Winters Main: Here you'll find summaries of winters, and links to storms within them. This starts with the Winter of 1892-93, and continues through the Winter of 1997-98 . Currently, there is nothing in this section.

Storms Main: Here you'll find summaries, totals, and (hopefully) maps for storms from 1893 to 1998. Some storms are major, others not so much, but they're all unique and interesting. When this database is completed (2012?), there will be over 1,000 storms contained within this section.

Case Studies: Case studies on events of the past.


This site is intended to be a reference for past weather, particularly winter, in the Eastern Lake Erie region (From the PA-OH line towards Buffalo). Data from about 1893 to present is by and large from NWS Cooperative Observer (COOP) Records. Data before that is almost exclusively from newspapers or the US Signal Service, which began in 1871.

COOP records are not perfect, or anywhere close to it. Most stations (I've got 129 currently saved, there are thousands in all) have periods of missing days (days where there were no observations). They can range from a day or two, to months, to years, to decades. Sometimes only one field is missing. Sometimes there are only temperature observations. In the case of my nearest COOP (Jamestown 4 ENE), from the early 1980s to 2000, observations were not taken on weekends or holidays. The missing data causes headaches for anyone creating totals or averages. In major cities that have dedicated (their job) observers, there is not this problem (we'll not talk about BWI or DCA...), so complete records can be made and, generally, be relied upon. COOP data is not kept to these stringent regulation, however, it's the best we've got, so we'll use it.

I have to make this one note about snow measuring, especially considering this site deals with largely winter weather. The average person would naturally think that how one measures snow is not a big deal (You just stick a ruler in the ground, right?). It is a big deal. For a very long time, snowfall obs were generally taken every 24 hours by just sticking a ruler in the ground before clearing the area for the next 24 hours. It is not that simple today, however. Today, snow measurements are taken every 6 hours on a snow-board (usually just a flat piece of wood). The snowboard must be cleared every 6 hours and it should be located in an area that does not see much wind. Snow measurements may be made within the 6 hours before clearing (for things like hourly snowfall rates), but the snowboard may not be cleared until the 6 hours is up. This makes a big difference in amounts that are measured today and of the past. I think this could also explain the dramatic increase in snowfall over the past few decades (especially the last one) in the Great Lakes. Anyone that has seen lake effect snow knows that it is usually fluff, unlike the cement of East Coast snowstorms. This fluffiness can lead to rapid settling of snow in a short time (sometimes faster than it falls). Now when you think about this, you can see how the 24 hour rule could result in much lower amounts than the modern 6 hour rule. I've done a 'reconstruction' of sorts to account for this, but it, in my opinion, is not accurate enough to use for anything. The problem also is that I don't know when we transitioned from 24 to 6 hours. This makes it even more hard to compare years. So in all, I just ask that you keep that in mind when looking at anything pertaining to snowfall in the past, here or anywhere.


Simply, you can use any of my work in any way you want as long as you attribute it to this site. All of my work is under the Creative Commons Attribution-Share Alike License or is public domain.

Credits and Sources

The vast majority of the data on this site comes from 4 sources:

The NOAA Central Library Data Imaging Project: This is where I get the "Daily Weather Maps" that you see with most storms. Started by the Signal Service in 1871, eventually taken over by the Weather Bureau and then the National Weather Service, maps were produced daily based on observations from all over the country. They were digitalized by the NOAA Central Library Data Imaging Project.

The Utah Climate Data Center: The most important site. This is where I get totals for nearly ALL of the storms here. They digitized the data from most COOPs into a simple format and I couldn't thank them more.

IPS- Record of Climatological Observations: I don't use this as often as the above, but it also has COOP data, but in the original forms. It makes it sort of difficult to use, but it is very useful when verifying things in the UCDC stuff.

The National Weather Service WFO in Buffalo: My home NWS office and source for a lot of Buffalo and Rochester data. They've put together far, far more historical stuff than most WFOs.

Community content is available under CC-BY-SA unless otherwise noted.