[stormer]

You are correct. However, the upward trend is even more pronounced if only one year’s data is used.

[wfu_deacons]

You can detect a "recent" trend by linear regression of data that is even a year old. I'm not a statistician but you may be able to make the scientific statement that the single month of data falls in (or outside) two standard deviations of the last year's data (i.e. is consistent with or not consistent with) the historical data. Tell me, did the data point for July 2007 fall above or below the regression line?