Choice of the appropriate group of signals to model the onset, the length and the volume of the Los Banos rainy season

Professorial Chair Lecture

Metro Manila Commission Professorial Chair Lecture

Place

CEAT Lecture Hall, CEAT, UPLB, College, Laguna

Date

6-29-2001

Abstract

Abstract: The onset, length and the total volume of water of the rainy season in the Philippines where there is a distinct dry and wet season vary from year to year. The process is quite natural and the annual numerical observations of the data form natural time series that would be eligible for time series modeling in both time and frequency domain. The various groups of signals tested to model each of the natural time series using frequency domain are:

(a) the 33 signals found from the first 100 data points of the 1163 to 1262 period of the CHIN001A.CRN tree-ring chronology;

(b) the 294 signals from the 1163 to 1982 period of the CHIN001A.CRN tree-ring data;

(c) the 1413 signals from the first 4000 data points of the CA535.CRN tree-ring data;

(d) the 2835 signals from the full 7979 data points of the CA535.CRN tree-ring data;

(e) the 273 signals from the 1198 to 1975 period of the AUS002.CRN tree-ring data;

(f) the 332 signals from the 1028 to 1975 period of the AUS006.CRN tree ring data;

(g) the 38 signals from the 1847 to 1983 period of the JAP004.CRN tree ring data;

(h) the 90 signals from the 1750 to 1985 period of the JAP007.CRN tree ring data;

(i) the 52 signals from the 1839 to 1995 period of the IND0001.CRN tree ring data; and

(j) the 161 signals from the 1839 to 1995 period of the IND0001.CRN tree ring data.

All the groups of signals are expected to reconstruct each of the series but not all or only some of them can consistently forecast any of the series even for one future season can be seen based on this study. Additional effort will be made to gather a few more groups of signals from other long natural time series data from other parts of the world. This work will also try to confirm the notion that no matter where a group of signals come from, as long as there are adequate number of frequencies that have adequate resolution of their location in the spectrum they are all capable of reconstructing a natural time series. However these do not necessarily have the same adequacy for prediction purposes as shown by the results.

Key words: spectral density, natural time series, precipitation, amplitude spectrum, phase c-.., spectrum, forecast, and spectral window.

Location

UPLB Main Library Special Collections Section

Call Number

Sp. Col.

Pages/Collation

22 leaves

This document is currently not available here.

Share

COinS