December 1917 through January 1918 still stands today as the coldest and snowiest December-January period ever recorded in Louisville, Lexington, Bowling Green, and several other locations across southern Indiana and central Kentucky. The 49 inches of snow that buried Louisville during those two months beats the 2nd snowiest December-January by more than a foot and a half!
The pattern that would bring the Ohio Valley such a miserable winter actually started to take shape before summer was over. Strong Canadian high pressure entered the northern Great Plains on the 8th of August, bringing temperatures in the 30s to Montana and Wyoming. Several additional highs followed through the rest of August and into September. On September 11 record lows were set in central Kentucky that still stand today: 42° at Lexington and Frankfort, and 44° at Louisville.
The parade of Canadian highs intensified in October, with snow accompanying the cold in the Great Lakes. Every station in Michigan reported snow during the month, with an even 12 inches falling on Ironwood on the Wisconsin border. On the morning of the 9th the freezing line dipped all the way down into Arkansas. A terrific surge of frigid air during the last few days of October brought sub-zero temperatures to the Rockies, sub-freezing temperatures to Texas and Louisiana, and middle 30s were reported in the Florida Panhandle. On the 30th Louisville had a high temperature of only 34° and Lexington didn’t make it above 30°, which is still the coldest October high temperature ever recorded in both cities. The month remains the coldest October on record at Louisville (#3 at Lexington, #5 at Frankfort, and #9 at Bowling Green).
A break in the weather came for the first 3 weeks of November, with generally mild conditions. A brief shot of cold air came through around Thanksgiving, followed by another brief warm-up into early December.
Then the cold returned with a vengeance.
A huge dome of high pressure gathered over northwest Canada during the first week of December and began to slide southeast on the 7th. Meanwhile, low pressure was forming over Texas and beginning to head to the east. On the 8th the low strengthened as it crossed the Tennessee Valley while the Canadian high pushed into the United States, spanning the entire length of the Great Plains all the way to the Texas coast. The low passing by to our south brought copious amounts of precipitation to the region, and the frigid high coming in from the northwest caused that precipitation to be in the form of very heavy snow. Over a foot of snow fell on the lower Ohio Valley, including 16.4” of snow recorded at the Louisville Weather Bureau station and about 10” at Lexington. The heavy snow on the 8th was accompanied by strong winds, gusting to 42mph at Louisville and 37 mph at Lexington, resulting in massive drifts.
Transportation was made extremely difficult by the heavy snow, resulting in shortages of fuel and other supplies. Country roads were impassable for several days and some rural schools were closed for a week or more. Young livestock suffered greatly and there was considerable loss, especially among pigs. Quail were reported to have starved and frozen.
Though temperatures started out around 30 degrees at the start of the day on the 8th, by midnight that night the mercury had fallen to -1° at Louisville and -4° at Lexington. On the 9th the monstrous high pressure system pushed into the southeast United States while the snowy low soared into New England while strengthening even further. Caught between these two systems, winds were even stronger on the 9th than they had been on the 8th with sustained winds up to 49 mph at Louisville and 50 mph at Lexington, and gusts to 52 mph at both locations. To add to the misery, those winds occurred while temperatures were mostly in the single digits, resulting in wind chills (using the modern formula) of 20 to 30 degrees below zero.
On the 10th wind speeds were “only” in the teens and 20s, but the high temperature at Louisville was 7°, and Lexington didn’t make it past 6° after a morning low of nine degrees below zero.
From the 8th to the 17th there was at least one station somewhere in Kentucky or Indiana reporting a sub-zero low temperature. Lowest temperatures around the area during the cold wave:
Alpha, KY -3° (10th) | Edmonton, KY -12° (10th) | Junction City, KY -16° (12th) | Louisville Downtown, -6° (9th) | St. John, KY -19° (10th) | ||||
Anchorage, KY -20° (11th) | Frankfort, KY -17° (11th) | Leitchfield, KY -15° (9th) | Madison, IN -15° (11th) | Salem, IN -23° (11th) | ||||
Bardstown, KY -14° (10th) | Greensburg, KY -12° (10th) | Lexington, KY -9° (9th-10th) | Paoli, IN -23° 11th) | Scottsburg, IN -22° (11th) | ||||
Beaver Dam, KY -19° (10th) | Irvington, KY -9° (9th-10th) | Loretto, KY -16° (12th) | Richmond, KY -9° (10th) | Shelbyville, KY -16° (9th, 11th) | ||||
Berea, KY -9° (11th) | Jeffersonville, IN -10° (11th) | Louisville Cherokee Park, -12° (11th) | Russellville, KY -13° (9th) | Taylorsville, KY -20° (11th) | ||||
Bowling Green, KY -10° (9th) |
Records set at our main climate sites December 8-11, 1917 that still stand today:
Location | December 8 | December 9 | December 10 | December 11 |
Bowling Green | Snowfall of 5" | Low of -10°, cold high of 17° | Low of -7°, cold high of 15° | Cold high of 19° |
Lexington | Low of -4°, snowfall of 9.4" | Low of -9°, cold high of 13° | Low of -9°, cold high of 6° | Low of -2°, cold high of 13° |
Louisville | Low of -1°, snowfall of 15.0" | Low of -6°, cold high of 10° | Low of -4°, cold high of 7° | Low of -3°, cold high of 11° |
Much of the length of the Ohio River froze over completely. By the 11th the ice was solid at Madison and gorged at Evansville, and by the 17th the river was solid at Cincinnati. At its mouth at Cairo, Illinois, the ice conditions exceeded in severity anything in the memory of the oldest inhabitant. The Kentucky and Licking rivers froze over as well.
A warm-up took place during the ten days leading up to the Christmas holiday, causing the snow pack to melt away. The warmth peaked on the 24th with highs in the 50s, and there was no snow on the ground Christmas Eve as a light rain fell.
It would be a month before we would see 50s again.
Another massive Canadian high pushed a cold front through southern Indiana and central Kentucky late Christmas Eve and brought another round of bitter cold to the region. Two to four inches of snow fell on southern Indiana and north central Kentucky on the 25th, helping to set the stage for the cold temperatures to come. By the morning of the 30th a second, reinforcing high came in and was centered just northeast of Lexington at dawn. In southern Indiana temperatures bottomed out at -14° at Madison, -20° at Paoli, and -19° at Salem and Scottsburg. In central Kentucky the coldest readings were -16° at Bardstown and Frankfort, -19° at Junction City, and -20° at Taylorsville.
A hundred years later, December 1917 still stands as the 3rd coldest December on record at Louisville and Lexington, #4 at Frankfort, and #6 at Bowling Green.
A few more inches of snow fell on the 1st and second of January 1918, and with the exception of a brief respite on the 5th and 6th, colder than normal conditions continued through the first 3 weeks of the new year. As a matter of fact, even after the blistering cold of early December, citizens of southern Indiana and central Kentucky hadn’t seen the worst cold of the season yet.
On the 11th yet another dome of incredibly cold air that had organized in northwest Canada pushed into the northern Plains and Rockies while strong low pressure developed rapidly near New Orleans. On the night of the 11th and into the 12th the low continued to deepen and became quite powerful as it shot through the Tennessee and Ohio Valleys to the Great Lakes while the Canadian high nosed as far south as the Texas coast.
Though the 11th started off near the freezing mark, the low traveled just to our east and joined forces with the western high to pull frigid air down from Canada. As the low deepened, those winds grew stronger…and stronger…and stronger. By evening sustained winds were in the 20 to 40 mph range across southern Indiana and central Kentucky, combined with blinding snow that had started falling around lunch time and intensified in the afternoon. As darkness fell the blizzard raged, with gusts to 52 mph at Louisville and 44 mph at Lexington. At 7pm in Louisville winds were sustained at 30 mph with a temperature of 13°.
The twelfth day of 1918 “was probably the coldest and most disagreeable day experienced in a century” (Monthly Weather Review), or at least since the intense cold of January 1, 1864 (Climatological Data). At 7am, winds were blowing at 20 to 30 mph while the temperature at Louisville was fifteen degrees below zero and Lexington was fourteen below. Using the modern calculation, that gives a wind chill around 45 degrees below zero!
Though the snow had ended and the sun came out, the wind continued to roar throughout the day at speeds of 20 to 40 mph with higher gusts. The temperature rose to only -2° at Louisville and -3° at Lexington. January 12, 1918 is still one of only two days in both cities’ histories during which the temperature stayed below zero all day.
After the storm of the 11th-12th abated, intense cold continued. Every day from the 12th to the 22nd the daily high temperature at Louisville was colder than the normal low temperature (using 1981-2010 normals). Heavy snows fell during this time as well, including 11.2” at Louisville on the 14th-15th
accompanied by winds gusting to 45 mph.
Transportation was practically paralyzed, including the inter-urbans and streetcars. In Louisville on the 14th three people were killed when two inter-urban cars collided, and the next day four people died when a snow-laden roof collapsed. Fuel shortages added to the misery, and schools were closed for several days. In northern and western sections of Kentucky some rural farmers were snowbound for up to two weeks.
Lowest temperatures January 10-23:
Alpha, KY -12°(12th) | Edmonton, KY -15° (12th, 21st) | Junction City, KY -14° (12th, 21st) | Louisville Downtown, -15° (12th) | St. John, KY -16° (12th) | ||||
Anchorage, KY -16° (12th) | Frankfort, KY -12° (12th, 21st) | Leitchfield, KY -16° (12th) | Madison, IN -17° (21st) | Salem, IN -23° (20th, 21st) | ||||
Bardstown, KY -14° (12th) | Greensburg, KY -16° (21st) | Lexington, KY -14° (12th) | Paoli, IN -25° (20th) | Scottsburg, IN -20° (21st) | ||||
Beaver Dam, KY -16° (21st) | Irvington, KY -15° (12th) | Loretto, KY -16° (12th) | Richmond, KY -13° (12th) | Shelbyville, KY -16° (12th, 13th, 21st) | ||||
Berea, KY -14° (21st) | Jeffersonville, IN -12° (12th) | Louisville Cherokee Park, -16° (12th) | Russellville, KY -15° (12th) | Taylorsville, KY -18° (21st) | ||||
Bowling Green, KY -15° (12th) |
The amount of ice on the Ohio and middle Mississippi rivers surpassed anything that had been seen before. At Cairo pedestrians were able to walk across the Ohio River to Kentucky, a feat unknown to even the oldest inhabitants of the region.
Though there was one more cold snap in early February when some locations dipped below zero on the 4th-5th, the remainder of the winter was relatively mild. On February 28 record highs were set that still stand today, including 75° at Lexington and 77° at Louisville.
(click on an image to see a larger version)
December 1917
Temperature Departure from Normal, Local | Temperature Departure from Normal, Midwest | Temperature Departure from Normal, USA |
January 1918
Temperature Departure from Normal, Local | Temperature Departure from Normal, Midwest | Temperature Departure from Normal, USA |
December-January
Temperature Departure from Normal, Local | Temperature Departure from Normal, Midwest | Temperature Departure from Normal, USA |
Temperature Plot for Louisville, December 1917-January 1918 |
Temperature Plot for Lexington, December 1917-January 1918 |
Temperature Plot for Madison, December 1917-January 1918 |
Temperature Plot for Greensburg, December 1917-1918 |
All maps on this page were generated by the Midwest Regional Climate Center. The graphics are smoothed, so values shown here may not exactly match individual values at a specific point.
(click on an image to see a larger version)
December 1917
Snowfall Departure from Normal, Local | Snowfall Departure from Normal, Midwest | Snowfall Departure from Normal, USA |
Snowfall Percent of Normal, Local | Snowfall Percent of Normal, Midwest | Snowfall Percent of Normal, USA |
January 1918
Snowfall Departure from Normal, Local | Snowfall Departure from Normal, Midwest | Snowfall Departure from Normal, USA |
Snowfall Percent of Normal, Local | Snowfall Percent of Normal, Midwest | Snowfall Percent of Normal, USA |
December-January
Snowfall Departure from Normal, Local | Snowfall Departure from Normal, Midwest | Snowfall Departure from Normal, USA |
Snowfall Departure from Normal, Local | Snowfall Departure from Normal, Midwest | Snowfall Departure from Normal, USA |
All of the graphics above were generated by the Midwest Regional Climate Center. The graphics are smoothed, so values shown here may not exactly match individual values at a specific point.
December 8-9, 1917 Snowfall
(click image to see a larger version)
December 1917
December 7, 1917 Low pressure begins to develop over west Texas while high pressure noses in from Canada. |
December 8, 1917 The Texas low has deepened and moved into the Tennessee Valley while the Canadian high invades the Great Plains. |
December 9, 1917 The low moves off to the St Lawrence Valley as cold air dumps into the eastern U.S. |
December 10, 1917 Frigid high pressure takes over from the Northern Plains to Dixie |
December 11, 1917 The center of the high moves right through the Ohio Valley atop deep, fresh snow cover. |
December 27, 1917 Shortly after Christmas another Canadian high appears and pushes a weak cold front from the Plains into the Midwest. |
December 28, 1917 The cold front continues to the east as high pressure advances. |
December 29, 1917 Circulation around the high across Great Lakes produces widespread snows. There may also be some lingering moisture left over from the departing cold front. |
December 30, 1917 High pressure centers itself over the Ohio Valley around dawn, atop a few inches of new snow. |
December 31, 1917 |
From Monthly Weather Review, December 1917:
High Pressure Tracks | Isobars, Isotherms, Winds | Low Pressure Tracks | Mean Temperature Departure | |
Percent Clear Daytime Sky | Total Precipitation | Total Snowfall |
January 1918
January 11, 1918 Low pressure spins up quickly over the Louisiana coast while high pressure moves in from Canada. |
January 12, 1918 The low deepens dramatically while the high continues to push into the U.S. |
January 13, 1918 The low moves off to the northeast as high pressure penetrates all the way to the Gulf of Mexico. |
January 14, 1918 Another low spins up along the western Gulf Coast. |
January 15, 1918 The low moves quickly to the mid-Atlantic after dropping heavy snow along its path. |
January 16, 1918 Cold high pressure follows the snowy low. |
From Monthly Weather Review, January 1918
High Pressure Tracks | Isobars, Isotherms, Winds | Low Pressure Tracks | Mean Temperature Departure | |
Percent Clear Daytime Sky | Total Precipitation | Total Snowfall |
(click image to see a larger version)
December 1917
Anchorage "Heaviest snow on record" |
Louisville Cherokee Park "Blizzard all afternoon" |
Ohio River at Cincinnati "Frozen" |
||
Madison "Blizzard" |
Paoli "The worst December on record" |
Frankfort "River frozen" |
January 1918
Louisville Cherokee Park "Blizzard" |
Frankfort "River frozen" |
Shelbyville "-16" |
(click image to see a larger version)
December 1917
December 7, 1917 Low pressure begins to develop over west Texas while high pressure noses in from Canada. |
December 8, 1917 The Texas low has deepened and moved into the Tennessee Valley while the Canadian high invades the Great Plains. |
December 9, 1917 The low moves off to the St Lawrence Valley as cold air dumps into the eastern U.S. |
December 10, 1917 Frigid high pressure takes over from the Northern Plains to Dixie |
December 11, 1917 The center of the high moves right through the Ohio Valley atop deep, fresh snow cover. |
December 27, 1917 Shortly after Christmas another Canadian high appears and pushes a weak cold front from the Plains into the Midwest. |
December 28, 1917 The cold front continues to the east as high pressure advances. |
December 29, 1917 Circulation around the high across Great Lakes produces widespread snows. There may also be some lingering moisture left over from the departing cold front. |
December 30, 1917 High pressure centers itself over the Ohio Valley around dawn, atop a few inches of new snow. |
December 31, 1917 |
From Monthly Weather Review, December 1917:
High Pressure Tracks | Isobars, Isotherms, Winds | Low Pressure Tracks | Mean Temperature Departure | |
Percent Clear Daytime Sky | Total Precipitation | Total Snowfall |
January 1918
January 11, 1918 Low pressure spins up quickly over the Louisiana coast while high pressure moves in from Canada. |
January 12, 1918 The low deepens dramatically while the high continues to push into the U.S. |
January 13, 1918 The low moves off to the northeast as high pressure penetrates all the way to the Gulf of Mexico. |
January 14, 1918 Another low spins up along the western Gulf Coast. |
January 15, 1918 The low moves quickly to the mid-Atlantic after dropping heavy snow along its path. |
January 16, 1918 Cold high pressure follows the snowy low. |
From Monthly Weather Review, January 1918
High Pressure Tracks | Isobars, Isotherms, Winds | Low Pressure Tracks | Mean Temperature Departure | |
Percent Clear Daytime Sky | Total Precipitation | Total Snowfall |
Meteorologists and climatologists often look to various oscillations in the global atmosphere to help explain certain weather patterns. One such oscillation that many of us are familiar with is the El Niño Southern Oscillation, or ENSO, which actually includes not only El Niño but also La Niña and neutral phases.
Another oscillation we can look at is the North Atlantic Oscillation, or NAO. The NAO compares air pressure over Greenland and Iceland to air pressure near the Azores. Changes in the strength and positioning of these two pressure centers affects the flow of the jet stream, which results in changes in the distribution of temperature and precipitation.
When the pressure centers weaken, the jet stream weakens and buckles, allowing cold air that might otherwise have stayed bottled up in Canada to come flooding south in to the eastern United States. This is called a "negative phase" NAO. Below is an illustration of typical weather conditions experienced when the NAO is negative:
In December 1917 and January 1918 the NAO was negative. Really negative. One study suggested that the NAO of January 1918 was in the top 2% of the strongest negative NAO's of the last few hundred years. The succession of bitterly cold Arctic outbreaks sweeping down from the northern Rockies into the central and eastern United States certainly seems to fit the typical weather pattern of a negative NAO. The index turned positive in February and winter gradually loosened its icy grip on the Ohio Valley.
Interestingly, the winter of 1917-18 was a La Niña season (source). La Niña typically brings warmer conditions to the southeast U.S. and Ohio Valley. It's possible that the NAO was so strong that it dominated that winter's weather, overriding the signal from La Niña.
The negative NAO and La Niña tend to fight against each other when it comes to precipitation, too. La Niña tends to make the Ohio Valley wetter than normal, whereas a negative NAO favors dry conditions. In December 1917-January 1918 our region's precipitation totals were around or just slightly below normal. However, because it was so very cold, much of the precipitation that did occur came down in the form of snow, resulting in snowfall totals that were well above normal.