Earth's long-term climate habitability is significantly influenced by the chemical weathering of silicate minerals, which acts as a negative feedback mechanism. A prevailing hypothesis posits that increased volcanic CO2 emissions raise surface temperatures through the greenhouse effect. Higher temperatures, in turn, enhance silicate weathering, leading to atmospheric CO2 drawdown. The strength of this climate-weathering feedback hinges on silicate weathering's response to temperature changes. While laboratory and small-watershed studies have shown a temperature dependence, this relationship is often obscured at larger spatial or temporal scales due to covarying factors or minor temperature fluctuations. Consequently, other mechanisms have been proposed, including hydrological regulation (precipitation and runoff), tectonic uplift, and land surface reorganization. However, validating these hypotheses over geological timescales is challenging because paleo-weathering records might lack robustness due to inconsistencies among proxies, and high-quality paleo-records of individual forcing factors are limited. This study aims to investigate the link between environmental factors and silicate weathering using a comprehensive modern dataset, spanning a broad environmental gradient, to overcome the limitations of existing studies. The researchers focus on two widely applied weathering proxies derived from siliciclastic sediments: the chemical index of alteration (CIA) and the weathering index of Parker (WIP). CIA represents the proportion of immobile Al2O3 to labile oxides, while WIP sums the molar proportions of labile elements, weighted by their weathering susceptibility. Both indices reflect silicate weathering intensity, particularly tracking feldspar hydrolysis, a significant process considering feldspar's abundance in the Earth's upper crust. The study compiled a large major element dataset (n=3828) from modern river sediments across six continents to calculate CIA and WIP values and extracted corresponding basin-scale environmental forcing factors (n=20) to assess the controls on silicate weathering intensity and the potential use of weathering indices as paleoclimate proxies.

The existing literature on the relationship between silicate weathering and climate change exhibits some discrepancies. While some studies emphasize the importance of temperature as a primary driver of weathering, others highlight the roles of precipitation, tectonic uplift, or land surface reorganization. These discrepancies arise partly from challenges in validating paleo-weathering records, which may show inconsistencies across different proxies. Also, the availability of high-quality paleo-records of individual forcing factors (rather than mixed signals) is limited, making it difficult to isolate the effect of each variable. Several studies have attempted to understand the influence of individual factors, proposing mechanisms such as temperature control, precipitation control, and combined climate control. However, the co-variation of temperature and precipitation, along with the limited variability of one factor compared to the other in local studies, hinder disentangling their respective effects. The use of weathering indices as paleoclimate proxies has been explored but faces challenges due to potential confounding factors like grain size, provenance, and lithology. While the latitudinal trend of silicate weathering intensity has been observed in various studies, these studies often rely on datasets with limited spatial resolution, which might not capture the full complexity of the interactions between weathering and environmental factors.

The study compiled a global dataset of fine-grained sediment major element compositions (n = 3828) from modern rivers across six continents. The data were used to calculate the Chemical Index of Alteration (CIA) and Weathering Index of Parker (WIP). The authors focused on CIA in the main text for brevity, noting that WIP showed consistent trends (presented in supplementary information). A total of 20 basin-scale environmental forcing factors were extracted from gridded datasets to assess their influence on silicate weathering. These factors were categorized into climate, geomorphology, lithology, and land cover. The dataset included silt and clay-sized sediments to minimize mineralogical differentiation caused by hydraulic sorting. The researchers calculated CIA values based on molar proportions of silicate-bound major elements. They observed a decreasing trend in CIA with latitude, reflecting stronger weathering in the tropics. The correlation between CIA and each environmental factor was assessed using correlation coefficients (R). Weak correlations were found between CIA and land cover, lithological factors, geomorphic characteristics (except soil thickness), and modeled sediment yield. However, a strong positive correlation (R = 0.60) was observed between CIA and mean annual temperature (MAT), indicating a primary temperature control on silicate weathering intensity. The study also analyzed the relationship between CIA and mean annual precipitation (MAP), revealing a more complex, non-monotonic relationship that they attribute to competing effects of increased water supply versus increased erosion. To better understand the influence of each climatic factor while minimizing others, the data were grouped by temperature and precipitation zones, revealing a monotonic increase of CIA with MAT (above 0 °C). A linear relationship between MAT and CIA was established from binned data (to reduce regional variability) for use in paleoclimate reconstruction. The utility of this relationship was tested by deriving a global-average CIA, which aligned with previous estimates. The potential biases associated with applying the MAT-CIA equation for paleo-temperature reconstruction were addressed. The authors also compared paleo-CIA records (n=10) from various climatic events with independent paleo-temperature records (biomarker or pollen), finding a good overall agreement within uncertainties. To connect temperature to weathering-driven CO2 consumption, they converted the MAT-CIA relationship into one between MAT and the percentage of feldspar dissolution (fdiss). The non-linear relationship observed was attributed to depletion of plagioclase and dominance of orthoclase at high MAT. A simplified feldspar weathering model, incorporating the MAT-fdiss relationship and feldspar hydrolysis reactions, was used to estimate the CO2 consumption change in response to temperature changes. The model suggests a 28% increase in CO2 consumption for a 3 °C temperature rise, falling within the range predicted by the Arrhenius law.

This study, using a large, globally distributed dataset of river sediments, provides robust evidence for the primary control of temperature on silicate weathering intensity. The key findings are: 1. A strong positive correlation (R = 0.60) exists between mean annual temperature (MAT) and the Chemical Index of Alteration (CIA), a proxy for silicate weathering intensity. This relationship is stronger than the correlations observed between CIA and other environmental factors such as precipitation, geomorphic characteristics, or lithology. 2. The relationship between CIA and MAT is monotonic and positive (above 0°C), indicating that higher temperatures consistently lead to stronger silicate weathering. Below 0°C, higher CIA values are attributed to glacial processes. 3. The impact of precipitation on CIA is complex and non-monotonic, likely reflecting competing effects of increased water supply versus increased erosion. 4. A linear equation was developed to estimate MAT from CIA values, useful for reconstructing past temperatures from sedimentary archives. This relationship was validated by comparing with independent paleo-temperature records, showing agreement within uncertainties. 5. A nonlinear response of feldspar dissolution to temperature was found, attributed to the depletion of the more reactive plagioclase mineral at higher temperatures, leaving less reactive orthoclase dominating. 6. A simplified feldspar weathering model, incorporating the MAT-feldspar dissolution relationship and stoichiometry of hydrolysis reactions, showed that a 3°C increase in MAT would result in a 28% increase in CO2 consumption through silicate weathering. This estimate is consistent with estimations from the Arrhenius law. These findings highlight the importance of temperature as a primary driver of silicate weathering and offer a new empirical tool for quantitative paleoclimate reconstruction and carbon cycle modeling.

The study's findings strongly support the hypothesis that temperature is the primary driver of silicate weathering at a global scale, resolving some existing discrepancies in the literature. The strong correlation between MAT and CIA, stronger than correlations with other factors, provides compelling evidence for the dominance of temperature control. The developed MAT-CIA relationship offers a valuable tool for quantitative paleoclimate reconstruction, particularly in deep time where other proxies are less reliable. The nonlinear relationship between temperature and feldspar dissolution, reflecting the relative proportions of different feldspar minerals, provides crucial insights into the dynamics of weathering processes. This nonlinearity is crucial for understanding the long-term climate-weathering feedback mechanism. The model developed to estimate CO2 consumption from silicate weathering provides a useful empirical constraint for carbon cycle modeling. The study suggests that the increased land surface reactivity during the late Cenozoic cooling could be attributed, at least in part, to a higher proportion of more reactive plagioclase minerals available for weathering. Further research should focus on refining the model by incorporating more detailed information on mineral composition and kinetics and testing it against a wider range of paleo-records.

This study presents compelling evidence for the primary control of temperature on silicate weathering intensity at a global scale. The authors establish a strong empirical relationship between mean annual temperature and the chemical index of alteration (CIA), allowing for quantitative paleo-temperature reconstruction. This relationship, combined with a model for CO2 consumption through silicate weathering, offers valuable new tools for paleoclimate studies and carbon cycle modeling. The non-linearity of the temperature effect highlights the importance of mineral composition in regulating weathering rates. Future work should focus on expanding the dataset to improve the precision of paleo-temperature reconstructions and on improving the weathering model by incorporating more complex mineral kinetics.

The study relies primarily on the CIA as a proxy for silicate weathering intensity. While CIA is widely used, other factors such as grain size and sediment provenance might affect its interpretation. The model used for CO2 consumption estimation simplifies complex weathering processes and does not fully account for all factors affecting weathering rates. The comparison of paleo-CIA with independent temperature records is limited by the availability of such records, particularly at the regional scale. Additionally, the use of basin-averaged environmental data might mask local variability in weathering processes.