Geoinformatics & Geostatistics: An OverviewISSN: 2327-4581

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

A Hybrid Model Using Time Series and Markov Chain for the Detection of the Main Periodicities on the Temperature Data in Nisyros Volcano, South Aegean

Nisyros volcano has shown a series of volcanic and seismo-tectonic events including recent volcanic phreatic eruptions intruding an impermeable layer near the surface. Geological and atmospheric phenomena contribute to the heat output from the hydrothermal system. To monitor the temperature fluctuations from a vent gas emission in Nisyros, a temperature sensor has been installed in a fumarole near the Lofos and Laki sites. Raw and decomposed temperature data were analyzed to determine the cycles of heat contribution at the surface from the endogenic volcanic activity and from the interaction with the atmospheric temperature. The time series decomposition using Kolmogorov-Zurbenko filter and the Markov chain approach have been utilized as complementary tools for the determination of the main periodicities on the temperature data. The Kolmogorov-Zurbenko filter is used for the time series decomposition of the temperature data into the longand
short term component of the time series. Markov chain analysis was used to determine probabilistically the periodicity in the temperature data and aggregate load impact in forecasting. Using spectral analysis and cluster analysis we determine the main periodicities of the longand short term cycles of heat contribution. Raw temperature data have shown an accelerated frequency of hourly temperature decreases at the surface. Small temperature decreases occur within the warmer atmospheric temperatures and dramatic temperature decreases occur during colder atmospheric temperatures. A physical or a mechanical phenomenon may  prevent heat from reaching the surface. The time series decomposition performs well in long term cycles while the Markov chain performs better in short term cycles.
Both periods are related to cyclones that occur in the Aegean Sea. Since Nisyros is an island, the temperature measurements are probably affected by various phenomena (cyclones, sea tides) that occur in the sea and affect the hydrothermal convection. 

Special Features

Full Text

View

Track Your Manuscript

Share This Page

Media Partners

Associations