Making Choices

In a certain online supermarket, there are over 100 different varieties of salad dressing that can be bought. When faced with all that choice, most people stick to what they know and buy the same variety each time. Research has shown that too much choice makes us unhappy and frequently leads to no decision being made at all. As the number of options increases, the cost, in terms of time and effort, of gathering the information needed to make a good choice, also increases.

When it comes to seismic attributes, the same problem exists. When there were only a few attributes, it was easy to calculate them all and work with each of them. Nowadays, there are over 100 attributes that we can calculate so geoscientists are faced with a dilemma. Do they use the same attributes that they have always used or do they calculate new ones. If they calculate new ones, which ones do they select. As you might expect, they typically just use the ones that they have always used. 

However, geology changes and the way that seismic responds to each geology will be different. As a result, the most appropriate attributes will change from area to area. Applying the attributes that worked last time is not an appropriate method of working. Fortunately, Principal Component Analysis is here to help!

Principal Component Analysis (PCA) a linear mathematical technique to reduce a large set of variables (seismic attributes) to a small set that still contains most of the variation of independent information in the large set.   We use it to determine which attributes are the most important and which can be ignored in the final analysis.

By running PCA first, we can determine which are the most important attributes and then only use those in our calculations. This reduces the effort required and ensures that the only the most important attributes are used. As a result, we get a much better analysis of the results than if we only used the attributes that we used for the last project. 

Paradise, from Geophysical Insights, not only calculates your attributes, it applies PCA to determine the most important ones and then runs self organising maps to combine them into geological features of interest. Please get in touch if you would like to know more.