Frost history / Harlem 20S, MT
Harlem 20S, MT: frost dates by year
Last spring and first fall frost at NOAA station USC00243934, 2001 to 2025, measured from daily minimum temperatures.
Elevation 2,695 ft, 48.238, -108.678. Download the data (JSON), NOAA daily records
Summary at 32°F
| Last spring frost | First fall frost | |
|---|---|---|
| Median | May 28 | Sep 26 |
| Earliest | May 6 (2025) | Aug 4 (2008) |
| Latest | Jul 21 (2006) | Oct 6 (2025) |
| Trend | 14.5 days earlier per decade | 8.8 days later per decade |
| Years with data | 16 | 16 |
At 28°F (hard freeze) the median last spring date is May 13 and the median first fall date is Oct 5.
The trend is a straight-line fit through the years with a freeze. A positive spring trend means the last frost now comes later. Short or patchy records give noisy trends.
Chart
Last spring frost (32°F)
First fall frost (32°F)
Dots are single years. The gray line is the median, the green line the fitted trend.
Year by year
| Year | Last spring 32°F | First fall 32°F | Last spring 28°F | First fall 28°F |
|---|---|---|---|---|
| 2025 | May 6 | Oct 6 | Apr 25 | Oct 6 |
| 2024 | May 20 | Oct 3 | May 4 | Oct 13 |
| 2023 | May 19 | Oct 6 | Apr 23 | Oct 6 |
| 2022 | May 22 | Sep 21 | May 22 | Oct 7 |
| 2021 | May 23 | Oct 2 | May 12 | Oct 12 |
| 2019 | no data | Sep 28 | no data | Oct 1 |
| 2017 | May 29 | Sep 24 | Apr 20 | Oct 8 |
| 2014 | Jun 12 | no data | May 14 | no data |
| 2013 | no data | Oct 2 | no data | Oct 4 |
| 2012 | May 29 | Sep 12 | May 11 | Oct 6 |
| 2011 | May 14 | no data | May 14 | no data |
| 2010 | May 26 | no data | May 13 | no data |
| 2008 | no data | Aug 4 | no data | Aug 4 |
| 2007 | Jul 11 | Aug 11 | Jul 10 | Aug 11 |
| 2006 | Jul 21 | no data | Jun 25 | no data |
| 2005 | Jun 14 | Sep 7 | May 30 | Sep 28 |
| 2004 | Jun 18 | Oct 1 | May 13 | Oct 1 |
| 2003 | May 20 | Sep 18 | May 7 | Sep 18 |
| 2002 | Jun 4 | Sep 22 | Jun 4 | Sep 23 |
| 2001 | no data | Oct 3 | no data | Oct 5 |
"none" means the station recorded no freeze at that threshold in that half of the year. "no data" means more than 10% of the season's daily readings were missing. Years with no usable data are left out. See the methodology for the rules, and the ZIP lookup for 1991 to 2020 frost probabilities.