Using hydrogeochemical signatures of stream water to assess pathways for rainfall events: towards a predictive model

Abstract

Thanks to its simple division into agricultural and forestry land use, the Corbeira catchment (Galicia, Spain) is used as a case study to build a predictive model using hydrogeochemical signatures. Stream data acquired under recessional flow conditions over a one year period were obtained from a sampling station near the downstream end of the catchment, and using principal component analysis, it is shown that some of the analytical parameters are covariant, and some are negatively correlated. These findings support inferences about the pathways of rainfall in the catchment. Specific signatures may be associated with the dominant hydrological source, either surface runoff or subsurface waters: additionally, the dominant land use in that part of the catchment, where the flow originated, can also be predicted. The dominant runoff shows a strong covariance between suspended solids (SS) and particulate phosphorus (PP), with a clear negative correlation with pH. Dissolved organic carbon (DOC) data are associated with this covariant set when these compounds are available in the soils in question. Dissolved phosphorus, total organic nitrogen and dissolved nitrates are also associated with the same covariant set when the runoff flows through areas of extensive agricultural use. The SS − PP covariance is less significant at lower flows. Typical base flow regimes show a significant covariance between salinity and pH, with a marked negative correlation with SS − PP set, confirming the dominance of subsurface waters in the baseflow, as expected. Seasonally divergent DOC − SS behaviour proves to be a useful tracer for rainfall regimes. The DOC trend shows a sinusoidal annual variation in amplitude, determined by the rainfall regime. As a result, flow from the catchment is dominated by surface water whenever there is synchronicity between the peaks of DOC and SS.