Climate science investigators Cowtan and Way  updated the HADCRUT4 global temperature time series to properly account for areas that had little representation, incorporating regions such as Africa and the Arctic using geospatial kriging techniques . The Arctic in particular has shown considerable anomalous warming that hadn’t been captured in that time series. They have a methods page that contains details here.
This impacts the CSALT model by adjusting the temperatures upward in the last few years to better match the movement out of the “pause” regime that the fluctuation components plus relentless CO2 forcing have been predicting would occur. See Figure 1.
The last year or two is still lagging in intensity but the trend is recently upward.
Incidentally, I believe that Kevin Cowtan is the Kevin C who contributed the SkS trend calculator linked in the sidebar. An online commenter complained that Cowtan was a chemist specializing in X-ray crystallography, as if that was a bad thing. I did diffraction modeling for my PhD and I can tell you that this is puzzle solving at its finest; trying to piece together our understanding of reality using limited data — by inverting wave interference diffraction patterns — prepares one for just about any challenge.
Fine work along with a companion piece here .