ARS | Publication request: TOWARDS SCALING-UP TOWER CO2 FLUX MEASUREMENTS IN GRASSLANDS OF THE NORTHERN GREAT PLAINS: PHENOMENOLOLGICAL MODELING USING FLUX PARTITIONING, VEGETATION INDICES AND GEOGRAPHIC INFORMATION SYSTEM (GIS)

Submitted to: Society for Range Management Meeting Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: January 6, 2004
Publication Date: January 15, 2004
Citation: GILMANOV, T.G., TIESZEN, L.L., WYLIE, B.K., FLANAGAN, L.B., FRANK, A.B., HAFERKAMP, M.R., MEYERS, T.P., AND MORGAN, J.A. 2004. TOWARDS SCALING-UP TOWER CO2 FLUX MEASUREMENTS IN GRASSLANDS OF THE NORTHERN GREAT PLAINS: PHENOMENOLOLGICAL MODELING USING FLUX PARTITIONING, VEGETATION INDICES AND GEOGRAPHIC INFORMATION SYSTEM (GIS). SOCIETY FOR RANGE MANAGEMENT MEETING ABSTRACTS #112.

Technical Abstract: Estimates of gross primary productivity (Pg) and ecosystem respiration (Re) were derived from 20 minute net CO2 fluxes monitored at five Northern Great Plains locations using light-response analysis. At the weekly scale, daytime ecosystem respiration ('day) was correlated (0.64<R2<0.95) with measured nighttime respiration ( 'night). 'day tended to be higher than 'night and regressions of 'day on 'night for all sites were different from the 1:1 line. Pg varied from 459 to 2491 g CO2 m-2 yr-1, Re from 996 to 1881 g CO2 m-2 yr-1, and net ecosystem exchange from '537 (source) to +610 g CO2 m-2 yr-1(sink) over all 13 site-years. Maximum daily light-use efficiencies 'd,max = Pg/PAR ranged from 0.014 to 0.032 mol CO2 (mol quanta)-1. Maximum weekly 'wk,max = Pg,wk/PARwk ranged from 0.008 to 0.019 mol CO2 (mol quanta)-1. Ten-day averages of PAR, Tair, Tsoil, soil moisture, temperature sum above 5 °C, and tower derived Pg and Re were compared to ten-day smoothed SPOT VEGETATION Normalized Difference Vegetation Index (NDVI). Pg was significantly higher correlated with NDVI than Pd (R2 0.46 to 0.77 and 0.05 to 0.58 respectively). A positive correlation was found between Re and NDVI (0.56<R2<0.77). Other ecological factors that are available as spatial-temporal data sets have potential to improve NDVI relationships using multivariate functions (Pg = fP(NDVI, X1, X2, ', Xn) and Re = fR(NDVI, Y1, Y2, ',Ym)). This will allow accurate scaling-up of local flux-tower measurements to larger geographical regions using GIS data, remote sensing, and environmental factors.