Bayesian estimation of uncertainty in land surface-atmosphere flux predictions

This studya ddressesth e assessmenot f uncertaintya ssociatedw ith predictions of land surface-atmospherfelu xesu singB ayesianM onte Carlo simulationw ithin the generalizedli kelihoodu ncertaintye stimation( GLUE) methodologyE. ven a simples oil vegetation-atmosphertrea nsfer( SVAT) schemei s shownt o lead to multiple acceptable parameterizationsw hen calibrationd ata are limited to timescaleso f typicali ntensivef ield campaignsT.h e GLUE methodologya ssignas likelihoodw eightt o eacha cceptables imulation. As more data becomea vailablet, hesel ikelihoodw eightsm ay be updatedb y usingB ayes equationA. pplicationo f the GLUE methodologyc an be shownt o reveald eficiencieisn model structure and the benefit of additional calibration data. The method is demonstrated with data sets taken from FIFE sites in Kansas, and ABRACOS data from the Amazon. Estimates of uncertaintya re propagatedf or each data set revealings ignificantp redictiveu ncertainty. The value of additional periods of data is then evaluated through comparing updated uncertaintye stimatesw ith previouse stimatesu singt he Shannone ntropym easure