- [Voiceover] In this lecture we are going to study what bootstrap is, how to conduct a Bootstrap Monte Carlo Simulation by examples and let's make a summary. Bootstrap is a statistical resampling Monte Carlo Simulation technique which quantifies uncertainty in statistics by resampling from the original data. That is pulling yourself by your bootstraps as shown in the picture. This an example of simulating mean for one variable, given 20 porosity values. Make a histogram, shown here. Calculate is a mean of 20 porosity values it is 0.20, standard deviation is 0.12. Randomly draw 20 values from the given samples with replacement some values, maybe try them more than once and other values may never be try to calculate the mean and save it. Repeat the last step many times, let's do 1000 times, obtain 1000 means, make a histogram of 1000 means, calculate the mean of 1000 means, it is 0.21. Standard deviation is 0.08. This quantifies the uncertainty of porosity mean. This is an example of simulating correlation coefficient of our two variables given 20 paired data of porosity and permeability. Make a cross plot, calculate the correlation coefficient CC is equal to 0.7. Randomly draw 20 values from the given samples with replacement. Some values may be tried more than once and other values may never be, try to calculate the correlation coefficient and save it. Repeat the last step many times, let's do 1000 times, obtain 1000 CCs, make a histogram of 1000 CCs, calculate the mean of 1000 Ccs. It is 0.69, standard deviation is equal to 0.13. This quantifies the uncertainty of correlation coefficient. This is an example of simulating oil in place. Oil in place is equal to gross pore volume times net to gross times net porosity times oil saturation. Uncertainty of gross pore volume can be quantified by geostatistical methods. Uncertainty of net-to-gross, net porosity, and oil saturation can be quantified by bootstrap. Randomly draw gross pore volume, net-to-gross, net porosity, oil saturation from the given samples with replacement, some values may be chosen more than once and other values may never be chosen, calculate the oil in place and save it. Repeat the last step many times, let's do 1000 times, obtain 1000 oil in places. Make a histrogram of 1000 oil in places. Calculate the mean of 1000 oil in places. It is 2100. Standard deviation is 700. This quantifies the uncertainty of oil in place. The early stage of uncertainty assessment is available in reservoir appraisal. Bootstrap is a statistical resampling Monte Carlo Simulation technique which quantifies uncertainty in statistics by resampling from the original data. Bootstrap assumes data are independent one from other and representative of the underlying population. When the above assumption is not applicable, the advanced bootstrap techniques should be considered.