mek6800d2

Except that doing so does NOT avoid signed overflow. You're converting effectively w-1 bits of precision (plus a sign bit) to w bits of precision, doing the calculation, and then converting w bits of precision back to w-1 bits of precision -- that only works for a limited range of input values.

Using int32_t and uint32_t, try adding -2147483648 and -1. The arithmetic result should be -2147483649, but a program converting back and forth between signed and unsigned will produce 2147483647 -- wrong answer and wrong sign.

I didn't read the blog post you linked to. It appears from the date in the link that the guy analyzed temperature data from 1890 to 2010, roughly. He threw out out data from weather stations that hadn't been continuously in service for at least 100 years -- to "stabilize" the dataset. This implies that NASA (or whoever compiled the data) hadn't already analyzed and corrected the data to ensure that the readings (i) over the 120-year span and (ii) from disparate locations and different instruments were commensurable. If so, limiting his exercise to 100-year-plus stations provides a surface veneer of illusory "stability", but doesn't speak to the accuracy of the reduced or full dataset -- and the accuracy is key to conclusions that can be drawn from analysis. (And if NASA did analyze and correct the readings in the full dataset, then the guy is cherry-picking.)

The guy then drastically restricts the dataset to only include rural weather stations, to less than a quarter of the original set of 100-year+ stations. At the least, there has to be some accounting of geograpical distribution in the analysis. The bigger question, though, is why would he remove the urban heat islands? Does their effect on the climate stop at the city limits, with a partition rising straight up into space? The whole concern about climate change is whether human activity (which certainly includes urban heat islands) will eventually lead to irreversible climate change that will have catastrophic effects.

(If folks feel "catastrophic" is alarmist, I suggest they Google "nine planetary boundaries", of which climate change is but one of the interdependent boundaries, for some truly alarming reading.)

Anyway, someone more knowledgeable than me can perhaps provide a better perspective on whether any climate-/weather-related elements have more of a role than temperature in affecting the initiation and spreading of forest fires. I'm particularly thinking of low humidity and drought caused by climate-change-induced storms, winds, etc. Are forest fires less likely to occur in cooler regions experiencing low humidity and drought?

Aah, starik36 and Excel Hero have gotten me interested in researching further; thank you, sincerely!