Just in case the PRISM map products are not trusted for one’s area of interest (i.e. wrong elevation gradient or lake effect gradient), or if my use of archived temperature didn’t correctly weigh missing data, I created maps from data from a couple other independent sources.

First, I calculated annual air temperatures from the newly released 1 km temperature grids from WorldClim Version 2: http://worldclim.org/version2

The reference period for this map is 1971-2000:
https://drive.google.com/open?id=0B3NKeT9nHj0ReFBIRTJIeWpFdFU

Then I hind-casted based on the annual temperature anomalies for different time frames from gridded data available at this site: https://data.giss.nasa.gov/gistemp/maps/
The reference period for this map is 1881-1910:
https://drive.google.com/open?id=0B3NKeT9nHj0RT0JReTN0R0x1Tms

The reference period for this map is 1996-2016*:
https://drive.google.com/open?id=0B3NKeT9nHj0RUWFVR0FmV05hMWc
*I intended to use the most recent 20 for a preview of the 1991-2020 norm, but used 21 by mistake. Perhaps, the middle 22, 1994-2016, would have been more appropriate to estimate the next 30 year norm. But I don’t feel like remaking the map.

Greg J. Schmidt
Ecological Site Inventory Specialist
Natural Resources Conservation Service
3260 Eagle Park Dr. NE, Suite 104
Grand Rapids, MI 49525
cell - (517) 285-7911
e-mail - greg.schmidt@mi.usda.gov
Ecological Site Descriptions - https://esis.sc.egov.usda.gov/