1AUGUST 2005 WATSON AND LUCKMAN 2847

Spatial Patterns of Preinstrumental Moisture Variability in the Southern Canadian Cordillera

EMMA WATSON Climate Research Branch, Meteorological Service of , Environment Canada, Downsview, Ontario, Canada

BRIAN H. LUCKMAN Department of Geography, University of Western Ontario, London, Ontario, Canada

(Manuscript received 7 May 2004, in final form 23 October 2004)

ABSTRACT

Extreme wet and dry intervals of the last 350 yr in the Canadian Cordillera and adjacent United States are examined using a network of 25 tree-ring-based precipitation and Palmer Severity Index (PDSI) reconstructions. Reconstructed twentieth-century-mapped patterns compare well with patterns based on the instrumental records at both annual and decadal scales. During the most extreme events, dry conditions occurred over the entire area. The longest widespread drought in the last 350 yr occurred from 1917 to 1941. Shorter intervals of more severely dry conditions occurred in the early 1720s, 1750s, 1790s, 1860s–70s, and the 1890s. Many of the driest individual years and most extreme dry periods of Ͻ7yrare reconstructed for the eighteenth century. The longest, wettest periods identified by these reconstructions occurred in the early twentieth century. In agreement with published studies that explore links between instrumental precipitation records from the region and conditions in the Pacific Ocean, the reconstructed records show that drier (wetter)-than-normal conditions are associated with El Niño (La Niña) events and the positive (negative) phase of the Pacific decadal oscillation (PDO).

1. Introduction States (Cook et al. 1996, 1999) has demonstrated the tremendous potential for tree-ring-based reconstruc- The majority of dendroclimatic work conducted in tions to document spatial as well as temporal patterns the southern Canadian Cordillera has focused on gen- of preinstrumental climate variability. Reconstruction erating estimates for individual climatic time series: ei- of such patterns allows us to compare the magnitude ther for a single meteorological station (e.g., Watson and spatial extent of documented extreme events (e.g., and Luckman 2001a) or a regionally representative the 1920s–1930s drought) with those in the preinstru- time series (e.g., Luckman et al. 1997; Wilson and Luck- mental period. This type of information cannot be elic- man 2003). Although these studies have greatly ex- ited from individual station reconstructions and is panded our knowledge of climatic variability at sites highly significant because of the relatively greater eco- within the region, they have not systematically explored nomic and social costs of widespread prolonged variability across the region. Climate reconstructions drought vis-à-vis shorter, more localized severe events. for individual records contain local-, regional- and Much of the southern Canadian Cordillera and the large-scale signals that cannot be differentiated, or veri- neighboring prairies lie in the rainshadow of adjacent fied, without the broader view and replication that a mountains and consequently have low mean annual network of sites affords. The recent development of a precipitation totals (often less than 400 mm). The spa- gridded network of Palmer Drought Severity Index tial and temporal variability of precipitation is there- (PDSI) reconstructions for the contiguous United fore a significant control of many economic activities and an important conditioner for forest fire activity within the region. The extreme forest fire season of Corresponding author address: Dr. Emma Watson, Climate Re- search Branch, Meteorological Service of Canada, 4905 Dufferin 2003 in the interior was preceded by Street, Downsview, ON, M3H 5T4, Canada. three years of drought and cost the province an esti- E-mail: [email protected] mated 700 million dollars, destroyed over 300 homes,

© 2005 American Meteorological Society

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2848 JOURNAL OF CLIMATE VOLUME 18 and led to the evacuation of over 45 000 people (Filmon 2004). Watson and Luckman (2001a, 2004) have developed tree-ring-based reconstructions of annual precipitation for 13 meteorological stations across this region. These data, combined with precipitation/PDSI reconstruc- tions for the Rocky Mountain foothills, the , and adjacent United States provide the first dataset that can be used to explore spatial aspects of preinstrumental precipitation in western Canada. This study complements earlier research into temporal char- acteristics of dry and wet events in the region using these data (Watson and Luckman 2004). In this paper, we present maps of the spatial extent of dry and wet periods in the southern Canadian Cordil- lera over the past 350 yr. The periods identified during the twentieth century can be compared with those in FIG. 1. Location of the tree-ring reconstructions used in this study. Annual (Jul–Jun) precipitation reconstructions developed the preinstrumental period. This allows us to determine by Watson and Luckman (2004) are PG ϭ Prince George, BC ϭ whether these earlier events were more extensive and/ Big Creek, WL ϭ Williams Lake, JA ϭ Jasper, NT ϭ North or severe than those identified in the instrumental rec- Thompson, LI ϭ Lillooet, LY ϭ Lytton, WE ϭ Westwold, BA ϭ ord (e.g., the infamous drought of the 1920s–30s). It Banff, SU ϭ Summerland, OL ϭ Oliver, CR ϭ Cranbrook, and ϭ also provides data about the past frequency of wide- WA Waterton. Here, CM is an annual precipitation reconstruc- tion (Aug–Jul for a combined Fort Macleod and Calgary record) spread dry and wet events that persist over different using limber pine (Case and MacDonald 1995). SAU is an annual time scales across the region. To improve our under- precipitation reconstruction (Aug–Jul) for Maple Creek, Sas- standing of the causes of widespread events, we provide katchewan, using white spruce from the Cypress Hills (Sauchyn a preliminary discussion of their possible association and Beaudoin 1998). The grid points (GPs) are gridded Jun–Aug with conditions in the Pacific Ocean. PDSI reconstructions from Cook et al. (1999). The shaded area in Canada represents the approximate location of the Palliser Tri- angle, which is an extension of the U.S. northern Great Plains 2. Datasets and reconstruction quality (also shaded). a. Datasets used Although proxy records of two different variables The primary dataset used in this analysis consists of (precipitation and PDSI) are used in this study, the two 13 annual (prior July to current June) precipitation re- moisture measurements are comparable. The PDSI constructions developed from a network of Pseudot- (Palmer 1965) is a measure of meteorological drought suga menziesii (Douglas fir) and Pinus ponderosa (pon- that incorporates precipitation, potential evapotranspi- derosa pine) tree-ring chronologies sampled at low el- ration, and antecedent and runoff (Ola- evation sites in the southern Canadian Cordillera dipo 1985). Monthly PDSI values have considerable (Watson and Luckman 2001b, 2004). In addition, to month-to-month persistence built into them (Hu and expand spatial coverage, we utilize data from 10 adja- Wilson 2000) and therefore represent conditions over a cent grid points from the June–August PDSI estimates much longer time period than monthly precipitation. in the northwestern United States (Cook et al. 1999) Annual instrumental precipitation totals (summer– plus annual (prior August to current June) precipita- summer) correlate highly with summer PDSI values in tion reconstructions for the Rocky Mountain foothills the southern Canadian Cordillera (Watson and Luck- (Case and MacDonald 1995) and Canadian Prairies man 2002). For example, the mean correlation between (Sauchyn and Beaudoin 1998; Fig. 1). Although these August and July precipitation at 17 stations in the area sites are unevenly distributed across the region (Fig. 1), and the June–August (JJA) PDSI (the period used by Watson and Luckman (2004) demonstrate that the re- Cook et al. 1999) is 0.79. As others report similar results constructions from the southern Canadian Cordillera (e.g., Stahle et al. 2000) we consider these two param- correlate significantly with gridded instrumental pre- eters to be comparable.1 cipitation data over the region and define coherent spa- tial patterns. This indicates that these reconstructions 1 Few instrumental PDSI records are available for the southern capture regional precipitation variability and are there- Canadian Cordillera and therefore preference was given to recon- fore appropriate for use in this study. structing precipitation variables (Watson and Luckman 2002).

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2849 b. Comparing instrumental and reconstructed years. Mean correlation and congruence coefficients precipitation patterns calculated between the pairs of decadal maps (similar to the maps of widespread anomalies presented later in The individual reconstruction models pass conven- this paper) are higher than for individual years (0.78 tional verification tests that evaluate the year-to-year and 0.79; Fig. 4). performance and levels of explained variance in the time series over several intervals (Watson and Luck- man 2004; Cook et al. 1999; Sauchyn and Beaudoin 1998; Case and MacDonald 1995). However, these tests 3. Results do not evaluate the collective ability of the reconstruc- a. Evaluation of the relative severity of prolonged tions to reconstruct spatial patterns. The degree of simi- dry and wet periods larity between spatial patterns may be assessed visually The magnitude of prolonged dry and wet intervals and by calculating correlation and congruence coeffi- can be defined by ranking the mean standardized cients (Richman 1986; Cook et al. 1999). Congruence anomalies for all 25 reconstructions (Tables 1a and 2a) coefficients provide more information on the similarity or the 13 records from the southern Canadian Cordil- of the magnitude of the anomalies in each annual map lera (Tables 1b and 2b) for equivalent intervals of time. pair than correlation coefficients that provide a metric Only 1 (entire area, Table 1a) or 3 (13 site network, of how well the annual map patterns covary (Richman Table 1b) of the 10 most extreme dry individual years 1986; Cook et al. 1999). The mean correlation and con- occur in the twentieth century compared with 5 (Table gruence coefficients over the interval 1920–94 are 0.60 2a) or 3 (Table 2b) of the 10 wettest years. These results and 0.67, respectively (Fig. 2). This difference is not suggest that the twentieth-century instrumental record unexpected, as congruence coefficients are biased to does not contain many examples of the most extreme higher values than the correlation coefficient (Richman dry years experienced in this region in the last 350 yr. 1986; Cook et al. 1999). These values are similar to At the decadal time scale, a dry period originating in those identified by Cook et al. (1999) between instru- the first quarter of the eighteenth century (ϳ1717) is mental and reconstructed patterns of summer PDSI identified as well as a period in the 1790s (Table 1a over a much larger region. These results indicate that only, it is identified in shorter intervals in Table 1b) and the reconstructions consistently capture large-scale an- the early twentieth-century drought. Shorter dry inter- nual spatial patterns providing a convincing verification Ͻ of both the signal at individual sites and its spatial rep- vals (i.e., 7 yr) are identified for the late 1750s to early resentativeness. The ability to replicate the annual pat- 1760s, and the late 1860s through early 1870s (Tables 1a terns is, in and of itself, a verification of the skill of the and 1b). The 3-yr mean for the period 1661–63 is iden- individual reconstructions. tified as dry for the southern cordillera alone (Table Comparison of the reconstructed and instrumental 1b). Several of these shorter intervals (those in the early annual patterns (Fig. 2) indicate that only about 10% and mid-1700s) are also considerably more severe than (around 8 of 75, depending on the criterion used) of the equivalent periods during the early twentieth century. Ͼ years are poorly modeled. An examination of the in- At longer time scales ( 10 yr means), the docu- strumental and reconstructed patterns for these years mented drought that occurred in the early twentieth- (Fig. 3) is instructive. The annual contour maps in Fig. century dominates (Table 1). The reconstructions indi- 3 show reconstructed and instrumental annual precipi- cate that it is clearly the longest, most severe dry spell tation/PDSI values expressed in standard deviation to have affected the region in the past 350 yr. Within units based on the period 1920–94. In all cases, the this drought, the 3-yr interval 1929–31 was particularly reconstructed patterns include some points that are the dry (Tables 1a and 1b). The next most severe 20-yr incorrect sign (i.e., a reflection of local climatic or non- periods in the entire region in the past 350 yr are 1717– climatic anomalies) and others that are the correct sign 36 (mean Ϫ0. 25, ranked 14th), 1781–1800 (mean but underestimate (or overestimate) the magnitude of Ϫ0.18, ranked 21st), and 1840–59 (mean Ϫ0.15, ranked deviation. However, if one were to coarsely summarize 29th; not shown in Table 1a). the entire region as being either wet or dry (or half wet The strongest, most prolonged wet interval appears and half dry), the reconstructions perform well even for to have occurred in the early part of the twentieth cen- these poor years. When the annual maps are compos- tury (ϳ1897–1916: Tables 2a and 2b). There are some ited over longer intervals (Fig. 4), local anomalies are differences in the order of severity of other prolonged reduced by the averaging process making longer-term wet intervals between the entire region (Table 2a) and patterns more comparable than those for individual the southern Canadian Cordillera (Table 2b). In the

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2850 JOURNAL OF CLIMATE VOLUME 18

FIG. 2. Measures of agreement between the reconstructed and instrumental annual maps. (a) The number of stations available in each pair of annual maps. (b) Correlations between the instrumental and reconstructed maps in each year. Years with nonsignificant correlation coefficients are labeled prior to 1980, after which correlations become a less reliable measure because of the decreased number of observations. Note that the significance levels displayed do not account for any spatial autocorrelation that may exist in the dataset. They are only a guide to identify poorly modeled years. (c) Congruence coefficients for the same data. There is no theoretical sampling distribution for testing the significance of congruence coefficients (Richman 1986).

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2851

FIG. 3. A comparison of the instrumental and reconstructed precipitation/PDSI anomalies in the poorly modeled years identified in Fig. 2b. Data are expressed as standard deviations from the 1920–94 mean (or, for shorter records, the longest possible portion of this interval). series for the entire region (Table 2a), anomalous wet b. Evaluation of mapped patterns periods occurred in the first quarter of the twentieth century, 1900–19 and in the late seventeenth century In the previous section, reconstructed dry and wet (ϳ1666–85). In the series for the southern Canadian events were ranked according to their severity across Cordillera, the second, third, and fourth wettest 20-yr the region for a number of time scales. This approach periods include the mid-twentieth century (ϳ1947–66), excludes information on their spatial extent and less the period 1801–20, and the late seventeenth century severe but more extensive may not be identi- (as seen in the regional series). At shorter intervals, the fied. It is also difficult to define the most appropriate latter part of the 1870s to the mid-1880s is also recon- length of period to identify dry and wet events as they structed as being particularly wet. overlap at a number of time scales (Tables 1 and 2). In

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2852 JOURNAL OF CLIMATE VOLUME 18

FIG. 4. A comparison of patterns of instrumental and reconstructed precipitation/PDSI anomalies by decade from 1920 to 1989. Here, r and cc refer to the correlation and congruence coefficients between each pair of decadal maps. Data are expressed as standard deviations from the 1920–94 mean.

the following sections, we present maps of dry and wet Conditions during two of the most extreme dry and wet events (defined according to their spatial extent) at years reconstructed (Fig. 5a; Tables 1a and 2a) display three time scales: annual events, 5–10-yr periods and large areas where precipitation/PDSI is more than one intervals Ͼ10 yr in length. standard deviation from the mean and roughly one quarter of the values are more than two standard de- 1) ANNUAL MAPS viations from the mean. These diagrams indicate that Patterns of precipitation anomalies may be mapped for individual years strong anomalies of the same sign over the network for each year from about 1640 to can occur across the entire region. 1996. Prior to 1640, there are Ͻ10 records available and Figures 5b and 5c show less extreme sequences of the contoured patterns are unduly influenced by the consecutive years that are highly significant in the his- distribution of the records. Between 1640 and 1996, the tory of settlement of the Canadian west. The Palliser number of records varies between 9 and 25 because of Triangle is a semiarid, drought-prone region of the Ca- the different lengths of the individual reconstructions. nadian Prairies located in southeast and south-

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2853

TABLE 1. The 10 lowest mean standardized anomalies ranked over different intervals for (a) the entire mapped area and (b) the southern Canadian Cordillera from 1650 to 1996. The number of reconstructions (N) used for area (a) ranges from 25 (1978) to 10 (1650). For the southern Cordillera (b) N decreases from 13 in 1994 to 6 in 1650. All reconstructions were converted to standardized anomalies relative to the 1831–1978 common period. The bold font identifies the first occurrence of a period within each interval. Italicized periods overlap with a period already identified in that interval. Gray shading identifies periods that overlap with the persistent (20-yr mean) drought identified for the years 1917–36.

(a) 1 Year 3 Year 5 Year 10 Year 20 Year 0.51؁ Ϫ1.46 1719–1721 Ϫ0.97 1717–1721 Ϫ1.01 1917–1926 Ϫ0.58 1918–1937 1869 1 Ϫ0.56 1917–1936 Ϫ0.50 1931–1922 0.83؁ Ϫ1.43 1929–1931 Ϫ0.93 1756–1760 1721 2 Ϫ0.46 1938–1919 0.55؁ Ϫ1.41 1717–1719 Ϫ0.89 1718–1722 Ϫ0.71 1717–1726 1756 3 Ϫ0.46 1941–1922 0.53؁ Ϫ1.37 1756–1758 Ϫ0.84 1922–1926 Ϫ0.68 1791–1800 1657 4 5 1741 Ϫ1.30 1739–1741 Ϫ0.83 1716–1720 Ϫ0.65 1918–1927 Ϫ0.52 1920–1939 Ϫ0.45 6 1931 Ϫ1.24 1718–1720 Ϫ0.80 1794–1798 Ϫ0.65 1929–1938 Ϫ0.52 1921–1940 Ϫ0.45 Ϫ0.65 1928–1937 Ϫ0.50 1916–1935 Ϫ0.39 1800–1796 0.79؁ Ϫ1.24 1869–1871 1800 7 8 1677 Ϫ1.24 1918–1920 Ϫ0.74 1918–1922 Ϫ0.64 1716–1725 Ϫ0.49 1915–1934 Ϫ0.35 9 1717 Ϫ1.22 1924–1926 Ϫ0.72 1793–1797 Ϫ0.62 1930–1939 Ϫ0.48 1923–1942 Ϫ0.35 10 1794 Ϫ1.10 1720–1722 Ϫ0.71 1755–1759 Ϫ0.61 1923–1932 Ϫ0.47 1924–1943 Ϫ0.32 0.25؁ 1736–1717 *(14) 0.19؁ 1800–1781 *(21) (b) 1 Year 3 Year 5 Year 10 Year 20 Year 0.56؁ Ϫ0.70 1917–1936 1931–1922 0.85؁ Ϫ1.78 1929–1931 Ϫ1.06 1756–1760 1869 1 2 1759 Ϫ1.34 1758–1760 Ϫ1.03 1757–1761 Ϫ0.80 1917–1926 Ϫ0.67 1918–1937 Ϫ0.53 3 1929 Ϫ1.33 1757–1759 Ϫ0.99 1922–1926 Ϫ0.78 1918–1927 Ϫ0.64 1919–1938 Ϫ0.50 4 1931 Ϫ1.25 1869–1871 Ϫ0.98 1929–1933 Ϫ0.75 1923–1932 Ϫ0.63 1916–1935 Ϫ0.48 5 1720 Ϫ1.18 1719–1721 Ϫ0.81 1918–1922 Ϫ0.71 1921–1930 Ϫ0.63 1920–1939 Ϫ0.47 Ϫ0.62 1921–1940 Ϫ0.47 1929–1920 0.70؁ Ϫ1.18 1759–1761 Ϫ0.79 1869–1873 1760 6 7 1741 Ϫ1.15 1924–1926 Ϫ0.79 1868–1872 Ϫ0.69 1924–1933 Ϫ0.60 1915–1934 Ϫ0.47 Ϫ0.46 1941–1922 0.59؁ Ϫ0.68 1717–1726 1932–1928 0.77؁ Ϫ1.06 1661–1663 1757 8 9 1922 Ϫ1.04 1922–1924 Ϫ0.74 1921–1925 Ϫ0.67 1919–1928 Ϫ0.55 1914–1943 Ϫ0.42 10 1663 Ϫ1.04 1930–1932 Ϫ0.74 1717–1721 Ϫ0.66 1925–1934 Ϫ0.55 1913–1942 Ϫ0.38 0.34؁ 1731–1712 *(13) 0.23؁ 1871–1852 *(31)

* Identifies the two most highly ranked nonoverlapping 20-yr periods. west . It is named after the British ex- persisted for several years. The examples presented in plorer Captain John Palliser who visited the region in Fig. 5 are not isolated extreme years but occur within the late 1850s and said that it was too dry to be suitable sequences where the pattern of departures may persist for agricultural settlement (Fig. 5b; Koshida 1992; Pal- for several years and thereby have a more pronounced liser 1859). In a later expedition during the 1870s, the impact than single year extreme events. In the rest of botanist John Macoun found the region to be suitable this paper we present a series of maps that explore for based on the comparatively wet condi- longer-term, widespread dry and wet anomalies across tions that prevailed during that interval (Fig. 5c; Nem- the region. anishen 1998; Koshida 1992). His findings encouraged settlement of the area that we now know receives an 2) WIDESPREAD DRY AND WET INTERVALS average of Ͻ350 mm of precipitation per year but pro- duces 75% of Canadian wheat (Koshida 1992). Agree- Major widespread dry and wet periods reconstructed ment between these documentary sources and the tree- for the southern Canadian Cordillera over the past ring-based reconstructions provides mutual verification ϳ350 yr are presented in Figs. 6 and 7. These periods while demonstrating the important context that these are defined as intervals longer than 4 yr when Ͼ50% longer records provide for evaluating more recent in- (Ͼ60% before 1700 when n Ͻ 8) of the 13 annual pre- strumental climate records. cipitation reconstructions developed for British Colum- Although it is important to examine the magnitude of bia and Alberta are below (dry) or above (wet) their departures in widespread patterns that occur in particu- 1831–1978 average (the period common to all 25 recon- lar years (e.g., for water resource planning), it is per- structions). In some cases, single years within the inter- haps more important to study widespread patterns that val fall below this 50% criterion. These intervals

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2854 JOURNAL OF CLIMATE VOLUME 18

TABLE 2. The 10 highest mean standardized anomalies ranked over different intervals. For explanation see Table 1.

(a) 1 Year 3 Year 5 Year 10 Year 20 Year 1 1715 1.38 1879–1881 0.86 1990–1994 0.78 1989–1988 0.54 1897–1916 0.40 2 1672 1.36 1877–1879 0.81 1877–1881 0.77 1666–1675 0.52 1898–1917 0.40 3 1697 1.32 1750–1752 0.81 1750–1754 0.73 1907–1916 0.49 1899–1918 0.35 4 1750 1.28 1994–1996 0.78 1912–1916 0.72 1667–1676 0.47 1978–1997 0.34 5 1996 1.22 1752–1754 0.75 1993–1997 0.72 1873–1882 0.46 1896–1915 0.32 6 1775 1.17 1666–1668 0.73 1876–1880 0.70 1664–1673 0.45 1979–1998 0.30 7 1913 1.16 1992–1994 0.73 1992–1996 0.69 1988–1997 0.45 1900–1919 0.28 8 1981 1.16 1878–1880 0.69 1666–1670 0.67 1876–1885 0.45 1977–1996 0.28 9 1916 1.16 1899–1902 0.68 1749–1753 0.62 1875–1884 0.44 1666–1685 0.27 10 1942 1.15 1911–1913 0.68 1875–1879 0.62 1665–1674 0.43 1894–1913 0.27 (b) 1 Year 3 Year 5 Year 10 Year 20 Year 1 1697 2.04 1696–1698 1.33 1694–1698 1.10 1691–1700 0.70 1897–1916 0.41 2 1672 1.67 1695–1697 1.22 1693–1697 1.01 1693–1702 0.66 1947–1966 0.39 3 1752 1.46 1697–1699 1.07 1696–1701 1.00 1690–1699 0.65 1898–1917 0.39 4 1996 1.42 1900–1902 1.00 1695–1699 0.97 1666–1675 0.65 1801–1820 0.39 5 1900 1.39 1879–1981 0.96 1900–1904 0.89 1692–1701 0.64 1948–1967 0.38 6 1789 1.34 1670–1672 0.93 1990–1994 0.80 1689–1698 0.64 1664–1683 0.37 7 1819 1.24 1750–1752 0.91 1877–1881 0.77 1665–1674 0.62 1663–1682 0.37 8 1670 1.23 1805–1807 0.88 1750–1754 0.76 1664–1673 0.60 1662–1681 0.36 9 1879 1.21 1804–1806 0.85 1899–1903 0.75 1667–1676 0.59 1802–1821 0.36 10 1981 1.20 1752–1754 0.84 1672–1676 0.72 1694–1703 0.57 1665–1684 0.35

are defined using the method outlined in Watson and ditional data are included to provide more information Luckman (2004). about the spatial scale of the reconstructed anomalies. The aggregate maps produced reduce the localized Maps of the dry and wet periods are arranged according anomalies created by poor estimates for individual to their length to enable direct comparison of periods of years in specific reconstructions. The standardized similar length. The patterns or the spatial scale of dry/ anomalies on these maps are defined by the standard wet events of different lengths should be more directly deviations of the annual series over the 1831–1978 com- comparable than the actual magnitude of departures. mon period. However, averaging reduces the absolute (i) Dry periods Ͻ10 yr in length magnitude of these anomalies that display a reduced range of departures from those seen in the annual maps The majority of the spatially extensive dry periods (e.g., Figure 5; Table 1). For this reason, the absolute identified are less than 10 yr in length (Fig. 6). The most magnitudes of the anomalies in the annual (Figs. 3 and severe are in the mid-1750s, which coincides with dry 5) and composite maps (Figs. 4 and 6–8) are not directly conditions over much of the western United States, and comparable as they express anomalies averaged over the 1790s. On average ϳ60% of the 24 reconstructions different time intervals. The absolute values are com- that extend back to the 1790s drought have values more pared for specific intervals in Tables 1 and 2. than half a standard deviation below their mean with The maps of the extended dry and wet intervals in- particularly dry conditions reconstructed in the Rockies clude data from precipitation and PDSI reconstructions and the foothills. A streamflow reconstruction for the in adjacent areas that were not used in defining these North Saskatchewan River, which flows eastward from events. The coincident occurrence of pronounced mois- this region, reconstructs very low flows during the 1790s ture anomalies in reconstructions from adjacent areas (Case and MacDonald 2003). verifies the significance of the periods identified using Wolfe et al. (2001) document a widespread period of the smaller 13-reconstruction dataset. If the spatial dune activity in the Great Sand Hills region of the Pal- scale was broadened and all 25 reconstructions were liser Triangle in Saskatchewan during the early 1800s. used to define extreme periods, fewer intervals would They attribute this activity to the dry conditions that be identified because it is less common to find droughts prevailed throughout much of the 1700s and the ex- that affect the entire region. Smaller-scale dry/wet tremely dry conditions of the 1790s [as based on pre- events unique to the southern Canadian Cordillera cipitation reconstructions in Case and MacDonald would not be as readily identified. Therefore these ad- (1995) and Sauchyn and Beaudoin (1998), also used in

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2855

FIG. 5. Annual maps showing reconstructed moisture conditions over a number of intervals. (a) Two of the driest and wettest years reconstructed (see Tables 1 and 2); n is the number of reconstructions used to produce the map. The percentage of records that are more than 0.5 standard deviations (sd) from their 1831–1978 mean are listed. (b), (c) Selected years during the nineteenth century (refer to text for discussion). The maps show reconstructed conditions for each record expressed in standard deviation units from the mean of the period common to all 25 reconstructions (1831–1978).

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2856 JOURNAL OF CLIMATE VOLUME 18

FIG. 6. Widespread dry periods in the southern Canadian Cordillera after 1650. The dry intervals are defined as periods when Ͼ50% (Ͼ60% before 1700) of the 13 reconstructions developed by Watson and Luckman (2004; Fig. 1) show below average precipitation for Ͼ4 consecutive years. Single years within some intervals fall below the 50% criterion. The z scores are derived using the mean and standard deviation calculated over the common period of all reconstructions (1831–1978). (a), (b) Dry intervals Ͻ and Ͼ10 yr are plotted sequentially, and (c) subdivisions of the 1917–41 period are shown. The number of years in each period is given in brackets.

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2857

FIG. 7. Widespread wet periods in the southern Canadian Cordillera after ϳ1650. For explanation see Fig. 6.

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2858 JOURNAL OF CLIMATE VOLUME 18 this study]. The results presented here support this in- conditions with the driest records located in British Co- terpretation as the greatest concentration of dry peri- lumbia. Mean moisture conditions are reconstructed as ods identified in the past 350 yr occurs during the 1700s. below normal for all records during the 15-yr interval In addition to the 1750s and 1790s, dry intervals are 1717–31. During the driest portion of this interval, identified from 1701–08 and 1768–72. During both of 1717–21, reconstructed values are on average more these periods, most of the PDSI records from the than one standard deviation below the common period United States show wetter than normal conditions. mean. Evidence for this drought is also indicated by The 1868–75 drought is not particularly striking in below-normal streamflow estimates for the North Sas- terms of the magnitude of the standardized anomalies katchewan, South Saskatchewan, and Saskatchewan or its duration, but it appears to have had a strong Rivers (Case and MacDonald 2003), and negative sum- impact on the landscape. In particular, the year 1869 mer PDSI estimates on the Canadian Prairies for this was very dry (80% of the 25 reconstructions have val- interval (Sauchyn and Skinner 2001). ues more than one standard deviation below the Prolonged dry conditions are estimated for much of mean).2 The widespread dry conditions estimated for the study area over the years 1839–59. This drought 1869 coincide with fire-scar evidence of an extreme fire corresponds with low streamflow estimates for the year in the Cariboo Forest region of southern British South Saskatchewan and Saskatchewan Rivers (Case Columbia (Daniels and Watson 2003; Gray et al. 2002) and MacDonald 2003) and a drought in the western and with documented crop failure and grasshopper Great Plains of the United States from 1845 to 1856 plagues on the Canadian Prairies in 1868 (Phillips that may be linked with the decline in the bison popu- 1990). lation of the area (Woodhouse et al. 2002). Fye et al. The period 1889–97 was also very dry across the re- (2003) document drought over large areas of the west- gion with the driest conditions estimated for the eastern ern United States for much of this interval. Dry condi- reconstructions. Major forest fires burned approxi- tions are also reconstructed for the Canadian Prairies mately 9% and 20% of Banff and Jasper National Parks during the 1850s and 1860s (Sauchyn and Skinner in the Canadian Rockies in 1889 (Luckman 1998). Phil- 2001). lips (1990) reports that 9 yr of drought during the 1890s The longest severe drought identified in the recon- led to farm abandonment on the Canadian Prairies and structions occurred during the period 1917–41. All re- severe drought also occurred in the U.S. Great Plains constructions show drier-than-normal conditions dur- during the 1890s (Woodhouse and Overpeck 1998). ing this well-known dry interval. Analysis of the annual The 1750s, 1790s, 1860s, and 1890s droughts are re- maps (not shown) reveals that the region of greatest corded in a tree-ring-based reconstruction of July PDSI dryness differs between the first and second half of the developed for the Canadian Prairies by Sauchyn and drought. During the 1920s, the most extreme dry con- Skinner (2001). ditions are seen for southern interior British Columbia The most recent short (i.e., Ͻ10 yr), dry interval whereas, during the 1930s, the driest conditions are identified for the region occurred in the 1980s (1985– seen in the southern part of the study region (Fig. 6c). 89).3 This drought had severe impact on agriculture in However, the comparison of the instrumental and re- western Canada and an estimated 10% of farmers and constructed decadal maps for the 1920s (Fig. 4) farm workers left the agricultural sector in 1988 (Phil- indicates that the severely dry area in the 1920s ex- lips 1990). The year 1985 was one of the worst fire years tended further south. Over the twentieth century, 7 of on record in British Columbia and cost roughly 300 the 10 driest years4 in the instrumental records and 8 of million dollars in timber losses and fire-fighting ex- those in the reconstructed records occur during this penses (Phillips 1990). drought.

(ii) Dry periods Ͼ10 yr in length (iii) Wet periods The mapped patterns in Fig. 6 show five discrete dry Eleven regionally extensive wet periods dating from periods greater than 10 yr in length during the past 350 the late seventeenth century onward are displayed in yr. The earliest reported here is 1641–53 during which Fig. 7. This diagram demonstrates that it is not uncom- time 8 of the 10 available records show below-normal mon to have wet periods that span the entire study

2 The reconstructed 1869 year is July 1868–June 1869. 4 Calculated as the mean standard deviation from the 1920–94 3 The PDSI reconstructions are not included in this map be- mean for the set of 25 instrumental records and the set of 25 cause they terminate in 1978. reconstructed records.

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2859 region. Sample depth is limited (n ϭ 10) during the tive PDO) is not as strong. Sample depth is limited for earliest of these intervals (1664–68). All 23 reconstruc- situations where the two patterns are out of phase, but tions covering the 1749–55 interval show above-normal generally the patterns are more mixed. Composite values averaging 0.58 standard deviations above the maps produced using the reconstructed records for the mean, with the greatest anomalies in the western part of same groupings are very similar (Fig. 8), again demon- the study region. Wetter-than-normal conditions are re- strating the skill of the reconstructions. Unfortunately, constructed for almost all of the records during the in- it is not possible to extend this analysis back over the terval 1778–82. The final two short wet periods occur in past 350 yr, as no definitive PDO reconstructions are the late twentieth century: 1980–84 and the more wide- presently available.5 spread wet period in 1990–94. Both of these periods are Drought conditions across much of the United States also seen in the instrumental records from the region. (particularly the southwest and extending into the Six wet periods greater than 10 yr in length are iden- Great Plains) have been linked to La Niña conditions in tified in the reconstructions. The 12-yr period 1689– the Pacific Ocean (Cole and Cook 1998; Cole et al. 1700 is wet across much of the region with the excep- 2002). As mentioned above, La Niña events typically tion of a few records in the southern interior of British produce wetter-than-normal conditions in the study re- Columbia and the most easterly sites. The next four wet gion (Fig. 8). Cole et al. (2002) provide a list of persis- periods span parts of the nineteenth century. Notable tent droughts that affect the continental United States among these is the period 1819–30, which coincides and the ENSO-sensitive southwest. In the discussion with the largest flood reconstructed for the Red River that follows, the timing of these droughts is compared in in 1826 (St. George and Nielsen 2003). with conditions in the southern Canadian Cordillera to Although the Red River drainage basin is east of the determine whether widespread anomalies of the same study region, these wet conditions could have extended sign occur across western North America and if such across the prairies as wet conditions are reconstructed anomalies coincide with documented La Niña events. across much of the western United States and the U.S. As one would expect (based on the ENSO response Great Plains for part of this interval (Fye et al. 2003). of the two regions described above), there are instances The final two wet periods (1898–1916 and 1942–60) when conditions in the two regions are opposite (e.g., bracket the 1920s–1930s drought. The wet period in the during the persistent U.S. droughts of the early 1750s, early 1900s, which appears to have been more severe in 1780s, 1820s, and the 1950s; see Figs. 4 and 7). There are the eastern part of the study region, coincides with high other periods when persistent drought in the United flows on the North Saskatchewan, South Saskatchewan, States coincides with near-normal conditions in the and Saskatchewan Rivers (Case and MacDonald 2003) southern Canadian Cordillera (e.g., 1855–65). This is and is also seen in PDSI records (both instrumental and not unexpected as the ENSO weakens reconstructed) from the western United States (Fye et northward; there is great variability in characteristics of al. 2003). different ENSO events, and ENSO to surface climate can be unstable as demonstrated by Cole and Cook (1998) for the continental United 4. Links with Pacific forcing States. Sometimes, however, persistent drought affects much of the United States and the southern Canadian Precipitation variability in western Canada and the Cordillera concurrently (e.g., the 1840s, 1890s, and U.S. Pacific Northwest has been linked with known cir- 1930s; Fig. 6). While it is beyond the scope of this paper culation changes in the Pacific Ocean, particularly the to explore the causes of such widespread drought, it is El Niño–Southern Oscillation (ENSO) and the Pacific interesting to note that the droughts of the 1840s and Decadal Oscillation (PDO; Mantua et al. 1997; Ger- 1930s are not particularly strong in the ENSO-sensitive shunov and Barnett 1998; McCabe and Dettinger 1999; region (i.e., the U.S. Southwest) and the concurrent Shabbar and Skinner 2004; Bonsal et al. 1993). The drought of the 1890s does not correspond with La Niña influence of these patterns can be seen in the composite conditions (Cole et al. 2002). These comparisons indi- maps of moisture anomalies in the instrumental records cate that widespread droughts (i.e., droughts that affect during the twentieth century (Fig. 8). These composites show that El Niño and positive PDO years are gener- ally drier, whereas La Niña and negative PDO years are 5 Several tree-ring-based reconstructions of the PDO have been generally wetter across the region. More of the records developed (e.g., Gedalof and Smith 2001; D’Arrigo et al. 2001; have negative mean anomalies when ENSO and PDO Biondi et al. 2001), but they often disagree with one another prior are in phase but the opposite pattern (La Niña, nega- to the twentieth century.

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2860 JOURNAL OF CLIMATE VOLUME 18

FIG. 8. Composites of instrumental and reconstructed precipitation and PDSI for sites in the southern Canadian Cordillera and northwestern United States over the twentieth century for (a) El Niño and La Niña years, (b) positive and negative PDO years, and (c) in- and out-of-phase ENSO/PDO pairings (Gershunov and Barnett 1998). The instrumental and reconstructed conditions are mapped in standard deviation units from the period 1920–94 (note that some records do not span the entire interval). El Niño and La Niña years are identified using the bivariate ENSO index (Smith and Sardeshmukh 2000) for 1871–2001. Nov–Mar PDO data for 1900–2003 are available online at http://tao.atmos.washington.edu/pdo/.

the western United States and the southern Canadian Atlantic multidecadal oscillation (AMO); Enfield et al. Cordillera) appear to occur in the absence of La Niña 2001; Gray et al. 2003].6 events that are typical of the twentieth century (i.e., something other than La Niña appears to have forced these very large-scale droughts). It is possible that their 6 Composite maps produced for the different phases of the synchronous occurrence is a coincidence. However, the AMO show weak but generally consistent anomalies across the persistence and scale of these droughts may be related southern Cordillera (wet with the positive AMO and dry with the negative AMO), which is consistent with recent work that shows to positive feedbacks that enhance initial land surface nonsignificant but generally positive correlations between the anomalies or to more complex interactions with long- AMO and summer moisture availability over the region (Shabbar term modes of atmospheric/oceanic variability [e.g., the and Skinner 2004).

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2861

5. Summary and conclusions 79 is unusual in that many of the more northern and coastal PDSI grid points indicate wetter-than-normal This paper has demonstrated (through comparisons conditions. of instrumental and reconstructed precipitation Overall, the wet periods that have been identified are anomaly patterns) that a newly developed set of annual more regionally coherent than the dry intervals. The precipitation reconstructions (Watson and Luckman extensive wet periods identified for 1778–82, 1879–88, 2004) can be combined with data from adjacent areas to 1898–1916, and 1990–94 are similar in that they show explore spatial patterns of preinstrumental moisture very strong anomalies in eastern portions of the study conditions in the southern Canadian Cordillera. This is region and generally weaker anomalies in the United the first time this type of information has been available States and interior British Columbia (particularly the for the area. Maps of the most spatially extensive dry first two intervals). In contrast, the strongest anomalies events reveal that it is not uncommon for anomalies of during the 1749–55 and 1980–84 wet intervals are found the same sign to affect the entire study region. west of the Rocky Mountains. The three intervals 1689– Previous studies and instrumental records have docu- 1700, 1801–14, and 1819–30 have a number of features mented the severity and duration of the droughts that in common; they show strong anomalies in both south- affected much of southwestern Canada and the United western British Columbia and southwestern Alberta States during the 1920s and 1930s. An examination of and weak negative anomalies in south-central British preinstrumental dry events in the region suggest that Columbia and the southeast corner of the study region. this drought is of unmatched length and severity over Based on these results, the twentieth-century wet peri- the past 350 yr. Shorter intervals of more severely dry ods do not appear to be unprecedented in this region. conditions include the early 1720s, 1750s, 1790s, late A comparison of the occurrence of widespread 1860s through early 1870s, and the 1890s. Many of the drought (i.e., drought that is coincident in the study driest individual years and short dry periods (i.e., Ͻ7 region and the western United States) suggests that yr) are reconstructed for the eighteenth century. The these widespread events are not related to La Niña longest, wettest period identified in the reconstructions events (a strong control of drought in the southwestern occurred in the early twentieth century (ϳ1898–1917). United States). Although this observation is based on Most of the dry and wet periods identified are corrobo- only three common drought events, it suggests that cau- rated by the occurrence of synchronous events in inde- tion is warranted when evaluating widespread prein- pendent records from western North America. The re- strumental periods of drought solely in terms of ENSO sults suggest that the most severe extended dry (1917– events [e.g., the sixteenth-century megadrought (Stahle 36) and wet (1897–1916) intervals of the past 350 yr et al. 2000) identified in tree-ring reconstructions from occurred during the twentieth century, thus indicating Mexico to western Canada]. The coincident events in that the instrumental record provides a broad range of the two regions may not be related to a common forc- conditions that is well-suited for the calibration of tree- ing. However, much more work is needed to evaluate ring-climate relationships. the complex causes and persistence of drought in west- It is difficult to compare the spatial patterns of the ern North America. In particular, a very limited anomalies without also considering their severity amount of research has been conducted into relation- (which is not directly comparable—see previous discus- ships between atmospheric and oceanic circulation pat- sion). However, the dry events of the 1720s, 1750s, terns (and their interactions) and summertime precipi- 1790s, and early twentieth century appear to have been tation anomalies (i.e., the period most important to tree the most extensive, showing strong anomalies through- growth) in western Canada. Most of the available stud- out the southern cordillera and the adjacent United ies focus on the agriculturally important Canadian Prai- States. Extensive negative anomalies also prevailed ries or concentrate on winter conditions. Until a clearer during the 1890s but appear to have been more intense picture is available of the interaction between circula- in the more easterly sites. These results indicate that tion patterns and moisture anomalies in the southern although there appear to be analogs to the twentieth- Canadian Cordillera over the instrumental record, century drought in terms of pattern and severity in the causes of variability in reconstructions of summer pre- study region, none of the earlier events were as pro- cipitation will remain difficult to interpret. longed as that event. The dry periods in the early 1700s and late 1760s–early 1770s are similar in that the dry Acknowledgments. This research has been funded by conditions are concentrated in the interior Canadian the Natural Sciences and Engineering Research Coun- sites, whereas the U.S. PDSI grid points actually show cil of Canada, the Meteorological Service of Canada, wetter conditions. The dry period identified from 1968– and most recently a grant from the Canadian Founda-

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 2862 JOURNAL OF CLIMATE VOLUME 18 tion for Climate and Atmospheric Sciences. We thank Gray, S. T., J. L. Betancourt, C. L. Fastie, and S. T. Jackson, 2003: D. Sauchyn, R. Case, and G. MacDonald for providing Patterns and sources of multidecadal oscillations in drought- their reconstructions, M. Colenutt for contributing data sensitive tree-ring records from the central and southern Rocky Mountains. Geophys. Res. Lett., 30, 1316, doi:10.1029/ from Waterton Lakes National Park, and Walter Skin- 2002GL016154. ner and Xuebin Zhang for their comments on an earlier Hu, Q., and G. D. Wilson, 2000: Effects of temperature anomalies draft of the manuscript. on the Palmer Drought Severity Index in the central United States. Int. J. Climatol., 20, 1899–1911. REFERENCES Koshida, G., 1992: About drought in Canada. Canadian Climate Biondi, F., A. Gershunov, and D. Cayan, 2001: North Pacific dec- Program, Environment Canada Publication, Downsview, adal climate variability since AD 1661. J. Climate, 14, 5–10. ON, Canada, 41 pp. Bonsal, B. R., A. K. Chakravarti, and R. G. Lawford, 1993: Tele- Luckman, B. H., 1998: Landscape and in the cen- connections between North Pacific SST anomalies and grow- tral Canadian Rockies during the 20th century. Can. Geogr., ing season extended dry spells on the Canadian Prairies. Int. 42, 319–336. J. Climatol., 13, 865–878. ——, K. R. Briffa, P. D. Jones, and F. H. Schweingruber, 1997: Case, R. A., and G. M. MacDonald, 1995: A dendroclimatic re- Tree-ring based reconstruction of summer temperatures at construction of annual precipitation on the western Canadian the Columbia Icefield, Alberta, Canada, A.D. 1073–1983. prairies since A.D. 1505 from Pinus flexilis James. Quat. Res., Holocene, 7, 375–389. 44, 267–275. Mantua, N. J., S. R. Hare, U. Zhang, J. M. Wallace, and R. C. ——,and——, 2003: Tree ring reconstructions of streamflow for Francis, 1997: A Pacific interdecadal climate oscillation with three Canadian prairie rivers. J. Amer. Water Res. Assoc., 39, impacts on salmon production. Bull. Amer. Meteor. Soc., 78, 703–716. 1069–1079. Cole, J. E., and E. R. Cook, 1998: The changing relationship be- McCabe, G. J., and M. D. Dettinger, 1999: Decadal variations in tween ENSO variability and moisture balance in the conti- the strength of ENSO teleconnections with precipitation in nental United States. Geophys. Res. Lett., 25, 4529–4531. the western United States. Int. J. Climatol., 19, 1399–1410. ——, J. T. Overpeck, and E. R. Cook, 2002: Multiyear La Niña Nemanishen, W., 1998: Drought in the Palliser Triangle: A Provi- events and persistent drought in the contiguous United sional Primer. Prairie Farm Rehabilitation Committee, Ag- States. Geophys. Res. Lett., 29, 1647, doi:10.1029/ riculture and Agri-Food Canada, 58 pp. 2001GL013561. Oladipo, E. O., 1985: A comparative performance analysis of Cook, E. R., D. M. Meko, D. W. Stahle, and M. K. Cleaveland, three meteorological drought indices. J. Climatol., 5, 655–664. 1996: Tree-ring reconstructions of past drought across the Palliser, J., 1895: Papers relative to the exploration of that portion coterminous United States: Tests of regression method and of British North America which lies between the northern calibration/verification results. Tree Rings, Environment and branch of the River Saskatchewan and the frontier of the Humanity,J.S.Dean,D.M.Meko,andT.W.Swetnam,Eds., United States; and between the Red River and the Rocky Radiocarbon, 155–169. Mountains. Report presented to both Houses of Parliament ——, ——, ——, and ——, 1999: Drought reconstructions for the by command of Her Majesty. continental United States. J. Climate, 12, 1145–1162. Palmer, W. C., 1965: Meteorological drought. Weather Bureau Daniels, L. D., and E. Watson, 2003: Climate–fire–vegetation in- Research Paper 45, U.S. Department of Commerce, 58 pp. teractions in the Cariboo forests: A dendrochronological Phillips, D., 1990: The Climates of Canada. Minister of Supply and analysis. Rep., Forest Innovation and Investment—Forest Services Canada, 176 pp. Research Program, Vancouver, BC, Canada, 56 pp. Richman, M. B., 1986: Rotation of principal components. J. Cli- D’Arrigo, R., R. Villalba, and G. Wiles, 2001: Tree-ring estimates matol., 6, 293–335. of Pacific decadal climate variability. Climate Dyn., 18, 219– Sauchyn, D. J., and A. B. Beaudoin, 1998: Recent environmental 224. change in the southwestern Canadian Plains. Can. Geogr., 42, Enfield, D. B., A. M. Mestas-Nuñez, and P. J. Trimble, 2001: The 337–353. Atlantic multidecadal oscillation and its relation to rainfall ——, and W. R. Skinner, 2001: A proxy record of drought severity and river flows in the continental U.S. Geophys. Res. Lett., 28, Can. Water Resour. J., 2077–2080. for the southwestern Canadian Plains. Filmon, G., cited 2004: Firestorm 2003 Provincial Report. [Avail- 26, 253–272. able online at http://www.2003firestorm.gov.bc.ca/.] Shabbar, A., and W. R. Skinner, 2004: Summer drought patterns Fye, F. K., D. W. Stahle, and E. R. Cook, 2003: Paleoclimatic ana- in Canada and the relationship to global sea surface tempera- logs to twentieth-century moisture regimes across the United tures. J. Climate, 17, 2866–2880. States. Bull. Amer. Meteor. Soc., 84, 901–909. Smith, C. A., and P. Sardeshmukh, 2000: The effect of ENSO on Gedalof, Z., and D. J. Smith, 2001: Interdecadal climate variabil- the intraseasonal variance of surface temperature in winter. ity and regime-scale shifts in Pacific North America. Geo- Int. J. Climatol., 20, 1543–1557. phys. Res. Lett., 28, 1515–1518. Stahle, D. W., E. R. Cook, M. K. Cleaveland, M. D. Therrell, Gershunov, A., and T. P. Barnett, 1998: Interdecadal modulation D. M. Meko, H. D. Grissino-Mayer, E. Watson, and B. H. of ENSO teleconnections. Bull. Amer. Meteor. Soc., 79, 2715– Luckman, 2000: Tree-ring data document 16th century mega- 2725. drought over North America. Eos, Trans. Amer. Geophys. Gray, R., B. Blackwell, K. Iverson, C. Wong, and K. MacKenzie, Union, 81, 121, 125. 2002: Past fire regimes in the interior Douglas fir, dry cool St. George, S., and E. Nielsen, 2003: Palaeoflood records for the subzone, Fraser variant (IDFdk3). Rep., Lignum, Ltd., 207 Red River, Manitoba, Canada, derived form anatomical tree- pp. ring signatures. Holocene, 13, 547–555.

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC 1AUGUST 2005 WATSON AND LUCKMAN 2863

Watson, E., and B. H. Luckman, 2001a: Dendroclimatic recon- struction of maximum summer temperatures from upper struction of precipitation for sites in the Canadian Rockies. treeline sites in interior British Columbia, Canada. Holocene, Holocene, 11, 203–213. 13, 851–861. ——, and ——, 2001b: The development of a moisture-stressed Wolfe, S. A., D. J. Huntley, P. P. David, J. Ollerhead, D. J. tree-ring chronology network for the southern Canadian Cor- Sauchyn, and G. M. MacDonald, 2001: Late 18th century dillera. Tree-Ring Res., 57, 149–168. drought-induced sand dune activity, Great Sand Hills, Sas- ——,and——, 2002: The dendroclimatic signal in Douglas-fir and katchewan. Can. J. Earth Sci., 38, 105–117. ponderosa pine tree-ring chronologies from the southern Ca- Woodhouse, C. A., and J. T. Overpeck, 1998: 2000 years of nadian Cordillera. Can. J. For. Res., 32, 1858–1874. drought variability in the central United States. Bull. Amer. ——,and——, 2004: Tree-ring based reconstructions of precipi- Meteor. Soc., 79, 2693–2712. tation for the southern Canadian Cordillera. Climate Change, ——, J. J. Lukas, and P. M. Brown, 2002: Drought in the western 65, 209–241. Great Plains, 1845–1856: Impacts and implications. Bull. Wilson, R. J. S., and B. H. Luckman, 2003: Dendroclimatic recon- Amer. Meteor. Soc., 83, 1485–1493.

Unauthenticated | Downloaded 09/30/21 05:21 PM UTC