The new hardiness map’s obsolete! The new hardiness map’s obsolete!

I posted back in January about my excitement about the update of the USDA Hardiness zone map.   While I acknowledged some of the shortcomings of the new map, I was excited because it was a big improvement over the old map, both in terms of content (more recent temperature data included) and presentation (interactive search features, better graphics).


Dr. Nir Krakauer, assistant professor of civil engineering in The City College of New York’s Grove School of Engineering, was less impressed, however, and has essentially declared the new hardiness map dead on arrival.  Dr. Krakauer conducted his own analysis of climatic data and determined that much of the country is already a half zone to one full zone warmer than the new USDA map.  Why the difference?  The USDA zone estimates were based on a simple average of annual minimum temperatures from the past thirty years.  Dr. Krakuar’s applied a regression analysis to account for more recent warming.  One of the key observations of the new analysis is that winter minimum temperatures are warming much faster than average temperatures, leading to the new hardiness zone projections.


So, what do we do with this information?   All of the limitations that we associate with hardiness ratings still apply.  Most landscapes experience micro-climates that may be slightly colder or warmer than the surrounding area.  Plus, winter hardiness is just one piece of information in plant selection (we also need to also consider sun exposure, pest pressure, soil factors – drainage, soil pH, and so on).  Most importantly, average temperatures don’t kill plants, extreme temperatures do.  So whether we look at long-term averages or try to account for the most recent trends; the extreme 1 year in 5 or 1 year in 10 is what will cause issues.  For example, if Krakuar’s projection is right and we’re a zone warmer on average, then the greater Lansing area is now zone 6b (0 to -5 deg. F).  We’ve been colder than that in 2 out of the last 4 years.


The press release that accompanies Dr. Krakuar’s study starts with an attention-getting line: “Gardeners and landscapers may want to rethink their fall tree plantings.”  But really things haven’t changed that much from the past.  How homeowners and landscapers use zone information is still a matter of risk tolerance.  If I chose a zone 5 plant and we’ve really warmed to zone 6, I’ve lost nothing.  If I gamble and plant zone 6 plants and we drop back down to -10 or -15 deg. I may be looking for replacements.  And, of course, there’s the tendency of many gardeners to ‘push their zone’, so we may have folks trying to grow zone 7 plants in Lansing.  Which is fine – as long as people understand the risk.  Actually I think a better approach to hardiness zones may be a probability rating: Are you willing to take a 1 in 10 chance your plant won’t make it? 1 in 4? 50-50?  As new climate data are added to the model, garden centers could post and update the odds, sort of like the latest line at a sports book in Vegas.  Those that want the ‘tried and true’ can stick with the old zone 5’s; those that want to live on the edge can pay their money and take their chances.

5 thoughts on “The new hardiness map’s obsolete! The new hardiness map’s obsolete!”

  1. Dr. Karkauer’s study and the headline drawn from it are overreaching from my perspective as an agricultural climatologist who is also a Master Gardener. Indeed the climate has been warming, substantially enough to be reflected in phenological data as well as an increasing growing season length. And that warming has been more apparent in the latter half of the period of record the USDA used for their study. However, insofar as winter temperatures, particularly when associated with a single extreme event, impact survivability of a plant, we are not as advanced as Krakauer suggests. Cregg’s suggestion to view hardiness zones from a probablistic perspective is an excellent idea that addresses the fact that any area can still have a winter as, or even more extreme, than your hardiness zone suggests. One’s ability to assume risk would be proportional to the investment. If I lose a $10 perennial after a year or two, no big deal. But if I lose a tree, especially after years of growth after an expensive start, that is another issue.

  2. Thanks for that link! Does anyone know why the USDA uses means instead of medians? For extreme weather separated by years, the means are really uninterpretable, where the median just says ‘50% of the time the lowest temp is in this range’. Then they could add quartiles, and the betting strategy would be clear (to follow the argument). Actually, quartiles would be really handy to have on hand. Also, comparing the low temp in your own microclimate would leas to more reliable conclusions. For example, if USDA says that 50% of years the low temp is between 0 and 5 F, and your observation is that the low temp in your garden has not gotten that low for 10 years, you would feel better assuming you have a warmer microclimate. But a mean low temp is misleading. It could be one year with -15 and 9 years with nothing below 5. It is all about the usability of the data.

Leave a Reply