Global Maps of Streamflow Characteristics Based on Observations from Several Thousand Catchments*

Abstract

Streamflow Q estimation in ungauged catchments is one of the greatest challenges facing hydrologists. Observed Q from 3000 to 4000 small-to-medium-sized catchments (10–10 000 km2) around the globe were used to train neural network ensembles to estimate Q characteristics based on climate and physiographic characteristics of the catchments. In total, 17 Q characteristics were selected, including mean annual Q, baseflow index, and a number of flow percentiles. Testing coefficients of determination for the estimation of the Q characteristics ranged from 0.55 for the baseflow recession constant to 0.93 for the Q timing. Overall, climate indices dominated among the predictors. Predictors related to soils and geology were relatively unimportant, perhaps because of their data quality. The trained neural network ensembles were subsequently applied spatially over the entire ice-free land surface, resulting in global maps of the Q characteristics (at 0.125° resolution). These maps possess several unique features: they represent observation-driven estimates, they are based on an unprecedentedly large set of catchments, and they have associated uncertainty estimates. The maps can be used for various hydrological applications, including the diagnosis of macroscale hydrological models. To demonstrate this, the produced maps were compared to equivalent maps derived from the simulated daily Q of four macroscale hydrological models, highlighting various opportunities for improvement in model Q behavior. The produced dataset is available online (http://water.jrc.ec.europa.eu/GSCD).

MSWEP: 3-hourly 0.25° global gridded precipitation (1979–2015) by merging gauge, satellite, and reanalysis data

“…Following this method, we inferred long-term P from Q records for 13 762 regions (12 233 catchments and 1529 interstation catchment areas) across the globe. Long-term Q estimates for these catchments were computed from the same three sources as those used by Beck et al (2015), namely (i) the US Geological Survey (USGS) Geospatial Attributes of Gages for Evaluating Streamflow (GAGES)-II database (Falcone et al, 2010), (ii) the Global Runoff Data Centre (GRDC; http://www.bafg.de/GRDC/), and (iii) the Australian Peel et al (2000) database. We only used catchments < 10 000 km 2 with a Q record length > 3 years (not necessarily consecutive).…”

“…With the exception of CMAP, all datasets listed in Table 1 employ only one or two of the main data sources -either gauge and satellite, or gauge and reanalysisand thus do not take full advantage of the complementary nature of satellite and reanalysis data identified in previous studies (e.g., Janowiak, 1992;Huffman et al, 1995;Arkin, 1996, 1997;Xie and Joyce, 2014;Adler et al, 2001;Ebert et al, 2007;Serrat-Capdevila et al, 2013;Peña Arancibia et al, 2013); 2. Many of the listed datasets do not explicitly and fully account for gauge under-catch and/or orographic effects, and consequently underestimate P in many regions around the globe (e.g., Zaitchik et al, 2010;Kauffeldt et al, 2013;Beck et al, 2015Beck et al, , 2016aPrein and Gobiet, 2016); 3. Many datasets (numbered 6-12 and 14-21 in Table 1) do not incorporate gauge observations or do so on a monthly basis, and hence may not make optimal use of valuable information on the daily P variability provided by gauges; Table 1.…”

Global-scale evaluation of 22 precipitation datasets using gauge observations and hydrological modeling

“…Our collection of Q observations was compiled from the same three sources as Beck et al (2015), viz.…”

“…For the calibration, we employed the (µ + λ) evolutionary algorithm (Ashlock, 2010;Fortin et al, 2012) with the population size (µ) set to 20, the recombination pool size (λ) set to 40, and the number of generations set to 12 (amounting to 480 model runs per catchment per P dataset and approximately 40 million model runs in total). See Beck et al (2016) and (2017b) for more details on the hydrological model, calibration algorithm, model parameter ranges, Q observations, E p forcing, and T a forcing. We recognize that using data from different sources may bias results as the water balances are unlikely to be closed.…”

The Global Streamflow Indices and Metadata Archive (GSIM) – Part 2: Quality control, time-series indices and homogeneity assessment

“…2013; Beck et al, 2015;Olden and Poff, 2003;Sawicz et al, 2011Sawicz et al, , 2014Westerberg et al, 2016). These hydrological signatures include e.g.…”