It’s a free app and currently only available for iPhone users. It will be available on the Apple app store in a few weeks and in the meantime, I’m looking for users who might help me test it. If you might be interested, please sign up here.
The Frost Lens app will send you notifications for 3 frost predictions: “mild frost”, “hard freeze”, and “clear” based on your location. I use several meteorological parameters to predict frost & freeze. Frost Lens will also give you hourly weather data & the frost prediction for the next 24 hours as well as the frost prediction for the next 4 days. Currently, you can only use this for 1 location.
I know many of us are now well into Spring and past frost danger in many areas, so I’m especially keen on finding some folks from the Southern Hemisphere. Even if you are currently in Spring, you can still help me test the basic functions of the app and get yourself set up for next Winter.
I’ve been meaning to build this app for years, primarily for myself. I moved into my first home with a yard here in Bay Area, California for the first time 4 years ago in early winter. I was so eager to plant fruit trees that within a week of moving in, I went out and bought 6 citrus trees and planted them that weekend. I grew up in Los Angeles and really had no clue about frost and how it could damage young trees. So just three days after I planted my citrus, we had several nights with heavy frost and in the coming weeks, I lost all 6 trees. Now, 4 years later I am still no expert but I’ve come a long way. And I still live in the bay area which means that every year I have considered the risk of frost for my garden. I still haven’t found any easy to use frost alert tools so I decided to create Frost Lens, the app.
I hope gardeners around the world will find this app as useful as it will be for me. I’m happy to answer any questions you might have and I would absolutely love to get some testers from the Growing Fruit community.
3 Likes
I hate to be that guy, but what is the prediction model you are employing. In the US we already receive frost and freeze warning regularly from 1 to several days in advanced. How do you plan to improve on the frost predictions. Also frost and freeze have ranges. this chart for example shows the temp vs development stage of several species. https://www.canr.msu.edu/uploads/files/PictureTableofFruitFreezeDamageThresholds.pdf . A prediction model tuned to apples has little benefit to the growers of sweet cherries. And neither model will help with out knowledge of the bud development level, which is also dependent on the cultivar and in some cases influenced by the rootstock.
Well not to be a total debby downer might I propose an IoT device that can be deployed in someone fields that would over time create a location based variation predictions. After a period of data collection one might determine that one field or plot trends 2degrees warmer the regional actual data. Or in other word the closes weather station logged 30degrees f but data collected in the home garden regularly showed a +2 degree differences in the location. But for that to be predictive, you would need to track how accurate the forcasts have been for any general region and at what delta those predictions actualy become accurate 3 days 5 days 10 days is just a wild guese really.
Thank you for sharing thoughts. You make some interesting points. And yes I am working with several limitations in making this app.
First, Frost Lens isn’t reinventing the wheel when it comes to frost predictions. I worked with 2 different meterologists to figure out how to predict frost and the number 1 takeaway from both was its a difficult prediction with many parameters. Frost is a localized phenomenon as you might have alluded to as well and even how it presents varies by region.
For this app, I’m using a combination of 7 weather parameters that include all day temps, temps at sundown/night, cloud coverage, wind speed, etc to determine the likelihood of frost or freeze. It’s not a perfect model but so far it has accurately predicted frost for me in my specific location and several others around the bay area, portland, and coastal california for this past winter. That said, it’s not perfect.
Which brings me to my next point: This is a free app. It’s free because I’m using free weather data. Few services offer comprehensive weather data for free for use like this in an app. So I am limited to what is available for free. I looked into what paid data would cost and essentially I need charge more than $100 annually per user and that does not make sense for this purpose. So it’s free using free data and intended to be a tool for the average home gardener.
Speaking of which, I should clarify that this app best serves a home gardener and is not enough for a commercial grower. This is intended for hobbyists and part time plantspeople who wish they could get an alert for frost or freeze without having to do the legwork of looking it up, or (as I found for my area) trying to get a more accurate frost prediction for my specific city instead of the broad frost prediction that my local weather channel does for my very large county. Frost Lens uses your degree coordinates to pull data and present you with a prediction. It will be at the mercy of that weather’s current accuracy as well. And so far, in my opinion, it predicts frost for the next 3 days with more than 80% accuracy but this accuracy decreases past 3 days. And I do think that this is because of the limiting factor of my free data.
I think the IoT device is a fantastic idea, especially for commercial growers. I also imagine that service could cost a farmer several hundreds if not thousands of dollars a year. I have family in the almond business in california’s central valley and they pay a lot of money for water tracking devices that help them with “smart irrigation”. Perhaps someone out there is already doing the likely costly r&d for a frost IoT device as well, and I look forward to when such technology will also be available to home gardeners at a reasonable cost.
1 Like