Coincidentally, before the newest NOAA news release came out, I was looking at and thinking about the climate data that many gardeners use to plan their planting schedules: the frost dates. The Farmer's Almanac has a handy frost date calculator. Just plug in your zip code and the widget reports your frost dates as observed by your nearest weather station. But there's some "fine print." The last spring frost date that the Farmer's Almanac provides is for when there is a 30% probability of 32F temperatures occurring after that date. Likewise, the first fall frost date is for when there is a 30% probability of 32F temperatures having already occurred. After all, gardeners want to know when freezing weather is no longer likely in the spring so that they may safely set out cold-sensitive plants as well as when it is no longer safe to grow cold-sensitive plants outdoors in the fall due to imminent freezing.
I wonder why the 30% probability level was chosen. Farmer's Almanac does not state their rationale and it seems like an arbitrary place to draw the line. For my zip code in DC, this 30% probability of frost occurring translates to April 7 in the spring and October 29 in the fall. For spring 2019, we had a close call on April 1 when temperatures fell to 34F...but the actual observed last spring frost of 32F was actually back on March 8, a month before the Farmer's Almanac last spring frost date. And for fall 2018, the actual observed first fall frost was November 11, two weeks after the Almanac's first fall frost date.
Note that the data set used to make these calculations spans from 1981-2010, which by definition does not include some of the hottest years on record. I think it will be fascinating to see what happens to the frost date calculations when the 1991-2020 data set is released.
Here's how to access the full Climate Normals report from the National Oceanic and Atmospheric Administration (NOAA):
Visit the U.S. Normals Data (1981-2010) Map website.
In the left sidebar navigation, make sure that only "Annual Climate Normals" is checked in the Layers tab. (The other options are interesting too and I encourage you to explore them.)
On the map, enter your zip code in the field labeled "Search for a Location" and click the magnifying glass icon. The map should zoom in to a blinking red dot indicating the approximate location of your zip code. There may also be one or more other dots on the map.
If there are no other dots besides the blinking dot visible, click the minus sign in the upper right corner of the map to zoom out until other dots are visible. These dots are the locations of your local weather stations.
To choose your nearest weather station, click the wrench icon in the left sidebar navigation to open the Map Tools. Click the "Identify" tool, then click on the dot representing your nearest weather station. The left sidebar navigation will automatically shift to the Results tab, displaying information about your selected weather station.
Click "View Station Details" in the left sidebar navigation. It will open in a new browser tab.
Scroll down to the drop-down menu that reads "Select Year" and choose "2010" (it's the only option they give you at this time) and click "View Data." The report will open in yet another browser tab.
Save a copy of this report for future use. Sometimes, the website goes down. And some people in positions of political power wish to limit the people's access to climate data.
On this map, the approximate location of my zip code is the red dot (above Maryland Ave NE) and my nearest weather station is the grey dot within the National Arboretum.
Here is what I think is most interesting about NOAA's frost date reporting. Remember that our most recent actual observed last spring frost was March 8, 2019 and our most recent actual observed first fall frost was November 11, 2018? From the calculations presented in the table for the National Arboretum weather station, the probability of frost occurring after March 8 exceeded 90%. We really should have seen more freezing weather this spring, but we did not. As for last fall, the probability of frost occurring before November 11 was greater than 70%.
Again, I await the release of the 2020 data set to see how much these calculations change.