hide
Free keywords:
LAST GLACIAL MAXIMUM; ATMOSPHERIC CO2; SEISMOTECTONIC INFLUENCE;
HYDROLOGICAL ALTERATIONS; WEATHERING PROCESSES; CONTINENTAL EROSION;
SURFACE-WATER; JAPAN; OCEAN; ROCKSJapan; Dissolved silica; Empirical model; Chemical weathering;
Abstract:
Silicate weathering and resulting transport of dissolved matter influence the global carbon cycle in two ways. First by the uptake of atmospheric/soil CO2 and second by providing the oceanic ecosystems via the fluvial systems with the nutrient dissolved silica (DSi). Previous work suggests that regions dominated by volcanics are hyperactive or even "hot spots" concerning DSi-mobilization. Here, we present a new approach for predicting DSi-fluxes to coastal zones, emphasizing "first-order" controlling factors (lithology, runoff, relief, land cover and temperature). This approach is applied to the Japanese Archipelago, a region characterized by a high percentage of volcanics (29.1% of surface area). The presented DSi-flux model is based on data of 516 catchments, covering approximately 56.7% of the area of the Japanese Archipelago. The spatial distribution of lithology-one of the most important first order controls-is taken from a new high resolution map of Japan. Results show that the Japanese Archipelago is a hyperactive region with a DSi-yield 6.6 times higher than the world average of 3.3 t SiO2 km(-2) a(-1), but with large regional variations. Approximately 10% of its area exceeds 10 times the world average DSi-yield. Slope constitutes another important controlling factor on DSi-fluxes besides lithology and runoff, and can exceed the influence of runoff on DSi-yields. Even though the monitored area on the Japanese Archipelago stretches from about 31A degrees to 46A degrees N, temperature is not identified as a significant first-order model variable. This may be due to the fact that slope, runoff and lithology are correlated with temperature due to regional settings of the Archipelago, and temperature information is substituted to a certain extent by these factors. Land cover data also do not improve the prediction model. This may partly be attributed to misinterpreted land cover information from satellite images. Implications of results for Earth System and global carbon cycle modeling are discussed.
