When looking at the 30 day trend, we pick out the top 15 SQL statements with the highest wait time to show taking into account the entire 30 days.
When you drilled into the 24 hour period on the 10th, that shows the top 15 SQL statements for that time period taking into account that 24 hours.
What I suspect happened between the 9th and the 10th is that some queries changed and we now see it as a different hash (if SQL Server sees it as a different query, we will also).
So, when you look at the 30 day trend, only a few of the queries from prior to the 10th that had been running with significant waits are represented because the new queries haven't run enough time to accumulate waits to get into the top 15 (thus not making the top 15 for the 30 day trend).
Were there any changes that happened to the SQL code between the 9th and 10th? Even adding another space or carriage return will result in the SQL hash changing.
Clear as mud? 8 )