schratz, Carl Boettiger, Ben Marwick, to whom we are grateful, and the workflowr package.

Comments and issues

To report any problem or ask questions on this website, please use the Github issues tracker.
For other inquiries on the article or the data, please email [e-mail]

messamat/globalIRmap documentation built on July 4, 2021, 10:48 a.m.