Probability of a value of x, given a mean value of x bar and standard deviation s

AI Thread Summary
To determine the probability of the Edwards aquifer dropping to between 660 and 641 ft, a normal distribution approach is insufficient due to the dependency of water levels over time. Instead, a time series model should be developed based on historical data from well J-17. This model allows for a more accurate assessment of the likelihood of extreme events, such as reaching a water level below the specified thresholds. The concept of extreme value theory can be applied to evaluate the probability of such rare occurrences. Accurate modeling is crucial for understanding the risks associated with drought restrictions.
moonman239
Messages
276
Reaction score
0
Let's say I monitors my local aquifer (Edwards) using well J-17 over the past 8 summers (not counting this summer). I found that the mean water level/day is 680 ft above sea level, with a standard deviation of 30. Assuming a normal distribution, how do I find the probability that the aquifer level will drop to between 660 and 641 ft sometime this summer, at which time Stage 1 drought restrictions are in place?
 
Physics news on Phys.org
Also, how do I find the probability that the level will be exactly 641 ft?
 
Assuming a normal distribution won't do you any good. The water levels in the aquifer at different days are certainly dependent so you need to model them as a time series. One you have estimated a suitable time series model from your observations, the question of how likely the water level is to drop to below a certain threshold can be answered in different ways.
One thing which usually comes up in this context is that the target region for which you want to estimate the probability is well outside the typical range of the time series under consideration and reaching it is therefore considered an extreme event, which then leads to the discipline called extreme value theory for time series.
 
Back
Top