Tornado Survey – “The Old School Way”

Take yourself back 75 years to August 10, 1939 when World War II was less than a month from getting underway, the U.S. was slowly climbing out of the Great Depression and on the brink of war, the Studebaker Champion was introduced and cost about $660 (or $11,312 in 2014), and the Cubs actually had a winning record! Sorry Cubs’ fans. The country was just learning about ALS or Lou Gehrig’s disease as he was just diagnosed and had to retire from baseball that summer. The price of gas was $0.10, a pound of hamburger was $0.14, and a loaf of bread was $0.08. The average cost of a new house in 1939 was $3800 but now, the average cost of a new home in 2014 is $339,100 (per U.S. Census).

Okay, so you’re back in August 1939 in Iowa when no weather radar coverage or tornado warnings were available to meteorologists. Folks literally could say “It struck without warning” and be honest about it! They still did damage surveys, but the Fujita scale wouldn’t be introduced until 1971.  So when C.D. Reed and/or S.E. Decker, from the Iowa Department of Agriculture or IDA (See Figures 1a and 1b) had the daunting task of surveying three destructive tornadoes that occurred on August 10, 1939 in central Iowa, they did an amazing job (See Figure 2).  Back then, surveying included talking with eyewitnesses hit by the tornado and whatever sort of geodetic survey equipment they had available. They had limited resources, but the detail of what buildings were hit, livestock killed, or persons injured was phenomenal. Granted there were less people and fewer buildings to destroy, but traveling and communication was more cumbersome in 1939 than 2014; especially since the survey covered several counties.

Figure 1a: C.D. Reed Author of the Iowa monthly climate review for August 1939.

Figure 1a: C.D. Reed Author of the Iowa monthly climate review for August 1939.

Figure 1b: S.E. Decker was the author of the damage survey or storm section in the Climatological Data: Iowa Section.

Figure 1b: S.E. Decker was the author of the damage survey or storm section in the Climatological Data: Iowa Section.

The hardest hit counties were Adair, Clark, and Warren from the tornadoes while Polk County endured significant damage due to heavy rainfall. Well what do you know  – heavy rain in August in Iowa? There’s a shocker.  Another county, Montgomery, was hit hard with large hail as noted on the tornado track from Figure 2 and suffered $10,000 worth of crop damage.

August 10-1939 TornadoPaths-resize

Figure 2: hand drawn maps of the tornado paths and hail swaths on August 10, 1939.

The first and second tornadoes occurred in Shelby and Adair Counties respectively. The twister in Shelby County damaged buildings on five farms resulting in a loss of $12,000 (see inflation rate table below) and injured one person, Mrs. Pete Anderson, according to the report.

The Adair County tornado started about 3:30 p.m. near the Summerset Township and traveled northeast through Summerset and Prussia to just east of Fontanelle, Iowa. Miss Mildred Bakerink was fortunate to photo the tornado when it was about 3 miles northeast of her location near Prussia (See Figure 3). Reports suggested the early life of the tornado that “the storm of pendent cloud was shaped more like a cone with a wide V-top.”  The survey determined the tornado path was about “12 miles long and 80 rods wide.” A rod is equivalent to 5 ½ yards or 16 ½ feet in length.  Hence, the tornado width was roughly 440 yards (1320 feet) wide or a quarter of a mile. That’s a pretty significant tornado. To compare it to a recent tornado that occurred in Iowa, the “Belmond” Tornado that passed through the north side of Belmond on June 12, 2013 was 200 yards wide with path length of 6.2 miles. This tornado was rated an EF-3 tornado with a 155 mph peak wind speed. You can draw your own conclusions on where to rate the Adair County Tornado from August 10, 1939. To help you out, damage was estimated to buildings on six farms ranged from $5,000 to $10,000 while the damage to crops, stock, implements “amounted to several thousand dollars”, according to the IDA report. Luckily, there was only one injury and no deaths.


The third and most destructive tornado was on the ground for roughly 35 miles and it originated southwest of Osceola, in Clarke County, and finally dissipated near Milo in Warren County.  The damage surveyor, likely C.D. Reed, visited Liberty Center where several eyewitnesses said they could see five funnel clouds visible at one time southwest of town. There were also several reports from Osceola that suggested seeing the five funnel clouds at the same time.  In fact, a writer from the Osceola Tribune depicted the funnel cloud as “bounding around like a rubber ball, alternately lifting and lowering.” Here’s how the eyewitnesses from Liberty Center described the funnels:

“…as being close together and joined to a common dark cloud mass. They were said to be suspended in the air without touching the ground as long as they remained separated, but that upon joining or merging the remaining funnel grew in length and extended down to the ground.”

This sounds a lot like a what modern day meteorologists call a multi-vortex tornado. It certainly did some damage to Clarke and Warren Counties. The IDA report said “buildings were demolished on at least ten farms” in Warren County. There were 18 of 22 buildings, on one livestock farmstead, “wrecked or seriously damaged.” There was a stretch of corn, roughly a mile wide and 15 miles long, which was completely flattened. Several trees were snapped or uprooted, power and telephone lines blown down, and most fences blown away in the tornado path. The description of the damage near Liberty Center gets even more detailed (See Figure 4). The total damage from the storm, including heavy rain and straight-line winds, in Warren County was estimated to be at least $102,000 as the IDA report stated “several thousand dollars more” to furniture, telephone lines, crops, etc.  There were several injuries but no related deaths. The table below shows the damage adjusted for inflation from 1939 to 2014.

August 10-1939 DamageDescription

Figure 4: very detailed description of the the tornado damage near Liberty, Iowa.

The surveyor calculated the speed of the storm itself at around 40 mph by determining when and where it originated and when and where it dissipated. Something we still do today but with the aid of radar data, satellite imagery, and aerial photos.  The surveyor estimated the “rotary winds indicated at least hurricane velocity of about 75 miles per hour.” From the description of the damage, there’s little doubt the peak wind speeds where likely higher.

As far as the meteorological setup, a surface analysis on the morning of the storms (See Figures 5-7) suggested a warm front draped across Arkansas into eastern Oklahoma. However, further surface analysis along with the description from the IDA report, the warm front extended north-northwest and connected to the cold front near the Grand Island area.  The storms developed along the warm front by the late afternoon as its surged north throughout the day. The cold front, according to the report, pushed through the Des Moines area around 6:30 p.m. time frame.

A fine job done by the folks at the Iowa Department of Agriculture, which was likely down by either C.D. Reed or S.E. Decker or both, on the storm survey from August 10, 1939.


Figure 5: North America Synoptic Weather Map on the morning of August 10, 1939 at 1230 UTC or 6:30 am.


Figure 6: Zoomed in morning surface analysis on the Corn Belt and Central Plains.


Figure 7: CONUS surface analysis on August 10, 1939 at 6:30 am. The map shows the area of low pressure over northern Kansas and a boundary extending from southwest to northeast Iowa. It also depicts thunderstorms over South Dakota, western Nebraska, and far northwest Iowa.

Tornado Damage Adjusted for Inflation

Township County Type of Damage 1939 Cost 2014 Cost
Polk Shelby Buildings $12,000 $205,700
Polk Shelby Crops $1,000 $17,140
Polk Shelby Livestock $100 $1,710
Red Oak Montgomery Crops $10,000 $171,400
Summerset/Prussia Adair Buildings $5,000-10,000 $85,700-171,400
Near Osceola Clarke Buildings $10,000-15,000 $171,400-257,100
Near Osceola Clarke Crops $5,000 $85,700
Near Osceola Clarke Livestock $500 $8570
Near Liberty Center Warren Buildings $75,000 $1.3 million
Near Liberty Center Warren Livestock $2,000 $34,280
Near Liberty Center Warren Crops $25,000 $428,500

Inflation rates rounded and based off cumulative rate of 1614.0%


Blog post by Kenny Podrazik – NWS Des Moines

Snow in September?

Believe it or not, there have been at least nine years in which snow has been recorded in Iowa in the month of September, most recently in 1995, as detailed in Table 1 at the end of this document. The most remarkable of these events is the very early snowfall of September 16, 1881, which was amazing not only for its earliness in the season but also for the extent and amount of snowfall. The track of the surface low pressure center associated with this storm system is illustrated in Figure 1, a reprint of the original “War Department Weather Map” from September of 1881. At the time weather observations and reports were filed by the U.S. Army Signal Service, the progenitor of the modern National Weather Service. In the report of the Chief Signal Officer for that month, the development of the low pressure center is detailed as follows:

“This storm, which pursued a very anomalous track, was first evident in Texas, where on the 14th it moved in a track nearly due east. At the midnight observation, while the storm center was near New Orleans, a barometric depression extended from the Gulf of Mexico to the Lake Superior region. At the same time [an area of high pressure] prevailed with fair weather in New England. These conditions were unfavorable to an eastern progress of the storm, and on the 15th the depression moved in a northerly course to Lake Michigan. On the 16th, with diminishing energy, the storm center moved into Iowa and Minnesota, and on the 17th into Manitoba. The track on the 16th and 17th is very remarkable, and probably for a storm of such energy will have no parallel in the history of the Signal Service.”

Figure 1: Highlighted track of the low pressure center from September 14-17, 1881.

Figure 1: Highlighted track of the low pressure center from September 14-17, 1881.

On September 14, as the low moved across the Gulf of Mexico, fair weather prevailed across most of Iowa until a cold front moved through late in the day. At Des Moines the high temperature was 80 degrees but then the official observer wrote that, “Low stratus clouds moved rapidly from north and northwest during the afternoon and evening.” On the 15th, as the low pressure center moved northward toward Chicago, it pulled down unseasonably cool air behind the front into Iowa and spread a cold rain across much of the upper Midwest. At Des Moines the temperature fell through the day, with a high of 58 degrees measured early in the morning. The observer noted, “Cloudy and threatening weather prevailed during the day, low stratus clouds moved from the north.”

By the morning of September 16, the low pressure center was moving slowly westward into southern Minnesota and northern Iowa, pushing cold air even further southward into the central U.S. Frost was noted as far south as Arkansas and Texas and at Fort Gibson, Oklahoma ice formed on standing water. Across eastern Nebraska, southern Minnesota, and about the northwestern two thirds of Iowa the colder air allowed rain to mix with or change entirely over to snow at times, mostly in the morning. At Des Moines the high temperature for the day was only 46 degrees and the observer recorded that, “Few flakes of snow was observed.” Further north and west the snow was heavier, in some areas melting as it fell but in others managing to accumulate for a short time. At Algona an estimated 4 inches of snow fell in the morning, breaking some tree branches, but all melted by noon. The snow was observed as “quite heavy” at Creston, while “several inches” were noted between Des Moines and Atlantic and 4-6 inches were estimated on the Rock Island Railroad between Stuart and Avoca.

This stands as one of only two occasions on which fairly widespread, measurable snow has fallen in Iowa in the month of September, the other being on September 25, 1942. In that storm most of the state received snow with amounts ranging up to 4 inches at Allison, Forest City, Mason City, and Millerton and scattered tree and utility line damage noted across the state.

Table 1: Years in which snow has been recorded in Iowa in September.
1881 – widespread measurable snowfall on the 16th (see above)
1895 – “first snowflakes” noted at Madrid on the 28th
1912 – “few flakes” observed at Storm Lake and Marshalltown on the 17th and 18th
1938 – flurries reported at Orleans and Maquoketa on the 18th and 19th
1939 – traces of light snow and sleet across northern IA, 0.1” at Sheldon and 0.2” at Sibley, on the 29th and 30th
1942 – widespread measurable snowfall on the 25th (see above)
1961 – light snow across northwestern half of IA on the 30th, a few measurable amounts ranging up to 3.0” at Swea City
1985 – a few flakes at Des Moines on the 24th, widespread wintry mix with 0.5” at Audubon and Storm Lake on the 29th and 30th
1995 – a few flakes and ice pellets mixed with rain across northern IA on the 22nd

Blog post by Jim Lee