If you look at a specific data point on a graph showing 10 minutes of data, then look at the exact same data point on a graph showing 24 hours of data, you may see a different in value.

This is because data in the second graph is sampled. The values for the longer time period have been aggregated, so the specific data point you're interested now contains a summary of other points in close proximity. This form of data sampling makes sure that whatever the time range displayed, peaks and troughs are maintained on the graph relative to each other.