The bootstrap or resampling method serves as a means for estimating how a statistic's value would change when derived from random samples drawn from an infinite population.

The resampled (mean, CI (lower), CI (upper)) provides us with a range of plausible values of a point estimate (in your case that is the mean of aod/ai/ri, etc or other statistical estimates).

So one way of interpreting the result is if you have computed bootstrapping CI from several methods (for example, empirical, percentile, bias-corrected accelerated bias-corrected) is you can compare the sample mean and CI results from these different methods, and discuss the coverage probabilities of point estimated value by the CI constructed by several methods in your manuscript.

This can be one of the ways to interpret.

For your second question, CI (x,y) is actually defining the uncertainty of a point estimate. Imagine you calculated bootstrap sample for mean aod. Your results are 0.85 and CI (0.81, 0.88). So the CI represents the uncertainty that through several sampling you reached an envelope of value 0.81 and 0.88 within which the population (mean) lies.

One critical remark, however, is bootstrap is not a data correction method, it is not going to improve the point estimates. The quality of bootstrap depends upon the quality of data collected. Bootstrap estimates will be skewed if sample data is biased and does not accurately represent the population data.

Hope this information will aid you in your objectives.

Cheers!!! :)