Skip to main content
ARS Home » Pacific West Area » Logan, Utah » Forage and Range Research » Research » Publications at this Location » Publication #189272
Title: SCALING UP OF SITE CARBON DYNAMICS TO PREDICT THE CARBON DYNAMICS IN KAZAKHSTAN, CENTRAL ASIA

Author
item WYLIE, B - USGS EROS DATA CENTER
item GILMANOV, T - SOUTH DAKOTA STATE UNIV
item STENSAAS, S - USGS EROS DATA CENTER
item SALIENDRA, N - USDA FOREST SERVICE
item Johnson, Douglas
item AKSHALOV, K - BARAKEV KAZAKH RESEARCH
item Frank, Albert
item ZHANG, L - USGS EROS DATA CENTER
item DOYLE, R - USGS EROS DATA CENTER
item LACA, E - AGRONOMY & RANGE SCI.
item DEMMENT, M - AGRONOMY & RANGE SCI.

Submitted to: International Grasslands Congress
Publication Type: Proceedings
Publication Acceptance Date: 2/12/2005
Publication Date: 6/24/2005
Citation: Wylie, B.K., Gilmanov, T.G., Stensaas, S.L., Saliendra, N.Z., Johnson, D.A., Akshalov, K., Frank, A.B., Zhang, L., Doyle, R.F., Laca, E.A., Demment, M.W. 2005. Scaling up of site carbon dynamics to predict the carbon dynamics in kazakhstan, central asia. International Grasslands Congress.

Interpretive Summary: Climate and management determine whether rangelands store or lose soil carbon. Regional carbon dynamics in northern Kazakhstan have not been documented previously, and this area has undergone significant land use changes since the dissolution of the Soviet Union in 1990. The objective of our study was to quantify the regional carbon flux dynamics of the extensive steppe area in northern Kazakhstan. Micrometeorological towers were used to measure the gain or loss of carbon between the ecosystem and the atmosphere in the steppes of northern Kazakhstan and in similar areas in the Northern Great Plains of the U.S. Carbon losses measured during the winter of 2001-2002 were used to estimate carbon losses from other years. These data from micrometeorological towers were combined with satellite imagery to construct maps of carbon gains and losses for the Kazakh steppe during 1998 to 2001. Results from our study showed that soil carbon steadily increased in the Kazakh steppe across the four years. Our results also suggested that soil carbon could be increased in northern Kazakhstan by converting land areas from spring wheat production (depletion of soil carbon) to grassland (carbon accumulation).

Technical Abstract: Climate and management determine whether rangelands are net carbon sources or carbon sinks. Regional carbon dynamics of the Kazakh steppe have not been documented previously, and the area has undergone significant land use changes with the collapse of agriculture following the dissolution of the Soviet Union. The objective of this study was to quantify the regional carbon flux dynamics of these extensive steppes. Carbon flux towers were used to measure the net flux of carbon dioxide (CO2) between the ecosystem and the atmosphere. Detailed light curve analysis was used to partition these net fluxes into ecosystem respiration (Re) and photosynthesis (gross primary production, Pg). Spatial and temporal data sets of photosynthetically active radiation (PAR), temperature, precipitation, Normalized Difference Vegetation Index (NDVI), and NDVI-derived metrics were combined with flux tower derived gross primary production (Pg) and ecosystem respiration (Re) into a 10-day time step training database. Regression tree models were parameterized using training databases from the Northern Great Plains and Kazakh steppe flux towers. These regression tree models were then applied across the Kazakh steppe to map growing season 10-day Pg and Re. Annual fluxes were estimated using gap-filled winter fluxes from the winter of 2001-2002 based on previously described methods. Inter-annual and spatial variability of annual net ecosystem exchange (NEE) was investigated using inter-annual statistics and spatial moving window statistics. Regression trees used for mapping of growing season 10-day Pg and Re were robust and accurate. The Kazakhstan winter flux estimate (1.25 g CO2/m2/day) was used with growing season NEE to estimate annual NEE for each year from 1998 to 2001. The areas associated with inter-annual carbon sinks were 55 times larger than inter-annual carbon source areas. Carbon sources surrounded wetland areas in Kazakhstan and may be attributable to wetland dynamics not being adequately captured by the MODIS land cover. The annual average NEE for the Kazakh steppe was relatively stable with the greatest value occurring in 2000. This resulted in an increase in accumulated carbon across the four years. The local variance images identified areas with lower NEE near some urban areas. Although this was not common, this may suggest overgrazed area with reduced carbon sequestration and potential for improvement. The results of this study showed that regression tree algorithms were effective when applied regionally to the Kazakh steppe. This approach maximized the utility of intercontinental flux towers for regional mapping. The Kazakh steppe has more average cumulative sinks than sources. Carbon sequestration could be increased in Kazakhstan with changes from production of spring wheat (depletion of soil carbon) to grassland (carbon accumulation).