Anticipating early & late frosts

Learning to understand and predict local weather is a really important skill for properly managing our diversified farm. There are so many ways that weather conditions can hinder or help our work, and general forecasts don’t always cover what we need to know. Case in point,  predicting when we’re going to have a frost in our valley regardless of whether it’s regionally expected.This skill can mean the difference between significant crop loss and success at extending/completing our growing season. You’d never know it from the muggy, high-80s conditions this week, but we saw last week’s three nights of frost coming.

We’ve known for years that our location gets a lot colder than the “official” National Weather Service temperatures recorded at the Columbia airport. That’s only about 20 miles away, but it’s a high, flat, open location with lots of concrete to keep things warm, nearly the opposite of our narrow, protected valley. We rely on two well-proven rules of thumb to forecast our own frosts:

1) When the forecast calls for low 40s or below and a clear, calm night, we’re going to get frost. The NWS can record near mid-40s at the airport and we can have frost.

2) When a cold front or air mass is forecast to move into our area, the NWS forecast will almost always continue to drop from the early expectations 4-5 days out. We’ve seen this over and over: the final forecast temperature for a given chilly night will be 4-5 degrees colder than initially forecast.

Both of these rules were well-demonstrated by the forecasts & conditions leading up to the three nights of mid-May frost we suffered this past week:


Look at Thursday night, whose forecast low dropped from 42 to 35 in the course of 4 days. Even a final temperature of 42 would likely produce a frost for us, but seeing that 4 days out really made us take notice that this would be serious, as we knew it would drop. Friday night went from 41 to 37, and another solid frost. Saturday night went from 46 to 44 to a sudden 41 late that afternoon, prompting a scramble to re-cover crops that might have been safe at 44, and indeed we had a third, if lighter, frost that night (see our Twitter feed for photos from these three nights).

The actual recorded low temperatures at the airport for those three nights were 37, 38, & 43. Notice that all were higher than the final forecast low, despite other weather stations throughout the region recording new record lows (which were 33, 35, 35 for Columbia). This encourages our opinion that the Columbia NWS recording location stays artificially warm and isn’t a reliable source for frost prediction in a county full of topography and sheltered areas.

We’ve also found that the NWS St. Louis office, whose territory includes (barely) Boone County, tends to downplay frost probabilities too much for a region full of topography and sheltered areas. For this past week’s event, the NWS Kansas City office started discussing frost potential publicly several days in advance, while St. Louis only picked up on it at the last minute. We’d told our workers the Sunday before that this would be a concern late in the week.

We hope CSA members appreciate our dedication to weather forecasting when they’re enjoying the strawberries, squash, potatoes, kohlrabi, and other sensitive crops that we saved through skill and caution, and all the other warm-weather items such as tomatoes and peppers that we held off on transplanting knowing that a stronger-than-forecast cold spell was coming. The difference is only obvious when we aren’t as good at our jobs…as with the pea shoots this time.

Learning which crops to cover when is a related skill to discuss in some future post. We’ll note one big surprise this time, though: Peas are generally very frost hardy, so we didn’t even think about trying to protect them. Our pole peas came through just fine. However, the peas that we had planted for pea shoots showed a great deal of damage after the 3rd morning of frost, and some of the Sugar Anns looked a little ragged, too. Why? Not sure; maybe a varietal difference, maybe because the overcrowded spacing of plants intended for shoots led to less healthy plants, maybe some early aphids weakened some of the plants. Or maybe three obnoxious, near-if-not-record-cold nights in mid-May just have consequences that we can’t always predict.

One thought on “Anticipating early & late frosts

  1. Thank you for this information. I am going to keep this with my gardening journal so I can refer to it this fall and most likely, from now on.