A few weeks ago I saw that the hi/low temps listed in the 10-day forecasts only roughly corresponded to the plotted values. Now I notice that the hi/lo for a given day doesn't even correspond to that day. The hi/lo for one day seems to correspond to the hi for that day and the lo for the next. See image below, where hi/lo for Tuesday is 93/61F, but the 61 actually comes (only roughly, one remembers) from Wednesday. So the low temp for Tuesday actually happens on Wednesday morning. Data management is not a strong point for the WU folks, I gather.