Drought in the Western Great Plains, 1845-56: Impacts and Implications

Grazing Bison.  Yellowstone National Park photo. Drought in the Western Great Plains, 1845-56: Impacts and Implications
Bulletin of the American Meteorological Society,
83, 1485-1493, October 2002.

Connie A. Woodhouse 1, 2, Jeffrey J. Lukas2, and Peter M. Brown3.

1 NOAA Paleoclimatology Program
2 Institute of Arctic and Alpine Research, University of Colorado
3 Rocky Mountain Tree Ring Research

A sustained mid-19th century drought in the western Great Plains has been indicated by a tree-ring analysis of trees flanking the western Great Plains, and in tree-ring reconstructions of drought and streamflow for eastern Colorado and the Colorado Front Range. The development of new tree-ring chronologies for the western Great Plains, in combination with existing chronologies, now enables a more detailed assessment of the spatial and temporal characteristics of this drought. The analysis of a set of drought-sensitive tree-ring chronologies ranging from the northwestern Great Plains to central New Mexico indicates a core area of drought from south-central Wyoming to northeastern New Mexico for the years 1845-56. Drought was particularly severe in the years 1845-48, 1851, and 1854-56, contracting and affecting smaller regions in intervening years. The impact of this drought on natural ecosystems and human activities is difficult to gauge because of the paucity of historical documents and the confounding effects of land use changes occurring over the same period. However, it is probable that this drought played a role in the decimation of bison herds in the second half of the 19th century. Were it to occur today, this relatively small but persistent drought would have significant impacts on the Colorado Front Range metropolitan area and the agricultural regions of eastern Colorado.
Download the eastern Colorado PDSI reconstruction data from the WDC Paleo Archive.

Many of the tree ring chronologies utilized in this study are available from the International Tree Ring Data Bank.

Figure 1 Drought and Streamflow reconstructions
Figure 1. a) Reconstruction of Middle Boulder Creek, mean annual flow in cubic meters/second (cms), 1710-1987 (Woodhouse 2001). b) Reconstruction of spring/summer droughts (May-June-July PDSI) for eastern Colorado, 1552-1995 (Woodhouse and Brown 2001). Series have been smoothed with a 5-weight binomial filter. The 19th century drought in both series is indicated by the striped bar. The 1930s and 1950s droughts are shaded for comparison of duration and severity to the 19th century drought.
Figure 2 Map of Great Plains drought extent and tree ring sites Figure 2. Locations of tree-ring chronologies. Outline indicates core drought region for 1845-1856. Chronology sites are numbered and correspond to those listed in Fig. 3. Chronologies selected were from species known to be sensitive to drought (ponderosa pine, Pinus ponderosa; Douglas-fir, Pseudotsuga menziesii; pinyon pine, Pinus edulis; and post oak, Quercus stellata), and were taken to be proxies of drought (generally winter/spring in the south grading to spring/early summer in the north). All but the three Montana chronologies (courtesy of David Meko) were obtained or are now available from the World Data Center for Paleoclimatology’s International Tree-Ring Data Bank (ITRDB). In all but three cases (the Montana sites, where raw data were not available), raw ring width measurements were used to generate tree-ring chronologies (ARSTAN; Cook 1985) to insure that the same standardization process and conservative detrending methods were used for all chronologies. Also included were 11 newly generated chronologies from isolated ponderosa pine, Douglas-fir, and pinyon pine woodlands growing in the Great Plains in Nebraska, eastern and central Colorado, and northeastern New Mexico (Woodhouse and Brown 2001). Except for the three Montana chronologies, residual chronologies, from which low-order autocorrelation presumed to be biological in origin has been removed (Fritts 1976), were used for this study.
Figure 1 Drought and Streamflow reconstructions Figure 3. Spatial distribution of low-ranking (narrow ring widths) years for 60 tree-ring chronologies, 1840-1870. The chronologies were arranged by geographic region to illustrate patterns of low growth, a proxy for drought. The chronologies are grouped by region, from roughly north to south; the northern and central Great Plains, Colorado Front Range, eastern Colorado, and northeastern New Mexico, western Colorado, central New Mexico, and Oklahoma. Annual ring-width indices for each of the 60 chronologies were ranked for the 285 years, 1680-1964. Rankings are shown for 1840-1860 (major drought years outline in black), the years that bracket the drought documented in the Colorado Front Range streamflow and eastern Colorado drought reconstructions, and for the 1860s central Great Plains drought, for comparison. The years in each chronology that fell within the lowest 50th, 25th and 10th percentiles of growth were highlighted. Red = ranking below the 10th percentile, orange = 10th - 24th percentile, yellow = 25th - 50th percentile, white = above median ranking.

To read or view the full study, please visit the Allen Press website.
It was published in Bulletin of the American Meteorological Society, 83, 1485-1493, October 2002.
This project was funded by grant ATM-9729751 from the
U.S. National Science Foundation.

Contact Us
National Oceanic and Atmospheric Administration
6 December 2002