Water chemistry evolution through the critical zone
Water as it passes through the critical zone - from top of the trees to the bottom of the groundwater table - plays a critical role in chemical weathering of rocks and in the global carbon cycle. Although the chemistry of surface water (i.e. rivers and lakes) has been intensively monitored at high-frequency (<1 day), the temporal and spatial variability of groundwater chemistry in the critical zone, especially in the weathered bedrock zone, has been rarely observed. In many catchments located at temperate regions, nearly all rainwater infiltrates into subsurface, thus identifying the processes and environmental controls for the groundwater chemistry will be fundamental to understand the stream chemistry dynamics.
Here, this dissertation directly monitored the groundwater chemistry flowing through a thick weathered argillite zone (5- 25m), underlain by a thin soil layer (<0.75m) in a hillslope in the Elder Creek catchment, located at Angelo Coast Range Reserve, Northern California. Groundwater samples from three locations along the hillslope (i.e. upslope, midslope and downslope) and the creek samples from adjacent Elder Creek were collected at 1-3 days frequency from late 2008 to early 2013 using a novel autonomous sampling methodology developed for this study. This study site has been intensively monitored (5- 30 minute frequency) for microclimate, groundwater table/ temperature, soil moisture/ temperature, and tree sap flows since 2008.
Observations of the dynamics of major cations (Na, Mg, K, and Ca), Si, and reactive trace metals, led to the identification of three key governing processes, each occurring at specific locations within the critical zone, that are responsible for water chemistry evolution in the critical zone. During the rainy season, rain water and throughfall - rainwater that has penetrated through tree canopy - as it passes through the vadoze zone, rapidly gains major cations through via cation exchange reactions enhanced by elevated subsurface pCO2. At the same time, rainwater increases its Si concentration through dissolution of amorphous silica phases. This new water recharges the groundwater and in upslope and midslope wells, raising their groundwater table by 4-6 m. During this high flow regime, the connectivity between the groundwater and stream increases significantly. Therefore, this groundwater rapidly flows into the stream and the groundwater and stream display the similar chemistry (fast-flowing groundwater). During the dry season, the groundwater table and the stream discharge recede to their lowest levels and the cation concentrations of the groundwater become higher but its Si concentration decreases. During this season, the chemistry of groundwater is governed by thermodynamic equilibrium with the argillite at high pCO2 (slow-flowing groundwater). The groundwater may become fully equilibrated with reactive minerals such as calcite and clay minerals but may be far from equilibrium with primary minerals (e.g. albite). Silicon in the groundwater may form secondary minerals, decreasing its concentration. When this equilibrated groundwater enters the stream, its chemistry will be dramatically altered via degassing of CO2 leading to abiotic and biotic carbonate mineral precipitation at the hillslope-stream interface. The biotic calcite precipitation appears to increase the partitioning of Mg in calcite, decreasing the groundwater's Mg concentration by 30%.
This fundamental hydrochemistry process framework explains the dynamics of reactive species, such as Fe and Mn. The Mn concentrations in the subsurface paralleled the behavior of major cations but varied by as much as 3 orders of magnitude vs. factors of 2 to 5 for major cations; the groundwater's Mn concentration rapidly decreased as the groundwater table rises and vice versa. Multiple lines of evidence suggest that Mn is mostly in dissolved form, and like major cations and that the controlling processes for major cations govern the Mn's behavior. However, even during the high flow regime, Mn in the fast-flowing groundwater is precipitated when it enters the stream, likely via Mn-oxide precipitation at the hillslope-stream interface. In contrast, Fe displayed no systematic correlation with the groundwater table dynamics or stream discharge. The highest Fe concentrations in both groundwater and stream were expressed at the beginning of the rainy season. In addition, the Fe concentration in the groundwater and stream display quite similar values. This suggests that Fe may be in colloidal form, likely organic-ligand bound. This organic ligand Fe- colloid may be more stable and less bio-available that dissolved elements; hence, Fe precipitation at the hillslope-stream interface may be negligible.
This dissertation demonstrated that the slow-flowing groundwater transported through the deepest weathered bedrock zone plays a significant role in fluxes of solute and solid to the adjacent creek. The atmospheric inputs of major cations and Si were insignificant compared with the Elder Creek' solute fluxes while that of Mn was 1 -2 orders of magnitude higher than in Elder Creek's solute fluxes; rain and throughfall data for Fe are unreliable. The first process in the vadose zone is responsible for 55% (12 t/km2/year) of the annual cation fluxes. The slow-flowing component is responsible for the rest of 45% annual cation fluxes (9.4 t/km2/year). The third process - water chemistry transition at the hillslope-stream interface - will precipitate 29 t/km2/year of Ca + Mg (mostly as calcite) from the slow-flowing groundwater. This process is also responsible for precipitating approximately 0.2 t/km2/year of Mn and 0.02 t/km2/year of Fe, which are greater than their atmospheric inputs. These findings demonstrate that the loss by carbonate precipitation at the transition zone 1) is greater than the dissolved cation fluxes estimated based on the Elder Creek chemistry observations (21.4 t/km2/year); and 2) Fe and Mn are not accumulating in the system, unlike previously considered. The significant solute precipitation at the hillslope-stream interface suggests that the role of chemical weathering of rock may play a much greater role in sequestrating CO2 than the previously quantified. In addition, a large pool of chemical weathering solid fluxes has not been taken into account properly in the global element budgets.