Large language model and domain-specific model collaboration for smart education

In this paper,we introduce the large language model and domain-specific model collaboration(LDMC)framework designed to enhance smart education.The LDMC framework leverages the comprehensive and versatile knowledge of large domain-general models,combines it with the specialized and disciplinary knowledge from small domainspecific models(DSMs),and incorporates pedagogy knowledge from learning theory models.This integration yields multiple knowledge representations,fostering personalized and adaptive educational experiences.We explore various applications of the LDMC framework in the context of smart education.

Carbon pools and fluxes in the China Seas and adjacent oceans

The China Seas include the South China Sea, East China Sea, Yellow Sea, and Bohai Sea. Located off the Northwestern Pacific margin, covering 4700000 km^2 from tropical to northern temperate zones, and including a variety of continental margins/basins and depths, the China Seas provide typical cases for carbon budget studies. The South China Sea being a deep basin and part of the Western Pacific Warm Pool is characterized by oceanic features; the East China Sea with a wide continental shelf, enormous terrestrial discharges and open margins to the West Pacific, is featured by strong cross-shelf materials transport; the Yellow Sea is featured by the confluence of cold and warm waters; and the Bohai Sea is a shallow semiclosed gulf with strong impacts of human activities. Three large rivers, the Yangtze River, Yellow River, and Pearl River, flow into the East China Sea, the Bohai Sea, and the South China Sea, respectively. The Kuroshio Current at the outer margin of the Chinese continental shelf is one of the two major western boundary currents of the world oceans and its strength and position directly affect the regional climate of China. These characteristics make the China Seas a typical case of marginal seas to study carbon storage and fluxes. This paper systematically analyzes the literature data on the carbon pools and fluxes of the Bohai Sea,Yellow Sea, East China Sea, and South China Sea, including different interfaces(land-sea, sea-air, sediment-water, and marginal sea-open ocean) and different ecosystems(mangroves, wetland, seagrass beds, macroalgae mariculture, coral reefs, euphotic zones, and water column). Among the four seas, the Bohai Sea and South China Sea are acting as CO_2 sources, releasing about0.22 and 13.86–33.60 Tg C yr^(-1) into the atmosphere, respectively, whereas the Yellow Sea and East China Sea are acting as carbon sinks, absorbing about 1.15 and 6.92–23.30 Tg C yr^(-1) of atmospheric CO_2, respectively. Overall, if only the CO_2 exchange at the sea-air interface is considered, the Chinese marginal seas appear to be a source of atmospheric CO_2, with a net release of 6.01–9.33 Tg C yr^(-1), mainly from the inputs of rivers and adjacent oceans. The riverine dissolved inorganic carbon (DIC) input into the Bohai Sea and Yellow Sea, East China Sea, and South China Sea are 5.04, 14.60, and 40.14 Tg C yr^(-1),respectively. The DIC input from adjacent oceans is as high as 144.81 Tg C yr^(-1), significantly exceeding the carbon released from the seas to the atmosphere. In terms of output, the depositional fluxes of organic carbon in the Bohai Sea, Yellow Sea, East China Sea, and South China Sea are 2.00, 3.60, 7.40, and 5.92 Tg C yr^(-1), respectively. The fluxes of organic carbon from the East China Sea and South China Sea to the adjacent oceans are 15.25–36.70 and 43.93 Tg C yr^(-1), respectively. The annual carbon storage of mangroves, wetlands, and seagrass in Chinese coastal waters is 0.36–1.75 Tg C yr^(-1), with a dissolved organic carbon(DOC) output from seagrass beds of up to 0.59 Tg C yr^(-1). Removable organic carbon flux by Chinese macroalgae mariculture account for 0.68 Tg C yr^(-1) and the associated POC depositional and DOC releasing fluxes are 0.14 and 0.82 Tg C yr^(-1), respectively. Thus, in total, the annual output of organic carbon, which is mainly DOC, in the China Seas is 81.72–104.56 Tg C yr^(-1). The DOC efflux from the East China Sea to the adjacent oceans is 15.00–35.00 Tg C yr^(-1). The DOC efflux from the South China Sea is 31.39 Tg C yr^(-1). Although the marginal China Seas seem to be a source of atmospheric CO_2 based on the CO_2 flux at the sea-air interface, the combined effects of the riverine input in the area, oceanic input, depositional export,and microbial carbon pump(DOC conversion and output) indicate that the China Seas represent an important carbon storage area.

Modeling the contribution of the microbial carbon pump to carbon sequestration in the South China Sea

The two key mechanisms for biologically driven carbon sequestration in oceans are the biological pump(BP) and the microbial carbon pump(MCP); the latter is scarcely simulated and quantified in the China seas. In this study, we developed a coupled physical-ecosystem model with major MCP processes in the South China Sea(SCS). The model estimated a SCSaveraged MCP rate of 1.55 mg C m^(-2) d^(-1), with an MCP-to-BP ratio of 1:6.08 when considering the BP at a depth of 1000 m.Moreover, the ecosystem responses were projected in two representative global warming scenarios where the sea surface temperature increased by 2 and 4°C. The projection suggested a declined productivity associated with the increased near-surface stratification and decreased nutrient supply, which leads to a reduction in diatom biomass and consequently the suppression of the BP. However, the relative ratio of picophytoplankton increased, inducing a higher microbial activity and a nonlinear response of MCP to the increase in temperature. On average, the ratio of MCP-to-BP at a 1000-m depth increased to 1:5.95 with surface warming of 4°C, indicating the higher impact of MCP in future ocean carbon sequestration.