Calibration of an early diagenesis model for high nitrate, low reactive sediments in a temperate latitude estuary (Great Ouse, UK)

The description and calibration of a reaction-diffusion model of early diagenesis are presented. Unlike previous models it has been developed for a temperate latitude estuary (Gt Ouse, UK) impacted by high nitrate concentrations (annual mean 700 μM). Five variables, O 2 , NO 3 - , NH 4 + , SO 4 2- and S 2- , are modelled from the steady state distributions of bulk total organic carbon (TOC) (i.e. a 1-G model). Three methods for deriving the first order rate constant, k, for TOC mineralisation are tested: (1) data calculated k values [i.e. (depth integrated total mineralisation rate) + (depth integrated TOC inventory)]; (2) an exponential formulation, k z = k 0 e -αz (k 0 = k at sediment surface, a = reactivity coefficient of decrease, z = depth); and (3) use of separate k values for individual mineralisation pathways. Method 1 underestimates observed fluxes of solutes across the sediment-water interface (SWI) by up to an order of magnitude. This is due to an inappropriate use of the calculated kin the model. The calculation of k yields an overall net value which implicitly accounts for all factors acting on mineralisation. Such factors (e.g. oxidant limitation of organic decay) are explicitly modelled. Consequently, k is significantly reduced by factors applied to it in the model which have previously been accounted for in the calculation. In Method 2, measured NO 3 - fluxes are overestimated by up to a factor of 7. To reproduce measured benthic oxygen demands and sulphate reduction rates, a cannot be simultaneously fitted to the NO 3 - fluxes. The high overlying NO 3 - concentrations result in model denitrification that cannot reproduce the degree of limitation that actually occurs. Method 3 reproduces the data (i.e. both stoichiometrically derived mineralisation rates and measured solute fluxes at the SWI) to a high degree (r > 0.99, p < 0.001), but at the expense of increasing the degrees of freedom in the model and conceptual simplicity. These results cast doubt over the universal applicability of diagenetic models for estuarine systems exposed to high NO 3 - concentrations. It is concluded that the use of commonly calculated first order rate constants (Method 1) and the frequently used exponential function (Method 2) in diagenetic models cannot be relied upon to reproduce observations in high NO 3 - estuaries. Previous stoichiometric calculations suggested that all of the measured ammonium fluxes across the SWI in the Gt Ouse could be accounted for with oxygen, nitrate and sulphate reduction alone. With these latter processes the model (Method 3) underestimates the observed ammonium fluxes by up to 44 % at 3 out of 4 sites. This suggests that other mineralisation pathways (e.g. nitrate ammonification) are active in the Great Ouse sediments.