Temporal and spatial variations of rock weathering and CO₂ consumption in the Baltic Sea catchment

Abstract

This study provides the first estimate of silicate and carbonate weathering rates and derived CO₂ consumption rates for the Baltic Sea catchment using river chemistry data of 78 rivers draining into the Baltic Sea. The silicate weathering rates (denoted as total dissolved solids) of individual river basin range from 0.014 Tg/year to 4.55 Tg/year and the carbonate weathering rates range from 0.079 Tg/year to 6.49 Tg/year. The total CO₂ consumption across the Baltic catchment is approximately 3.9 Tg C/year and is almost equally shared by silicates and carbonates. Uncertainty associated with the weathering estimate is around 32%, which is mainly caused by incomplete pollution correction for a few major rivers in the south. The calculations for the boreal river basins have higher certainties because of less human impacts. The CO₂ consumption rate of individual river basin vary between 0.53 and 5.66 g C/m²/year with an average of 2.97 g C/m²/year, in which carbonates consume CO₂, 1.4 g C/m²/year and silicates take 1.5 g C/m²/year. This is in the same range as has been reported for the Mackenzie River and Siberian river basins, but at the lower range of tropical rivers, suggesting the Baltic Sea catchment holds solid weathering signals for high-latitude systems, especially in the pristine boreal silicate-dominated areas. The amount of CO₂ consumed by weathering in the Baltic Sea catchment accounts for approximately 3–30% of the net ecosystem carbon exchange (10–100 g C/m²/year), implying that weathering contributes as a significant sink of atmospheric CO₂. Although many studies have shown the positive relation between temperature and weathering rates in various river catchments, multiple regression analysis using the 40-year continuous records of river chemistry in the boreal area of the Baltic Sea catchment reveals a strong correlation between weathering flux and precipitation, but no statistically significant correlation between weathering and temperature. This suggests not only that temperature is not necessarily to be primary controlling factor for weathering rates, but also besides precipitation, other factors, like increased soil organic matter contents and water path changes may have high impact on weathering rates. The 40-year data analysis also shows generally increasing weathering fluxes by 10–20% in the pristine boreal area over the past decades. This indicates that increased CO₂ consumption by weathering and the resulting elevated dissolved inorganic carbon delivery to the ocean act as a negative feedback for ocean acidification, such as the Arctic Ocean that has become more acidic due to high terrestrial organic carbon delivery together with increased river water input.

Introduction

Evidence for weathering of minerals is given in river discharge and shows results from interactions between the atmosphere, the hydrosphere, the biosphere and the lithosphere. Weathering is responsible for the chemical compositions of soil water, ground water, river water (Conley, 2002, Livingstone, 1963, Meybeck, 1982) and over geological time scales, the chemical compositions of the ocean and the atmospheric CO₂ level (Berner, 1991, Berner, 2003, Walker and Hays, 2007, Walker et al., 1981). To understand the carbon cycle and how the atmospheric CO₂ level have evolved over Earth's history, we need to first unravel and quantify the weathering rates (Berner, 1991, Berner, 1994, Berner and Kothavala, 2001, Walker et al., 1981). So far, many studies have estimated catchment- and global-scale weathering rates and the associated CO₂ consumption using riverine chemical fluxes (e.g. Amiotte Suchet et al., 2003, Gaillardet et al., 1999, Hartmann et al., 2009, Meybeck, 1987, Meybeck, 2003, Moon et al., 2014, Mortatti and Probst, 2003).

The most prominent changes in climate are foreseen for high latitudes due to the polar amplification of global warming and therefore high-latitude areas are especially prone to elevated temperatures and possible changes in precipitation regimes. The climatic response of weathering is considered as a function of temperature and river discharge. Thus, it can be hypothesized that the rates of chemical weathering may increase in high-latitude systems. Here, especially the subarctic areas are of great interest to both local and global weathering estimates and the carbon budgets. Many studies have demonstrated that weathering is not controlled by a single factor although positive correlations between weathering and a single factor like temperature, water discharge and precipitation have been reported in both lab-controlled experiments (Kump et al., 2000) and field observations (Li et al., 2016, Millot et al., 2003, Moon et al., 2007, Oliva et al., 2003, White and Blum, 1995). Erosion rates have been shown to interfere simple interpretations of factors controlling weathering (Ferrier and Kirchner, 2008, Riebe et al., 2004, WEST et al., 2005, West, 2012). Millot et al. (2003) have observed that soil organic matter is able to superimpose on the effect of temperature in the Canada Shield and Mackenzie River systems. Strong chemical weathering has also been reported for organic rich permafrost soils beneath central Siberia (Botch et al., 1995, Pokrovsky et al., 2005, Zakharova et al., 2005). Vegetation can further increase weathering rates through producing organic acids and soil CO₂ (Humborg et al., 2004, Smedberg et al., 2006). All of these compounding factors have challenged estimations of weathering rates at the river basin scale and subsequently hindered global extrapolation of weathering rates and the associated CO₂ consumption.

Good estimate of weathering and CO₂ consumption rates requires high-resolution temporal samples of river chemical fluxes with synchronous water discharge and good knowledge on geological settings of studied areas. However, it is always difficult to continuously collect temporal and spatial samples and measure water discharge during field expeditions. The chemical fluxes are calculated by multiplying concentrations of collected samples either with mean annual precipitation or discharge data. These data could be either originated from satellite derived precipitation estimates or from temporally averaged discharge from on-site/nearby gauging stations (Hren et al., 2007, Huh et al., 1998a, Huh et al., 1998b, Moon et al., 2007). Further models may simulate the amount of major elements from various potential sources including precipitation, silicates, carbonates, evaporates and anthropogenic pollution (Bickle et al., 2005, Gaillardet et al., 1999). Studies using limited samples with asynchronous discharge show large uncertainties in the estimates of silicate weathering rates (Eiriksdottir et al., 2008, Ollivier et al., 2010, Qin et al., 2006, Tipper et al., 2006, West et al., 2002); and for the subsequent modelling, the uncertainties are largely dependent on the ability to correctly assume the compositions of the pre-assigned source reservoirs (Négrel et al., 1993, Wu et al., 2005). So far, relatively continuous temporal and spatial sampling and well-constrained estimates have been documented for a few well-studied rivers in tropical and temperate areas (Moquet et al., 2011, Mortatti and Probst, 2003, Qin et al., 2006, Tipper et al., 2006) and high-latitude catchments (Huh et al., 1998a, Huh et al., 1998b, Millot et al., 2003) although the way of sampling strategy still varies in these rivers, such as low resolution of spatial sampling, less frequency of temporal sampling, analysed chemical species etc. However, good weathering estimates in high-latitude systems are still scarce and it can be hypothesized that positive trends in weathering fluxes and CO₂ consumption may be a negative feedback to atmospheric CO₂ concentrations in regions prone to polar amplification of climate change.

In this contribution, we compile monthly river chemistry data and synchronous discharges of 78 rivers in the Baltic Sea catchment. This dataset covering almost the entire catchment and including continuous monthly sampling and measurements for years allows for a well-constrained weathering estimate for such a large subarctic area. In particular, the good weathering estimate is obtained in the silicate-dominated boreal area, which are relatively undisturbed by human activities although large uncertainties likely exist in the calculation for the southern carbonate-dominated areas given the insufficient corrections for heavy industrial pollutions. The main objective of this study is for the first time to estimate rates of silicate and carbonate weathering and CO₂ consumption across the Baltic Sea catchment. Moreover, the unique data record of monthly river chemistry and water discharge in the unpolluted boreal area over four decades provides us opportunities to conduct long-term time series analysis of weathering trends as well as major controlling factors, offering new constrains on the CO₂ budget in high-latitude systems.