A Synthetic Global Spatiotemporal Sampled River Discharge Database for Different Satellite Altimetry Mission Orbits

Corresponding peer-reviewed publication

This dataset corresponds to all the RRR input and output files that were used in the study reported in:

• Sikder, Md. S., Bonnema, M., Emery, C. M., David, C. H., Lin, P., Pan, M., et al. (2021). A Synthetic Data Set Inspired by Satellite Altimetry and Impacts of Sampling on Global Spaceborne Discharge Characterization. Water Resources Research, 57(2), e2020WR029035. https://doi.org/10.1029/2020WR029035

When making use of any of the files in this dataset, please cite both the aforementioned article and the dataset herein. 

Note that this dataset makes extensive use of the river network and RAPID simulations that were produced in the following study, and the paper is gratefully acknowledged here:

• Lin, P., Pan, M., Beck, H. E., Yang, Y., Yamazaki, D., Frasson, R., et al. (2019). Global Reconstruction of Naturalized River Flows at 2.94 Million Reaches. Water Resources Research, 55(8), 6499–6516. https://doi.org/10.1029/2019WR025287

Version of record and details of this version

The version of record for this dataset (i.e. the one used in the aforementioned paper) is version V1.1 available at https://doi.org/10.5281/zenodo.4064188. This version V2.1 was produced to facilitate testing of the RRR software (https://github.com/c-h-david/rrr). Notable details regarding this version compared to V2.0 are as follows:

• The temporal sequence files (seq_TIM*.csv) of observations for regular temporal sampling now all have a sampling mean time of 0 second for every river reach instead of the previous value which corresponded to the cycle of observations (e.g. 259,200 seconds for a three-day regular temporal sampling). This allows to start sampling at the onset of each simulation instead of at the end of the first cycle. This change does impact the findings of the study.
• The sampled discharge files (Qout*.nc) where produced with an updated version of rrr_anl_spl_mod.py which now selects the time step at which a sample is retained using a slightly different approach. The update only impacts sampling results when the sampling time matches the river model output time step exactly, and is more accurate now. This change does impact the findings of the study.