TITLE:

HILBERT-HUANG TRANSFORM ANALYSIS OF GLOBAL PRECIPITATION DATASETS SINCE 1900

DATE:

Friday, Oct 28th, 2011

TIME:

3:30 PM

LOCATION:

GMCS 214

SPEAKER:

Samuel Shen
Department of Mathematics and Statistics
San Diego State University

ABSTRACT:

Average precipitation over the entire Earth is an important measure of global climate change. However, significant differences exist among the satellite observation data, rain gauge data, climate model data, and the reconstructed data from statistical methods and climate model assimilations. By using the Hilbert-Huang Transform (HHT) method, we made time-frequency analysis on four datasets of the global average annual total precipitation: CMAP (Climate Prediction Center Merged Analysis of Precipitation), GPCP (Global Precipitation Climatology Project), REOF (EOF-based reconstruction anchored on Global Historical Climatology Network station data), and MERG (the merged reconstruction from different latitude zones). The reconstructed datasets REOF and MERG span from 1900 to current, and the satellite-based datasets CMAP and GPCP span from 1979 to current. Our HHT analysis indicates the following: (1) The HHT intrinsic mode functions (IMF) can be used to identify the instrument change and help explain the temporal shifts of variances. (2) The Hilbert spectral and energy analyses help to quantify the errors of reconstruction and explore the optimal methods of blending different sources of observational data. (3) The HHT analysis and the probabilistic distribution analysis are critical to make statistical inferences regarding the trends of both mean and variance. This work was done in collaboration with David New, Thomas Smith, and Phillip Arkin.

HOST:

Dr. Andrew Cooksy.

DOWNLOAD:

PDF File