The previous warmest January was in 1956 with 18.8C.
In November, December and January, the temperature had come in at 17.7C, or 2.2C above average.
The hot early summer of 1934/35 - long used to compare warm seasons to - was 17.C.
"The main reason for the hot early summer and the marine heatwave around New Zealand over this period was a strongly positive Southern Annular mode, causing the storms whirling around the Southern Oceans to contract towards the Antarctic continent, with no cold outbreaks into the Tasman Sea and New Zealand area," Salinger said.
"The frequent anticyclones tracking over the Tasman Sea across the South Island and to the east have allowed the seas to heat up, producing the marine heatwave."
A marine heatwave also happened in 1934-35, but this period had been warmer by 0.4C, because of global-warming-producing climate change, Salinger said.
The New Zealand temperature series - better known as the seven-station series - was developed in 1981, as part of Salinger's PhD work. The stations are at Auckland Airport, Masterton Airport, Kelburn in Wellington, Hokitika Airport, Appleby near Nelson, Lincoln in Canterbury and Musselburgh in Dunedin.
"These are complete since 1909, and that is why Niwa uses the records from 1909, and they have been adopted as the standard," Salinger said.
"I have extended this record back to 1867 using a smaller network of available stations. So this is the warmest month on record, since reliable records commenced in 1867."
Early New Zealand scientist Sir James Hector established a network of 12 climate observatories, using Stevenson's screens and precision thermometers.
The self-registering maximum and minimum thermometers are made by Casella in London.
"New Zealand was the first country in the world where a standard instrumentation network was installed and measurements made that were 'robust'," Salinger said.
"For example, in the Australian states it was quite haphazard, with different methods of observing in different states. Only after federation and from 1910 onward does Australia have reliable records."
Salinger was able to crunch the data using six stations that recorded data from 1867 to 1880, four which were operating between 1881 and 1893, five that had been running since 1894, and today's seven that had been running together since 1909.
"Using the extended series back to 1867 then we can say this is the warmest month in over 150 years of records."