The weird spike is not from instrument tampering, time of day, or sunlight vs. clouds vs. fog. This looks like a classic frontal passage (FROPA in METAR code). Since I was bored today, I did a little investigating like I used to do when doing consulting meteorology for court cases. First take the site closest to the incident which in this case was the summit of Mount Washington, which worked out great since there is a station there. Grab 48 hours worth of data and focus the search down to 24 hours with the window of interest inside of it. Then grab three triangulated stations around the site. In this case: Berlin, NH (North), Whitefield, NH (Northwest), and Fryeburg, ME (Southeast). All stations in the same time frame showed a warm up and cool down of about 10 degrees during the morning hours peaking at roughly 11 am EDT to noon and then declining for the rest of the day until beginning to warm once again. Looking at weather maps from this day and today and from personal observation living at the base of the mountain, I can say that this spike was from a quasi-stationary front from the north on the 28th. Warmer air rode ahead of the front with colder northern air moving down after the front. Things then warmed again on the 29th as a result of a warm front that passed this morning. And it should be noted that although our digital thermometer we use for the auto-road display (the display with the graph in question) showed a max temperature of about 62.5F, the actual high for the time period was 59.18F according to our observers sending out the METAR’s. There are reasons behind this but for another time. Hope this helps. And “Bill O” said it best; it isn’t really sharp when spread over 4-6 hours warming and about 2 hours of cooling. To see a sharp decline, check the graph after a severe thunderstorm.
Staff Meteorologist/Night Observer, KMWN (Mt Washington Obs., NH)