"Everything Geoscientists and Data Scientists Need to Know About Geostatistics"

Prof. Michael J. Pyrcz, Ph.D., P.Eng. Hildebrand Department of Petroleum & Geosystems Engineering University of Texas at Austin

This repository includes the lectures for this short course. The suppliemental materials with workflows, demonstrations, hands-on and data are here: https://github.com/GeostatsGuy/2DayCourse_Exercises.

Michael Pyrcz is an associate professor at the University of Texas at Austin. He teaches and consults on the practice of geostatistical reservoir modeling and conducts research on new geostatistical methods to improve reservoir modeling and uncertainty for conventional and unconventional reservoirs. He has published over 40 peer reviewed technical articles, a textbook with Oxford University Press, and is an associated editor with Computers & Geosciences. For more details see www.michaelpyrcz.com or follow him on Twitter @GeostatsGuy.

Class will be accessible to geoscientists and data scientists with no previous experience with geostatistics. We will build up from data integration to spatial estimation and simulation along with uncertainty modeling to support decision making. After completion the students will understand: (1) the benefits and uses of geostatistics, (2) the common spatial and uncertainty modeling workflows, (3) how to better integrate their domain knowledge into the geostatistical model.

For the following 2-day class outline it is assumed that: (1) the class will be taught in English without translation, and (2) 1/3 of time will be guided hands-on practice

1. Exploratory Data Analysis (a) Sampling theory, stationarity, data debiasing (b) Random variables and functions (c) Univariate statistics, multivariate statistics (d) Geostatistics and big data analytics

2. Spatial Data Analysis (a) Trend modeling (b) Variogram calculation, interpretation and modeling (c) Scaling relations (d) Training images

3. Estimation (a) Interpolation (b) Kriging estimation (c) Predrill prediction

4. Stochastic Simulation (a) Simulation paradigm (b) Gaussian simulation (c) Minimum acceptance and uncertainty checks

5. Uncertainty Management (a) (Spatial) bootstrap (b) Model post-processing (c) Uncertainty workflows

6. Machine Learning for Subsurface (a) Estimation variance (b) Multivariate analysis (c) Decision Tree

I am open to revisions to the outline to accommodate student learning needs.