Arctic Climate at Peak Summer 2025
Wildfire, Sea Ice and Greenland Updates
This post is an update of my mid-July post on wildfire and sea ice, here. I’ll post the July temperature and precipitation analyses once the ERA5 reanalysis for the month is online.
Wildfire
Northern North America wildfire at the end of July, in terms of area burned, is ahead of last year though not dramatically so. Most of the difference is in Alaska.
The largest fires at the end of July by region:
Alaska: the three wildfires that have burned in excess of 50,000 acres (20,000 ha) were all located north of the Yukon River. The Klikhtentotzna fire west of Alatna was at 118,022 acres (47,752 ha), the Moldy fire near Huslia 72,843 acres (29,479 ha) and the Christian fire northeast of Ft. Yukon at 52,468 acres (21233 ha).
Yukon Territory: the North Ferry Hill north of Stewart Crossing was 13,000 ha (32,124 acres). In The Dawson area, the Rabbit Creek fire was at 12,496 ha (30,878 acres) and the Mount Leotta fire at 7,161 ha (17,695 acres). Also of note is the Mount Dave fire near the Alaska Highway just east of the Alaska-Yukon border at 5,997 ha (14,819 acres).
Northwest Territories: the Laird Complex in the southwest NWT, south and east of Fort Simpson, consisted of two separate fires with a combined area burned of 165,502 ha (408,964 acres). The ZF007-25 fire south of Great Bear Lake was at 53,932 ha (133,269 acres). These fires were not a significant threat to any communities at the end of the month.
Based on available news reports and remote sensing, early and mid-summer wildfires in Arctic Asia have not been extensive and what has occurred has been mostly, though not entirely, in western Siberia.
It’s worth noting that while high latitude wildfire tends to be “early summer loaded”, significant fire remains possible in August and very rarely into September when and where mid-summer has been warm and dry and especially when there is already significant wildfire on the land. Last summer, about two-thirds of the 1.6 million ha that burned in the Northwest Territories occurred after the end of July. In Alaska during the past 20 years, only twice (2007 and 2023) has August and September accounted for more than 20 percent of the total areas burned in a season.
July is the peak of the Alaska wildfire season in terms of area burned in about 44 percent of years since 1993 and is virtually certain to be again this year. The 633,000 acres (256,000 ha) burned is well above the 1993-2024 median but far short of the big years.
Lightning-started wildfires nearly always account for the vast majority of area burned in the boreal regions of North America. July lightning activity in Alaska and northwest Canada was only slightly below the 2012-24 median. As often happens in July, most of the lightning occurred on a few days during the first half of the month.
Sea Ice Data Availability Update
On July 29, for the third time in two months, the U.S. DoD and NOAA announced a different date for the termination of the Special Sensor Microwave Imager/Sounder (SSMIS) data that underlies some of the global passive microwave-derived sea ice data sets. In the latest pronouncement, DoD states SSMIS data will continue to be processed and made available until the satellites reach their end of life or September 2026. Obviously, at this point, neither DoD or NOAA has the slightest bit of credibility on this issue.
In spite of this latest flip-flop, the U.S. National Snow and Ice Data Center operationalized the switch to using the AMSR2 data as base input at the end of July, though the change is retroactive to January 1, 2025. The NSIDC announcement is shown below.
The European Ocean and Sea Ice Satellite Application Facilities (OSI SAF) announced on July 29 they were also operationalizing AMSR2 input data into their long term record, but on August 1 rescinded that change and will continue to use SSMIS data. The Japanese Aerospace Exploration Agency (JAXA) sea ice products don’t use the real-time SSMIS data and so are unaffected by the bureaucratic chaos.
Arctic Sea Ice
Arctic-side sea ice extent for July was about 25 percent lower than in June, which is a slightly larger than average decrease. In NSIDC data, the average extent was the sixth lowest for July.
During the first week of the month the Arctic-wide ice extent was either the lowest or second lowest in the satellite era (since 1979). By the end of the month (fig. 4), the sea ice edge was poleward of 1981-2020 median almost everywhere except the eastern Beaufort Sea. Arctic-side sea ice extent at the end of July is the fourth lowest in satellite record (since 1979).
While the split between lower ice extent in the Atlantic sector and higher extent on the Pacific side continued through July, by month’s end it wasn’t nearly as dramatic as earlier in the summer due more rapid ice loss in the Chukchi and Beaufort Seas. Figure 5 makes clear that the Beaufort Sea is the only region with unusually high ice extent, while the Barents and Kara Seas and the Canadian Archipelago all have the lowest ice extent on record for the end of July
The Northwest Passage, which connects to the Canadian Atlantic to the Bering Strait, was not yet open at the end of July because extensive ice remains in McClintock Channel. This is typical for this point in the season.
Greenland
Daily melt area on the Greenland ice sheet was quite variable in July, with below average melt area early in the month punctuated by several days with melt on more than 40 percent of the ice sheet.
For the month overall, the daily average melt area was very similar to last year.
Important and critical information that not only do you provide for the wider audience but also well done to make it clear and accessible. Thank you for your excellent work.