Principal Component Analysis in 6 Steps

The Principal Component Analysis (PCA) is equivalent to fitting an n-dimensional ellipsoid to the data, where the eigenvectors of the covariance matrix of the data set are the axes of the ellipsoid. The eigenvalues represent the distribution of the variance among each of the eigenvectors. To understand the method, it is helpful to know something about matrix algebra, eigenvectors, and eigenvalues. Here is a n=2 dimensional example to perform a PCA without the use of the MATLAB function pca, but with the function of eig for the calculation of eigenvectors and eigenvalues. Continue reading “Principal Component Analysis in 6 Steps”

How to Become a Geoscience Data Analyst

Working with quantitative data, which requires sophisticated mathematical and computer-assisted evaluation methods, came very late in the geological sciences, compared to other scientific disciplines. Unfortunately, in many geology courses worldwide university-level mathematics and computational geosciences is not included, as my experience – as the current chair of the examination committee of our geoscience masters program – from processing this year’s masters applications suggests. Continue reading “How to Become a Geoscience Data Analyst”

Public Lecture on Climate Change and Human Evolution at U Potsdam Children’s University

On September 29, 2017, the University of Potsdam will once again open its lecture halls for children to inspire them with the world of science. On this day, children can listen to interesting lectures, discuss with the scientists, and can experience exciting experiments. On Friday 29 September 2017 I will give a public lecture on tectonics, climate and human evolution for 8–10 year old children. Please visit the webpage of the children’s university to learn more about the event and how to register.

Data Voids and Spectral Analysis: Don’t Be Afraid Of Gaps!

During a workshop on time series analysis on paleoclimatic data, I was asked how data gaps affect the results of spectral analyzes. The good news is, the expected climatic cycles such as Milankovitch cycles do not shift when the time series has gaps. However, it is important that the data is properly pre-treated before being examined with a method of spectral analysis. Continue reading “Data Voids and Spectral Analysis: Don’t Be Afraid Of Gaps!”