Frost history / Atlanta, MI
Atlanta, MI: frost dates by year
Last spring and first fall frost at NOAA station USC00200342, 2005 to 2025, measured from daily minimum temperatures.
Elevation 879 ft, 45.005, -84.141. Download the data (JSON) · NOAA daily records
Summary at 32°F
| Last spring frost | First fall frost | |
|---|---|---|
| Median | May 25 | Sep 24 |
| Earliest | Apr 27 (2024) | Aug 31 (2009) |
| Latest | Jun 13 (2012) | Oct 18 (2021) |
| Trend | 12.3 days earlier per decade | 16.4 days later per decade |
| Years with data | 20 | 21 |
At 28°F (hard freeze) the median last spring date is May 18 and the median first fall date is Oct 9.
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 20 | Oct 9 | May 10 | Oct 9 |
| 2024 | Apr 27 | Oct 9 | Apr 27 | Oct 10 |
| 2023 | May 27 | Oct 12 | May 18 | Oct 23 |
| 2022 | May 23 | Sep 24 | May 9 | Oct 3 |
| 2021 | May 30 | Oct 18 | May 30 | Oct 23 |
| 2020 | May 14 | Sep 18 | May 14 | Sep 20 |
| 2019 | May 14 | Oct 5 | May 11 | Oct 25 |
| 2018 | May 13 | Oct 5 | May 11 | Oct 18 |
| 2017 | May 20 | Oct 12 | May 20 | Oct 26 |
| 2016 | May 23 | Oct 9 | May 18 | Oct 10 |
| 2015 | Jun 3 | Oct 1 | May 24 | Oct 2 |
| 2014 | May 18 | Sep 14 | May 18 | Oct 11 |
| 2013 | Jun 5 | Sep 15 | Jun 3 | Sep 17 |
| 2012 | Jun 13 | Sep 19 | May 18 | Sep 27 |
| 2011 | May 28 | Sep 7 | May 27 | Sep 16 |
| 2010 | May 19 | Sep 15 | May 11 | Oct 4 |
| 2009 | Jun 5 | Aug 31 | May 31 | Oct 1 |
| 2008 | May 29 | Sep 10 | May 29 | Oct 5 |
| 2007 | Jun 6 | Sep 16 | May 21 | Oct 12 |
| 2006 | Jun 11 | Sep 10 | May 23 | Sep 29 |
| 2005 | no data | Sep 24 | no data | Oct 9 |
"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 1991–2020 frost probabilities.