After analyzing the way scientists calculate the average earth temperature change (summing anomalies to daily 'normal' temperature, averaged over 30 yrs), and after hearing that every year during the last decade or so the earth temperature increase is breaking the record set the year before, it came to my mind that we may be witnessing not a continuously warming trend but something else.
As farmers, we have noticed that the growing season is slowly shifting, even in the tropics where we live. That may be more discernible in temperate zones. Let's say, the day January 1st will start a decimal of a degree warmer in 2017 than that of 2016. Assume that is not a warming, but simply an offset in the calendar-date (say, the temperature pattern that corresponded to January 2nd). The usual 30-yr averaged temperature of that day would not absorb such a shifting in temperature and we would treat Jan 1, 2017 as a positive anomaly. If this shift continues from then on, we would gather positive anomalies daily, easily breaking the 'temperature increase' record the year before. This may happen not necessarily because of a warming, but because the yearly temperature pattern has a minor offset with the calendar year.
If you observe this 'temperature increase' at one geographic point, it is easier to compare it with other evidence and discard such false alarms, but when you sum this up globally, we have no other way to verify it. I am sure somebody must have thought about this before, but I could not find a reference to it in usual places. Hope I stated the problem sufficiently clear, so that those of you who work on this issue on a daily basis could indicate if there is an error in this argument.