TY - JOUR
T1 - Recent drought phase in a 73-year record at two spatial scales
T2 - Implications for livestock production on rangelands in the Southwestern United States
AU - McClaran, Mitchel P.
AU - Wei, Haiyan
N1 - Funding Information:
Precipitation and livestock use records were provided by the Santa Rita Experimental Range Digital Database, which is supported by USDA FS Rocky Mountain Research Station and University of Arizona. Research funded by USDA-CSREES Conservation Effects Assessment Project (CEAP) Program and USDA-AFRI Climate Change Program. D.P. Guertin provided suggestions that led to this analysis. M. Crimmins, D. Goodrich, Moran, G. Ruyle, and an anonymous reviewer provided valuable suggestions on earlier versions of the manuscript.
PY - 2014/10/15
Y1 - 2014/10/15
N2 - Livestock producers on rangelands are interested in drought at spatial scales of pastures (<25km2) and water developments (<14km2), and at temporal scales of seasons because summer convective storms are more spatially variable than cyclonic winter storms, and most forage production occurs in summer. Using a 73 y record of monthly precipitation from a dense network of rain gauges (0.1km-2) and temperature from PRISM we interpolate drought conditions across a large-scale 225km2 area in southern Arizona USA, and at a small-scale of 100 separate 1.5km×1.5km cells across the large area. We (1) compared the standardized precipitation index (SPI) and standardized precipitation and evapotranspiration index (SPEI) because the latter includes trends in temperature, (2) calculated the indices for each of the 100 grid cells for winter (Oct-May), summer (Jun-Sep) and water year (Oct-Sep) periods, and (3) compared the most recent 17 y (1996-2012) to the previous 56 y (1940-1995) because drying and warming trends appeared since 1996. We defined drought as the ≤20th percentile, which is the 15 driest years in the 73 y record for each small-scale cell and the large-scale 225km2 area. At the large-scale since 1996, temperature increased for all seasons by ~0.9°C, frequency of water year and winter drought increased >3 fold to 40-65% of years but frequencies did not differ between SPI and SPEI; and the frequency of summer drought did not change after 1996. At the small-scale, the extent of drought increased in winter and water years and decreased in summer since 1996 when using SPEI, but did not change when using SPI. Since 1996, the chances of a management unit-sized drought patch (1-10 contiguous cells) increased in winter, water year, and summer when using SPEI; but not in summer when using SPI. Scaling relationships show extensive drought patches (>20% of large area) when the large-scale is near average conditions, and those patches are larger in summer than winter. Increased drought frequency and patchy spatial distribution of drought have implications for herd structure and herd movements among pastures to avoid economic losses and overgrazing in drought patches. We propose increased efforts to detect drought patches with on-site gauge networks and temperature monitoring as well as remote sensing of precipitation patterns and vegetation indices.
AB - Livestock producers on rangelands are interested in drought at spatial scales of pastures (<25km2) and water developments (<14km2), and at temporal scales of seasons because summer convective storms are more spatially variable than cyclonic winter storms, and most forage production occurs in summer. Using a 73 y record of monthly precipitation from a dense network of rain gauges (0.1km-2) and temperature from PRISM we interpolate drought conditions across a large-scale 225km2 area in southern Arizona USA, and at a small-scale of 100 separate 1.5km×1.5km cells across the large area. We (1) compared the standardized precipitation index (SPI) and standardized precipitation and evapotranspiration index (SPEI) because the latter includes trends in temperature, (2) calculated the indices for each of the 100 grid cells for winter (Oct-May), summer (Jun-Sep) and water year (Oct-Sep) periods, and (3) compared the most recent 17 y (1996-2012) to the previous 56 y (1940-1995) because drying and warming trends appeared since 1996. We defined drought as the ≤20th percentile, which is the 15 driest years in the 73 y record for each small-scale cell and the large-scale 225km2 area. At the large-scale since 1996, temperature increased for all seasons by ~0.9°C, frequency of water year and winter drought increased >3 fold to 40-65% of years but frequencies did not differ between SPI and SPEI; and the frequency of summer drought did not change after 1996. At the small-scale, the extent of drought increased in winter and water years and decreased in summer since 1996 when using SPEI, but did not change when using SPI. Since 1996, the chances of a management unit-sized drought patch (1-10 contiguous cells) increased in winter, water year, and summer when using SPEI; but not in summer when using SPI. Scaling relationships show extensive drought patches (>20% of large area) when the large-scale is near average conditions, and those patches are larger in summer than winter. Increased drought frequency and patchy spatial distribution of drought have implications for herd structure and herd movements among pastures to avoid economic losses and overgrazing in drought patches. We propose increased efforts to detect drought patches with on-site gauge networks and temperature monitoring as well as remote sensing of precipitation patterns and vegetation indices.
KW - Drought patches
KW - Rotational grazing
KW - Santa Rita experimental range
KW - Standardized precipitation and evapotranspiration index
KW - Standardized precipitation index
UR - http://www.scopus.com/inward/record.url?scp=84903482052&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84903482052&partnerID=8YFLogxK
U2 - 10.1016/j.agrformet.2014.06.004
DO - 10.1016/j.agrformet.2014.06.004
M3 - Article
AN - SCOPUS:84903482052
SN - 0168-1923
VL - 197
SP - 40
EP - 51
JO - Agricultural and Forest Meteorology
JF - Agricultural and Forest Meteorology
ER -