Supporting Datasets produced in Allen et al. (2018) Global Estimates of River Flow Wave Travel Times and Implications for Low-Latency Satellite Data"

Supporting datasets for Allen et al. (2018) - Global Estimates of River Flow Wave Travel Times and Implications for Low-Latency Satellite Data, Geophysical Research Letters, https://doi.org/10.1002/2018GL077914

The code used to produce these data is available as a Github repository, permanently hosted on Zenodo: https://doi.org/10.5281/zenodo.1219784

Abstract

Earth-orbiting satellites provide valuable observations of upstream river conditions worldwide. These observations can be used in real-time applications like early flood warning systems and reservoir operations, provided they are made available to users with sufficient lead time. Yet, the temporal requirements for access to satellite-based river data remain uncharacterized for time-sensitive applications. Here we present a global approximation of flow wave travel time to assess the utility of existing and future low-latency/near-real-time satellite products, with an emphasis on the forthcoming SWOT satellite. We apply a kinematic wave model to a global hydrography dataset and find that global flow waves traveling at their maximum speed take a median travel time of 6, 4 and 3 days to reach their basin terminus, the next downstream city and the next downstream dam respectively. Our findings suggest that a recently-proposed ≤2-day latency for a low-latency SWOT product is potentially useful for real-time river applications.

Description of repository datasets:

1. riverPolylines.zip contains ESRI shapefile polylines of river networks with outputs from main analysis. These continental-scale shapefiles contain the following attributes for each river segment:

• "ARCID" : unique identifier for each river segment line, defined as the river reach between river junctions/heads/mouths. The first 10 attributes are taken from Andreadis et al. (2013): https://doi.org/10.5281/zenodo.61758
• "UP_CELLS" : number of upstream cells (pixels)
• "AREA" : upstream drainage area (km²)
• "DISCHARGE" : discharge (m³/s)
• "WIDTH" : mean bankfull river width (m)
• "WIDTH5" : 5th percentile confidence interval bankfull river width (m)
• "WIDTH95" : 95th percentile confidence interval bankfull river width (m)
• "DEPTH" : mean bankfull river depth (m)
• "DEPTH5" : 5th percentile bankfull river depth (m)
• "DEPTH95" : 95th percentile confidence bankfull river depth (m)
• "LENGTH_KM" : segment length (km)
• "ORIG_FID" : original ID of segment
• "ELEV_M" : lowest elevation of segment (m). Derived from HydroSHEDS 15 sec hydrologically conditioned DEM: https://hydrosheds.cr.usgs.gov/datadownload.php?reqdata=15demg 
• "POINT_X" : longitude of lowest point of segment (WGS84, decimal degrees)
• "POINT_Y" : latitude of lowest point of segment (WGS84, decimal degrees)
• "SLOPE" : average slope of segment (m/m)
• "CITY_JOINS" : an index associated with how likely a city/population center is located on the segment. Population center data from:  http://web.ornl.gov/sci/landscan/  and http://www.naturalearthdata.com/downloads/10m-cultural-vectors/10m-populated-places/ 
• "CITY_POP_M" : population of joined city (max N inhabitants) 
• "DAM_JOINSC" : an index associated with how likely a dam is located on the segment. Dam data from Global Reservoir and Dam (GRanD) Database: http://www.gwsp.org/products/grand-database.html 
• "DAM_AREA_S" : surface area of joined dam (m²)
• "DAM_CAP_MC" : volumetric capacity of joined dam (m³)
• "CELER_MPS"  : modeled river flow wave celerity (m/s)
• "PROPTIME_D" : travel time of flow wave along segment (days)
• "hBASIN" : main basin UID for the hydroBASINS dataset: http://www.hydrosheds.org/page/hydrobasins
• "GLCC" : Global Land Cover Characterization at segment centroid: https://lta.cr.usgs.gov/glcc/globdoc2_0 
• "FLOODHAZAR" : flood hazard composite index from the DFO (via NASA Sedac): http://sedac.ciesin.columbia.edu/data/set/ndh-flood-hazard-frequency-distribution
• "SWOT_TRAC_" : SWOT track density (N overpasses per orbit cycle @ segment centroid). Created using SWOTtrack SWOTtracks_sciOrbit_sept15 polygon shapefile, uploaded here.
• "UPSTR_DIST" : upstream distance to the basin outlet (km) 
• "UPSTR_TIME" : upstream flow wave travel time to the basin outlet (days)
• "CITY_UPSTR" : upstream flow wave travel time to the next downstream city (days)
• "DAM_UPSTR_" : upstream flow wave travel time to the next downstream dam (days)
• "MC_WIDTH" : mean of Monte Carlo simulated bankfull widths (m)
• "MC_DEPTH" : mean of Monte Carlo simulated bankfull depths (m)
• "MC_LENCOR" : mean of Monte Carlo simulated river length correction (km)
• "MC_LENGTH" : mean of Monte Carlo simulated river length (m)
• "MC_SLOPE" : mean of Monte Carlo simulated river slope (-)
• "MC_ZSLOPE" : mean of Monte Carlo simulated minimum slope threshold (m)
• "MC_N" : mean of Monte Carlo simulated Manning’s n (s/m^(1/3))
• "CONTINENT" : integer indicating the HydroSHEDS region of shapefile

2. hydrosheds_connectivity.zip contains network connectivity CSVs for river polyline shapefiles. The tables do not contain headers:

• Col1: segment unique identifier (UID) corresponding to the ARCID column of the riverPolylines shapefiles
• Col2: Downstream UID
• Col3: Number of upstream UIDs
• Col4 – Col12: Upstream UIDs

3. SWOTtracks_sciOrbit_sept15_density.zip contains a polygon shapefile derived from SWOTtracks_sciOrbit_sept15_completeOrbit containing the sampling frequency of SWOT (number of observations per complete orbit cycle). Polygon attributes correspond to each unique shape formed from overlapping swaths:

• FID : unique identifier of each polygon
• CENTROID_X : polygon centroid longitude (WGS84 - decimal degrees)
• CENTROID_Y : polygon centroid latitude (WGS84 - decimal degrees)
• COUNT_count: SWOT sampling frequency (N observations per complete orbit cycle)

4. USGS_gauge_site_information.csv : table containing the list of USGS sites analyzed in the validation and obtained from http://nwis.waterdata.usgs.gov/nwis/dv Header descriptions contained within table. 

5. validation_gaugeBasedCelerity.zip contains polyline ESRI shapefiles covering North and Central America, where USGS gauges provided gauge-based celerity estimates. These files have FIDs and attributes corresponding to riverPolylines shapefiles described above and also contrain the folllowing fields:

• GAUGE_JOIN : an index associated with how likely a gauge is located on the segment. Gauge location information is contained in USGS_gauge_site_information.csv
• GAUGE_SITE: USGS gauge site number of joined gauge
• GAUGE_HUC8: which hydrological unit code the gauge is located in
• OBS_CEL_R: gauge-based correlation score (R). Upstream and downstream gauges were compared via lagged cross correlation analysis. The calculated celerity between the paired gauges were assigned to each segment between the two gauges. If there were multiple pairs of upstream and downstream gauges, the the mean celerity value was assigned, weighted by the quality of the correlation, R. Same weighted mean was applied in assigning R. 
• OBS_CEL_MPS: gauge-based celerity estimate (m/s). 

6. tab1_latencies.csv contains data shown in Table 1 of the manuscript.

7. figS3S4_monteCarloSim_global_runMeans.csv contains the mean of the Monte Carlo simulation inputs and outputs shown in Figure S3 and Figure S4. Column headers descriptions are given in riverPolylines (dataset #1 above). Some columns have rows with all the same value because these variables did not vary between ensemble runs.

8. figS5_travelTimeEnsembleHistograms.zip contains data shown in Figure S5. Each csv corresponds to a figure component:

• tabdTT_b.csv : basin outlet travel times for all rivers
• tabdTT_b_swot.csv : basin outlet travel times for SWOT
• tabdTT_c.csv : next downstream city travel times for all rivers
• tabdTT_c_swot.csv : next downstream city travel times for SWOT
• tabdTT_d.csv : next downstream dam travel times for all rivers
• tabdTT_d_swot.csv : next downstream dam travel times for SWOT