Using tracers to develop a parsimonious coupled flow-biogeochemistry model with reduced uncertainty
Abstract
Using combined hydrologic and tracer data can help develop models that are structurally consistent with the dominant processes of water and solute transport and storage. In this paper we present the development and application of a parsimonious conceptual rainfall-runoff model coupled with a biogeochemistry sub-routine able to simultaneously simulate streamflow and DOC dynamics in northern catchments with organic-rich soils. Conceptual model development was conditioned by isotopic and geochemical tracers which discriminate dominant water sources and by statistical time series models (differenced autoregressive moving average models with external regressors, ARIMAX) which identified important drivers on weekly stream DOC dynamics at two UK upland catchments. This strategy used temperature and streamflow as the main proxys to conceptualize DOC supply and transport. Quick near-surface and deeper groundwater runoff generation processes in combination with a DOC mass balance resulted in a coupled and low-parameter mechanistic model. Model tests showed mostly sensitive parameters and reasonable simulation results with seasonally controlled DOC supply. DOC transport is facilitated even for smaller events by saturated histosols connected to the stream network. However, during prolonged dry periods near-surface runoff generation mechanisms switch off and stream DOC is dominated by low concentration groundwaters. Furthermore, the model simulations were able to explain subtle differences in DOC dynamics between the calibration and a validation catchment mainly reflecting the distribution of saturated soils and available storage resulting in a flashier streamflow hydrograph and higher peak DOC concentrations in the larger validation catchment. We conclude that tracers and statistical time series models can successfully guide the development of parsimonious yet structurally consistent conceptual models if low parameterization is a key goal. Further, more parsimonious models provide tools for estimating future changes to DOC dynamics and loads with reduced uncertainty compared to more complex and higher parameterized models.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.H44B..06B
- Keywords:
-
- 1804 HYDROLOGY Catchment;
- 1847 HYDROLOGY Modeling