Skip to content

History Of Megadrought In Kansas

September 20, 2015

By Paul Homewood 




The Kansas Geological Survey have this very interesting study about the climatic history of Kansas over the last millennium, in particular drought trends.


Environmentally and economically, drought is one of the most costly natural disasters in North America. Yet it rarely gets the same public attention that other, more spectacular, natural disasters receive. While tornadoes, floods, wildfires, and hurricanes leave behind well-defined swaths of devastation in relatively short order, droughts whittle away at water quality and quantity, topsoil, crop yields, and other natural and socioeconomic resources over months and years, even decades.

In any given year, drought conditions of some degree are occurring somewhere in North America. For 1988–midway through a three-year drought in the central and eastern United States–the estimate of national drought damage was a record $40 billion (National Climatic Data Center, 2012), or $78.5 billion in 2013 dollars. In 2011, losses in Kansas alone exceeded $1.7 billion (Kansas Department of Agriculture, 2011). The 1930s and 1950s droughts, however, remain the benchmarks for the 20th century in terms of duration, severity, and spatial extent. The historically unprecedented dust storms in the 1930s, exacerbated by the rapid spread of farming practices unsuitable for the semi-arid High Plains, helped make the Dust Bowl the most memorable drought in modern times (fig. 1). Two decades later, 1956 surpassed 1934 as the single most severe drought year ever documented statewide based on instrumental precipitation and temperature readings.


Figure 1–Drifts of wind-blown soil on a farm near Liberal, Kansas, March 1936 (photo by Arthur Rothstein: Library of Congress).

Black and white photo of farmstead overcome by sand dunes.


Yet the major 20th-century droughts, as impressive as they were, do not rank as the two most intense or enduring droughts to hit Kansas.

For decades, scientists have been collecting and analyzing data to reconstruct paleoclimates–past climates dating back thousands of years–from clues in tree rings, sediments, and other proxies. Several past drought episodes, it turns out, have exceeded those of the 1930s and 1950s in severity, extent, and duration (Layzell, 2012). If such a drought occurred today, reductions in surface-water and groundwater resources would threaten municipal, industrial, and agricultural water supplies and cause widespread crop failure.

Being able to better forecast and plan for severe drought conditions is vital. Temperature and precipitation records are indispensable for understanding climate change but are largely restricted to the past 100 years. To assess the full range of drought variability that has occurred over 1,000 or more years, climatologists and other scientists measure the severity of pre-20th-century droughts using reconstructed paleoclimatic data from proxies and analyze the results in combination with more than a century of instrumental data.


Drought Measurement and The Palmer Drought Severity Index (PDSI)

Several indices have been developed to measure drought. The Palmer Drought Severity Index (PDSI), one of the most widely used indices in North America, was introduced in 1965 and is used to measure the severity of a drought occurrence for a specified period. PDSI values can be calculated from weather data collected using thermometers, rain gauges, and other instruments–available for most of North America since about 1895–or from paleoclimatic data reconstructed from tree rings and other proxy evidence.

PDSI values, representing relative wetness and dryness, are assigned after recorded precipitation and temperature data or proxy data are analyzed to determine how much soil moisture was available at a specific time compared to how much would be available under average conditions. The values typically range from -4 (extremely dry) to 4 (extremely wet), although the range is unlimited. Although a PDSI value of -4 or less (even more extreme) is daunting, a persistent drought averaging moderate (-2) to severe (-3) PDSI values over many years may actually cause more damage than a more severe but shorter episode. Plotted PDSI values provide a picture of climate variability over time and can be used to calculate the duration of drought conditions. The extent of the 1930s and the 1950s droughts–and the relatively wet years in between–are evident in the PDSI trends for Kansas from 1895 to 2011 (fig. 2).

Figure 2–Palmer Drought Severity Index (PDSI) trends in Kansas, 1895-2011 (High Plains Aquifer Atlas, 2012).

Chart showing changed in PDSI from 1895 to 2011.

Drought Severity and Duration

A key characteristic distinguishing the 1930s and 1950s droughts from other modern drought periods is aridity that was not only severe but also long lasting. The negative effects of one extremely dry year can be overcome relatively quickly when it is preceded or followed by a wetter year, but several years of nearly uninterrupted drought can lead to serious long-lasting socioeconomic and environmental problems. The PDSI value for 2002 in southwestern Kansas was -7.1, compared to -5.0 for the peak year of the Dust Bowl, yet the situation was not as dire in 2002 because it was bounded by years with positive PDSI values.

Year to year, climatic conditions vary across the state, with droughts hitting some regions harder than others. Since 1000 AD, southwestern Kansas has experienced a greater number of extreme droughts than southeastern Kansas (fig. 4). This west to east trend mirrors the strong longitudinal climate gradient in Kansas (High Plains Aquifer Atlas, 2012). Average precipitation today gradually increases from about 15 inches along the Colorado border to 45 inches in the southeastern corner of the state (High Plains Aquifer Atlas, 2012).


Figure 4–Annual PDSI reconstructions showing drought severity in southwestern Kansas and southeastern Kansas. Dashed lines indicate 1934 (black) and 1956 (red) PDSI values (Layzell, 2012). PDSI values are from Cook and Krusic, 2004. Reconstructions for all six Kansas regions are online at


PDSI reconstruction for southeast Kansas from 1000 to present.


PDSI reconstruction for southwest Kansas from 1000 to present.






Droughts of unusually long duration are commonly referred to as "megadroughts," although there is no standard definition of the term (Stahle et al., 2007). This public information circular considers a megadrought to be any multi-year drought that significantly exceeds the duration of the most extreme droughts of the 20th century. Lasting 20 or more years, these extreme episodes do contain individual years of normal or even above-average precipitation.

Megadroughts appear to be most prevalent in Kansas between 850 AD and 1500 AD (fig. 7). The longest one occurred in north-central Kansas from 1317 to 1427. As north-central Kansas was enduring near-continuous drought for 110 years, northwestern Kansas experienced two long-term droughts separated by a wetter period and southwestern Kansas conditions did not reach megadrought proportions (see fig. 6). These differences underscore how much circumstances can vary over a short distance.


Figure 6–Smoothed PDSI reconstructions showing drought durations for southwestern Kansas (top) and southeastern Kansas (bottom). Light gray bars indicate episodes of similar duration to the 1930s and 1950s droughts and dark gray bars represent episodes of greater duration. Annual PDSI values have been smoothed to filter out anomalous high and low values over a 10-year range (blue) and a 50-year range (red) (Layzell, 2012). PDSI values are from Cook and Krusic, 2004. Reconstructions for all six Kansas regions are online at

PDSI reconstruction for southeast Kansas from 1000 to present with smoothing functions and gray colors to indicate megadroughts.

PDSI reconstruction for southwest Kansas from 1000 to present with smoothing functions and gray colors to indicate megadroughts.

Many of the known megadroughts in North America occurred during the Medieval Warm Period (MWP), a time of significant climatic variability that lasted from about 900 to 1300. First identified in northern Europe, the MWP was later documented in other areas of the world, including parts of the western United States (Cook et al., 2004). A shift around 1500 to droughts of shorter duration may coincide with the onset of cooler climatic conditions during a period known as the Little Ice Age. Many dune records from the central Great Plains show significant sand dune activation–a sign of increased aridity and reduced vegetation–during these periods. A variety of sand-dune mobilizations have been documented from the 9th to the early 20th century in Kansas (fig. 7).


Figure 7–Synthesis of reconstructed PDSI data for the six regions of Kansas showing the severity and duration of droughts since about 850 AD. Events identified using geomorphic (sand dune), archaeological, and historical proxies are marked (Layzell, 2012). Sand-dune reactivation information from 1. Forman et al., 2008; 2. Lepper and Scott, 2005; 3. Hanson et al., 2010; 4. Arbogast, 1996; and 5. Halfen et al., 2012.


PDSI reconstruction for six regions of Kansas with several historical events marked as signposts.




Archaeological and Historical Corroboration of Megadroughts

Evidence that megadroughts destabilized North American civilizations between 850 and 1500 AD is found in the archaeological record. Although drought probably affected populations in the Great Plains during that time, clues there are sparse. Archaeological evidence of agricultural societies in adjacent regions, however, provides signs of widespread drought conditions that most likely also afflicted the plains people.

Several major droughts may have undermined Native American cultures between the 11th and 15th centuries. The population of the Fremont cultures in the Four Corners region of the U.S. Southwest declined around 1000 AD in the midst of a multi-decade drought (fig. 7). The 13th-century drought commonly referred to as the "Great Drought" contributed to the abandonment of Anasazi agricultural settlements in the same region and also appears to have impaired Mississippian agricultural societies hundreds of miles to the northeast (Benson et al., 2007). Further megadroughts in the 14th and 15th centuries likely contributed to the abandonment of Cahokia near present-day St. Louis by 1450 (Cook et al., 2007).

Widespread drought during the Stephen Long expedition in 1819-1820 probably influenced the explorers’ perception of the western Great Plains as the "Great American Desert." "The chief produce of these tracts of unmixed sand, is the sunflower, often the dense and almost exclusive occupant," wrote expedition member Edwin James. Jacob Fowler noted that on his way to Santa Fe in 1821, the sand hills along the Arkansas River in south-central Kansas were "distetute of vigetation as they are Bald" (Muhs and Holliday, 1995).

Set-tan (Little Bear) of the Kiowa recorded in his 60-year calendar history that during the hot "sitting summer" of 1855, the prairie grasses dried out and the Kiowa had to stop frequently to rest their emaciated horses (Stahle et al., 2007). Accounts from early settlers in eastern Kansas Territory also expounded on drought conditions that lasted from at least 1854 into the early 1860s, with only short reprieves. Newspapers reported suffocating dust storms, crop failures, prairie fires, "scorching, withering, blighting" winds, and the outward migrations of many newly arrived settlers (Malin, 1946). By the mid-1890s, locals around Garden City observed that area sand hills were becoming less extensive. Following brief reactivation of sand in small areas during the Dust Bowl years, dunes on the south side of the Arkansas River from just east of Pueblo, Colorado, to near Wichita are now mostly stabilized by vegetation (Muhs and Holliday, 1995).




It is a rather chilling reminder of how we are all at the mercy of Mother Nature.

  1. Elaine Supkis permalink
    September 20, 2015 12:11 pm

    The Southwest desert echos the Kansas droughts. The Long Drought you point out for Kansas was echoed in Arizona. Most of the farming Indians fled or died during that 100 year drought with virtually no rain.

    In the 1950’s we had a drought, only 5 years long. My brother, David was born during this drought and when it suddenly rained while we were in church one Sunday, we all ran outside to dance and yell in the rain.

    My poor brother thought it was the Apocalypse and he cried and tried to hide from the weird stuff falling from the sky!

  2. September 20, 2015 1:28 pm

    Interesting. For me, a big part of the story of this region is not only “how much rain is it getting?” But “what are we trying to do here?” One of the things that jumped out at me was that the Dust Bowl corresponded with a dry period — but nothing really outside the range of the normal variation. The issue, in other words, wasn’t that it suddenly got drier than it had ever been before. The issue was that farmers were using the land in ways that were inappropriate for a region with that type climate and that range of normal variation. And they’re doing it again. Doesn’t matter if we have a megadrought or a regular dry cycle. When the aquifer is empty, it’s all over.

  3. John F. Hultquist permalink
    September 20, 2015 3:41 pm

    The comment by danallosso ends with “… it’s all over.”

    Many times I have seen such statements and, always, I am left to wonder what the writer has in mind. One thing we know is that things change. About 60 years ago there were USA state puzzles that seem odd today, see here:

    Look at Ohio, with a rubber tire. Or note Washington State with Fir trees – The Evergreen State. As major industries in these states, those images are no longer relevant. Wine grapes might better represent WA. Also, today North Dakota would have oil derricks, and the fellow with a pick would be gone from Pennsylvania – being replaced by finance and biomedical workers.

    60 years from now things will be different in many places but to claim “it’s all over” is to dismiss the concepts of dynamic forces and disruptive innovation.

    • sandyS permalink
      September 20, 2015 7:43 pm

      That’s a good observation, to me me oil seed rape is a new feature of arable farming in the UK. To my children it’s the way it’s always been.

  4. September 21, 2015 2:05 pm

    I was familiar with the drought-caused abandonment of Mesa Verde in southern CO due to prolonged drought when we visited the cliff-dwellings in 1965. More recently, while State American Indians chairman for Daughters of the American Revolution, I picked up a book on “The Moundbuilders” as we have Indian mounds in WV and also the famous “Serpent Mound” in nearby southern OH. That led to a book “Cahokia” about the city in IL across the Mississippi River from St. Louis, MO which is mentioned in the article. Archaeology, including the ethnobotanical studies of plant materials being used by the Indian cultures, certainly gives a picture of climate affecting cultures. One of the more amusing anecdotes from the book is the opening of a garbage dump in a “borrow pit” created when soil was excavated to build the mounds. The pit was 195′ long x 62′ wide x 10′ deep originally. One area, more deeply buried in a wet portion and oxygen deprived for the past 900 years was remarkably preserved and extremely rank when opened.

Comments are closed.

%d bloggers like this: