Temperature dependence of silicate weathering in nature: How strong a negative feedback on long- term accumulation of atmospheric CO2 and global greenhouse warming?

Estimation of the temperature dependence of natural feldspar weathering in two catchments at different elevations yields an apparent Arrhenius activation energy of 18.4 kcal/mol (77.0 kj/mol), much higher than most laboratory values. This finding supports recent suggestions that hydrolytic weathering of silicate minerals may consume carbonic acid and thereby remove atmospheric carbon dioxide more rapidly with increasing temperature than previously thought. This result provides a stronger negative feedback on long-term greenhouse warming than has been assumed in most models of global carbon cycling. The present estimate was determined from the ratio of feldspar weathering rates (determined by geochemical mass balance) in the southern Blue Ridge Mountains of North Carolina, United States. Temperature (a function of elevation) is the only factor that differs between the two catchments; parent rock type, aspect, hillslope hydrology, and vegetation type and successional stage are the same in both. INTRODUCTION Silicate-mineral weathering reactions are fundamental processes that determine the contributions of continental crustal weathering to global geochemical cycles. Silicate minerals weather by hydrolysis, thereby consuming naturally-occurring acids, including carbonic acid. Because carbonic acid forms by the reaction of CO2 with water, reactions that consume carbonic acid remove CO2 from the atmosphere. Silicate weathering is responsible for over half (58%) of the carbonic acid consumption by continental weathering (Berner and Berner, 1987); weathering is therefore an important component of the global long-term carbon cycle and associated environmental consequences such as global ("greenhouse") warming. In spite of the importance of weathering in models of the long-term global carbon cycle, the response of silicate-mineral weathering rates to changes in Earth's surface temperature is not well known. This paper reports the determination of the activation energy of feldspar weathering from natural systems. GEOLOGIC BACKGROUND Suites of catchments along elevation gradients have recently been studied to investigate the effect of elevation-dependent factors (temperature, soil thickness, vegetation) on silicate-mineral weathering rates (Drever and Zobrist, 1992). A similar approach is used here. Catchments 2 and 34 (C2 and C34) of the U.S. Department of Agriculture (Forest Service) Coweeta Hydrologic Laboratory were used in this study. Coweeta is located —16 km southwest of Franklin, North Carolina, in the southern Appalachian Blue Ridge province. The geology, hydrology, topography, climate, and land-use history of the study area, summarized here, are described in detail elsewhere (Velbel, 1985a, 1985b, 1992; Swank and Crossley, 1988; Hatcher, 1979, 1980, 1988; Ciampone et al., 1992). C2 and C34 range in elevation from 709 to 1004 m and from 866 to 1184 m, respectively. Both have been free of any intentional anthropogenic disturbance since 1924, and both have a south-facing aspect. Annual average rainfall is high (~2 m), and the two catchments are similar in their topography (as measured by the relief ratio; Grantham and Velbel, 1988). All rock types and mineral assemblages of the catchments are those of the Tallulah Falls Formation (late Precambrian), a high-rank (amphibolite facies) metamorphic unit consisting principally of metasedimentary quartz-mica schists and mica-feldspar-quartz gneisses. Plagioclase feldspar (sodic andesine, An32) is common in C2 and C34; its composition is areally very homogeneous. Both catchments lie along structural strike, such that both transect the same parts of the lithostratigraphic unit. Because parent rock type, aspect (south-facing), and vegetation type and successional stage are as similar as is possible for natural systems, temperature is the only factor that differs between the two catchments. Weathering profiles in C2 and C34 consist of a thick (average thickness 6 m; range 0 to 20 m) mantle of soil and saprolite overlying a hydrologically impervious base of unweathered crystalline rock (Velbel, 1985a, 1985b). More resistant and remnant primary minerals and relatively insoluble weathering products retain much of the primary texture and preserve the volume of the parent material (Velbel, 1985a, 1990a). Despite the thick saprolites and mature soils, even the uppermost (oldest) soil horizons on the thickest (oldest) saprolites in the Blue Ridge landscape still contain weatherable primary minerals, including plagioclase feldspar (Velbel, 1992). Hillslope hydrology at Coweeta is such that stream-water solute loads are depth-integrated samples representing weathering throughout the entire thickness of the profile. Interflow and overland flow (runoff) are not significant processes at Coweeta; precipitation infiltrates almost immediately, even on steep slopes (average 27°). Base and storm flows at Coweeta are sustained principally by water from subsurface soils and saprolites (Velbel, 1985b; Hewlett, 1961; Hewlett and Hibbert, 1963). This water continues to move downward through the saprolitic weathering profile until it reaches the hydrologically impervious unaltered bedrock. At this point the water is shunted laterally toward the stream (Velbel, 1985b). GEOLOGY, v. 21, p. 1059-1062, December 1993 1059 TABLE 1. FELDSPAR WEATHERING RATES, AND TOPOGRAPHY AND TEMPERATURE DATA FOR COWEETA CATCHMENTS 2 AND 34