Premium
The Joint Estimation of Soil Trace Gas Fluxes
Author(s) -
Hossler Katie,
Bouchard Virginie
Publication year - 2008
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj2007.0232
Subject(s) - joint (building) , trace (psycholinguistics) , nonlinear system , flux (metallurgy) , estimation theory , least squares function approximation , soil science , trace gas , non linear least squares , soil gas , mathematics , environmental science , statistics , soil water , chemistry , physics , quantum mechanics , architectural engineering , philosophy , linguistics , organic chemistry , estimator , engineering
Soil gas flux is commonly measured by monitoring the change in headspace gas concentration over time within a sealed compartment at the soil surface. Often, more than one trace gas is monitored at a time (e.g., CO 2 and CH 4 ), but the data fit separately. Flux estimates for CO 2 and CH 4 were obtained simultaneously by minimizing a weighted sum‐of‐squares error. The approximation of one model parameter for CH 4 , through theoretical relationship to the respective CO 2 parameter, reduced the total parameter count by one and allowed for the joint estimation of one parameter using the combined CO 2 and CH 4 datasets. The method of joint optimization was compared with separate optimization for two nonlinear models, using both real and simulated data. The datasets were best fit with the jointly optimized models. Furthermore, the jointly optimized models more accurately estimated initial soil–air fluxes (simulated data only). The method of joint optimization is recommended as a means to apply better‐fitting nonlinear models to typically small gas sample sets. This method is applicable to any number of trace gases monitored simultaneously.