Christian Haug Eide, Reidar Müller, William Helland-Hansen

Models relating sediment-supply to catchment-properties are important in order to use the geological record to deduce landscape evolution and the interplay between tectonics and climate. Water-discharge (Qw) is an important factor in the widely used BQwART-model of Syvitski and Milliman (2007), which relates sediment load to a set of measureable catchment parameters. Although many of the factors in this equation may be independently estimated with some degree of certainty in ancient systems, water-discharge (Qw) certainly cannot. An analysis of a world database of modern catchments (Milliman and Farnsworth, 2011) shows that the commonly applied equation relating catchment area (A) to water-discharge (Qw=0.075A0.8), does not predict water-discharge from catchment area well in many cases (R2=0.5 and an error spanning 4 orders-of-magnitude). Neither does the equation incorporate the effect of arid and wet climate on this relationship. The inclusion of climate-data into such estimations is an opportunity to refine these estimates, because generalized estimates of palaeoclimate can often be deduced on the basis of sedimentological data such as palaeosol types, mineralogy and palaeohydraulics.
This paper investigates how the relationship between catchment size and river discharge vary with four runoff categories (arid, subarid, humid, and wet) which are recognizable in the geological record, and modify the coefficient and exponent of the abovementioned equation according to these classes. It follows from this analysis that water-discharge from arid catchments is so variable, that water-discharge cannot be predicted from catchment area. Our modified model yields improved results in relating discharge to catchment size (R2=0.95 and error spanning 1 order-of-magnitude) when core-, outcrop- or regional palaeoclimate reconstruction data are available in non-arid systems. In conclusion, this model, in contrast to the previous, is sufficient for many geological applications and will lead to a higher degree of confidence in the application of mass-balance models in ancient systems.