Refine
Year of publication
- 2003 (2) (remove)
Document Type
- Article (2)
Language
- English (2)
Has Fulltext
- no (2)
Is part of the Bibliography
- no (2)
Institute
Begutachtungsstatus
- peer-reviewed (1)
In this paper we discuss how finite data sets influence experimental measurements of the autocovariance function. Autocovariance estimators are biased, meaning that the expectation value for any measured autocovariance function is not identical to the actual autocovariance function. In this work we show that the measured autocovariance function for a finite length time series must become negative for some lag times. We derive analytic corrections to these finite time errors for different types of correlated random sequences. Our results explain the apparent anticorrelated noise observed in experimental observations.
We present the results of kinetics experiments on quantum-size-effects (QSE) Pb islands formed on Si(111). We have looked at the evolution of seven-layer Pb islands from five-layer Pb islands due to the transport of Pb atoms from the interisland region, overcoming a barrier at the island edges, and moving to the island tops. By analyzing low-energy electron-diffraction and scanning-tunneling microscopy data, we have estimated the barrier for layer formation in this system to be ∼0.32eV. This energy is much larger than that expected from QSE electronic contributions alone.