An extended data set of river discharges for validation of general circulation models

Abstract

Using existing measurements of streamflow data, we present a new data set of fiver discharges to the ocean, for use in validating climate models. This data set includes annual mean discharges for 981 rivers, representing approximately 824,500 m3/s of global river discharge. This is about 65% of contemporary estimates of continental precipitation minus continental evaporation (P-E). Using our new data set, we present a self-consistent water budget over continents. We find that rivers with annual mean discharges between about 250 m3/s and about 20,000 m3/s closely follow a power law size distribution. We estimate the total flow from rivers smaller than 250 m3/s by assuming they obey the same power law size distribution. Under this assumption, the estimated total flow from these rivers is about 33% of P-E. Including an existing estimate of soil moisture entering the ocean underground, our estimated total flow into the ocean is within about 2% of contemporary estimates of continental P-E. Our data set is available in digital format. flux of fresh water into the oceans affects the ocean thermohaline circulation. Land-surface hydrology partially controls the rates of growth and decomposition of most vegetation species, and hence affects the global carbon cycle. Despite its importance, large-scale hydrology is typically treated with less sophistication than other aspects of global climate models. One reason for this is that validation of largescale land-surface hydrologic calculations is difficult. Most of the physical processes being modeled (evapotranspiration, soil moisture content, surface runoff, etc.) are extremely heterogeneous, and hence are difficult to measure on the large scale; thus, it is difficult to determine if model calculations agree with the observed hydrologic properties.The hydrologic cycle in climate models can be validated in part by using observed river discharge data. Computed continental river discharge can be compared to measurements of river discharge at gauge stations near significant river outlets. If the simulated flows agree closely with the observed river discharges, then confidence in the simulated hydrologic system is increased.

Swept under the carpet: organic matter burial decreases global ocean biogeochemical model sensitivity to remineralization length scale

“…13. Distribution of river runoff (as fraction of total runoff, coloured boxes), as calculated from Perry et al (1996) for rivers south of 60…”

“…To account for mass preservation on long timescales, we re-supply it again via river runoff, using a subset of river runoff data of Table 2 of Perry et al (1996), i.e., the volumetric flow rates of the largest rivers in the world. We did not use direct P-or N-discharge rates, because these may exhibit a strong anthropogenic influence, e.g., due to fertilizers, and therefore may not correspond to the steady-state system envisaged here.…”

Nitrous oxide in the North Atlantic Ocean

“…The productivity in the eastern basin is fueled not only by coastal upwelling (Signorini et al, 1999) but also by dust deposition off the West African coast (Mills et al, 2004). Moreover, nutrient input by major tropical rivers such as the Senegal, Gambia and Niger (Perry et al, 1996) contribute to enhanced production off the West African coast, indicated by enhanced chlorophyll a concentrations (for the 2002 seasonal cycle of chlorophyll a see monthly data set of Sea-viewing Wide Field-of-view Sensor (SeaWiFS): http://earthobservatory.nasa.gov/Observatory/Datasets/ chlor.seawifs.html).…”

Reconstruction of land-to-sea radionuclide flux by POSEIDON/RODOS model chain: The Thermaikos Gulf base

“…So, according to Poulos et al [4] the mean annual value is 158 m 3 /s whereas this parameter was estimated to be 135 or 146 m 3 /s in Ref. [9].…”

“…Using this estimate and mean annual discharge one could estimate mean turbidity as 5.8 kg/m 3 (6.8 kg/m 3 using discharge data from [9]). Data on surface contamination by radio-caesium for the Greek territory of the Axios watershed are available in the "Atlas" [10] whereas no published data for FYROM territory were found.…”