July global surface TempLS up 0.045°C from June.

The TempLS FEM anomaly (1961-90 base) was 1.097°C in July, up from 1.052°C in June. It was the warmest July in the record, just ahead of 2023.

Once again there were difficulties with GHCN data posting. GHCN usually posts a single gz file, named ghcnm.tavg.latest.qcu.tar.gz, which should contain the latest files, with names including the date. That is good for scripting when it works, with a constant url for the package. But last month, I found that new such files were posted, but with old data. So I switched to another archive directory where files had the date in their names, and seemed to be more up to date.

But early this month, the archive was not getting new files, and the old system was better. I found also a directory called access which seemed ideal - it has the latest files, dated, but uncompressed. However, eventually the archive caught up, and meanwhile the access directory vanished. However, the number of non-US stations, which is my test, has stalled at about 2700, whereas last month it went fairly quickly to 3200 Still, 2700 isn't bad, so I'll post.

Here is the corresponding stacked graph:


Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.

June global surface TempLS up 0.062°C from May.

The TempLS FEM anomaly (1961-90 base) was 1.038°C in June, up from 0.976°C in May. It was the second warmest June in the record after 2024.

Results are late this month. The main reason is again GHCN. I normally download a file called ghcn4.tavg.latest.qcu.tar.gz, which untars to give a directory with data, all with a date in the name. Usually it is the latest, with sometimes a day or two lag. But on July 3, it stopped updating. The wrapper file showed a new date, but the names on the contents remained July 3, and the file contents didn't change. There were only 1176 stations outside the US, so I couldn't post.

The wrapper is useful, because I don't have to keep updating the URL. However, I found the archive which had the actual date-named files, and they were up to date.So I downloaded the latest one. I made an error which led, for a while, to me still using a copy of the old data, but when I had fixed that, the new file had over 3000 stations outside US - good to go. I'll have to cope with the need to update the URL.

On blog matters, I've been updating the mechanics of the latest data page, so things sometimes were not showing properly. Nearly finished now, I think.

Here is the corresponding stacked graph:


Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.

May global surface TempLS down 0.04°C from April.

The TempLS FEM anomaly (1961-90 base) was 0.962°C in May, down from 1.002°C in April. It was the second warmest May in the record, behind 2024.

It is a bumpy ride with GHCN V4 lately. The file posted on May 30 was empty, and then nothing postd until June 4. At first there were few stations outside USA, but then there was a big jump to over 3000, high by recent standards.

Here is the corresponding stacked graph:


Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.

April global surface TempLS down 0.021°C from March.

The TempLS FEM anomaly (1961-90 base) was 1.005°C in April, down from 1.026°C in March. It was the third warmest April in the record, equal to 2020 and behind 2024 and 2025.

Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.

March global surface TempLS down 0.023°C from February.

The TempLS FEM anomaly (1961-90 base) was 1.018°C in March, down from 1.041°C in February.

Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.

February global surface TempLS up 0.15°C from January.

The TempLS FEM anomaly (1961-90 base) was 1.041°C in February, up from 0.89°C in January. It was the fifthwarmest February in the record, just behind 2024.

I mentioned an intention to compile monthly averages from GHCN Daily, to try to overcome the fitful posting of GHCN V4 monthly lately. I tried that as February results came in, but it was actually slower.

Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.

January global surface TempLS up 0.04°C from December.

The TempLS FEM anomaly (1961-90 base) was 0.886°C in January, up from 0.845°C in December.

There was again a strange sequence in the GHCN V4 monthly reports. Until last August, by about 6th of the month I would have about 3200 or so land stations reporting outside the US (ie ROW), which was enough to publish. After August, data was much more sketchy, with countries like China missing, and struggling to reach about 2600 stations. By about 10th month, I could get a reasonable result with merging JMA (Japan) data.

In January, there was suddenly an early flush of ROW data, even more than before August. But this month, by the 7th, there were only 1164 ROW stations. Then the situation improved, and now there are 3194, almost back to the old days. So I have confidence in publishing.

However, I looked into using GHCN Daily data, which is where GHCN-M comes from. I found, when GHCN-M had only 1164 stations, that the number of monthly averages I could get from GHCN-D was very dependent on how many days I insisted the month should have. 25 did not do much better than 1100, but if I relaxed to 20, I could get up to 3000. I presume that the last days of the month are coming in slowly.

So I may switch to daily to get a better idea of what is happening, and maybe post earlier. However, for now I am still using GHCN-M (+JMA)/

Here is the temperature map, using the FEM-based map of anomalies. Use the arrows to see different 2D projections.

As always, the 3D globe map gives better detail. There are more graphs and a station map in the ongoing report which is updated daily.