An empirical model of carbon fluxes in Russian tundra

This study presents an empirical model based on la GIS approach, which was constructed to estimate the large-scale carbon fluxes over the entire Russi an tundra zone. The model has four main blocks: (i) the computer map of tun dra landscapes; (ii) data base of long-term weather records; (iii) the subm odel of phytomass seasonal dynamics; and (iv) the submodel of carbon fluxes . The model uses exclusively original in situ diurnal CO2 flux chamber meas urements (423 sample plots) conducted during six field seasons (1993-98). T he research sites represent the main tundra biome landscapes (arctic, typic al, south shrub and mountain tundras) in the latitudinal diapason of 65-74 degreesN and longitudinal profile of 63 degreesE-172 degreesW. The greatest possible diversity of major ecosystem types within the different landscape s was investigated. The majority of the phytomass data used was obtained fr om the same sample plots. The submodel of carbon fluxes has two dependent [ GPP, Gross Respiration (GR)I and several input variables lair temperature, PAR, aboveground phytomass components). The model demonstrates a good corre spondence with other independent regional and biome estimates and carbon fl ux seasonal patterns. The annual GPP of Russian tundra zone for the area of 235x10(6) ha was estimated as -485.8+/-34.6x10(6) tC, GR as +474.2 +/- 35. 0x10(6) tC, and NF as -11.6 +/- 40.8x10(6) tC, which possibly corresponds t o an equilibrium state of carbon balance during the climatic period studied (the first half of the 20th century). The results advocate that simple reg ression-based models are useful for extrapolating carbon fluxes from small to large spatial scales.