You are here:
NEW SAMPLING THEORY FOR MEASURING ECOSYSTEM STRUCTURE
Citation:
Mulholland, R. NEW SAMPLING THEORY FOR MEASURING ECOSYSTEM STRUCTURE. U.S. Environmental Protection Agency, Washington, D.C., EPA/600/3-78/104.
Description:
This research considered the application of systems analysis to the study of laboratory ecosystems. The work concerned the development of a methodology which was shown to be useful in the design of laboratory experiments, the processing and interpretation of the results of these experiments, the development of model structures for microcosms based upon experimental data, the testing of the predictive capabilities of evaluative models, and the analysis of measurement errors in the presence of noise. Development of the experimental design methodology was based upon a non-Nyquist sampling theory which gave a priori computable bounds on the sample period for data collection. Results of the research project included a definition for the bandwidth of an ecosystem in the frequency domain, an indication of the effect of measurement noise on the sample period for data acquisition, a scheme for reducing this noise, and a computer code for model identification based upon the experimental design developed. The project also produced the conditions under which two new schemes for model identification were applicable to the construction of linear compartment models.