1) The global mean surface temperature is increasing. For instance the 10 years 2013-2023 are all hotter than every year prior to 2013.

2) However if you look at the 1.88% of the earth's surface that is the continental United States, the randomness provides a stronger component of the variation. This is partly due to having a smaller sample. But also there's conservation of energy constraints if you take the whole planet, so the variation has to be heat moving into the oceans.

3) For the continental US, 1934 and 1998 were within error of the same temperature. NOAA has 1998 warmer.. NASA, apparently, has them the other way around.

4) As improvements and error corrections come to light, the rankings of years might change.