1. Introduction
Drought is a chronic concern in Canada but rarely has it been as serious or extensive as in 2001. According to Phillips (2002), for the western and central Canadian Prairies, “… it was the worst of times. Even in the dust bowl of the 1930s, no single year between Medicine Hat, Kindersley, and Saskatoon was drier than in 2001. Astonishingly, Saskatoon was 30% drier this year than any other over the last 110 years.” As well, Garnett (2002) presented details of the 2000/01 drought from an agricultural and economic loss point of view for the province of Saskatchewan. He pointed out that the economic damage caused by this drought is in the $4–$5 billion range if a multiplier effect is applied.
By definition, a drought is an extended period of abnormally dry weather sufficiently prolonged for the lack of water to cause a serious hydrologic imbalance (i.e., crop damage, water supply shortage, etc.) in the affected area (Glickman 2000). Although a substantial amount of research has been done on this subject, there is still no comprehensive theory that explains the physical mechanism of drought formation and maintenance. However, droughts are considered to be the result of a chain of interacting physical events (see, e.g., Nemanishen 1998).
One possible chain of events for droughts over the Great Plains and the Canadian Prairies is the following. Large-scale processes cause perturbations in sea surface temperature (SST) over the North Pacific Ocean, and this leads to large-scale perturbations in atmospheric pressure (e.g., affecting the strength of the Aleutian low). This in turn affects the large-scale circulation over the North Pacific Ocean and the North American continent, with one consequence being the development of extended periods of dry conditions (i.e., drought) over some regions of the continent. Many previous studies of drought over the U.S. Great Plains and the Canadian Prairies were prepared from this viewpoint and therefore examined teleconnections (Trenberth et al. 1988; Knox and Lawford 1990; Trenberth and Branstator 1992; Bonsal et al. 1993; Trenberth and Guillemot 1996; Chen and Newman 1998; Nemanishen 1998; Garnett 2002). In particular, Garnett (2002) claimed that the 2000/01 drought over Saskatchewan was well forecasted by his Niño-3 SST composite analysis–based teleconnection index. Knox and Lawford (1990), Bonsal et al. (1999), Nkemdirim and Weber (1999), and Chang and Smith (2001) revealed some of the relations between dry periods and the midtroposphere circulation at various time scales (monthly or seasonal). The common conclusion from all of these studies is that the region is covered by an anticyclonic circulation in the mid- or upper troposphere. In particular, based on their 40-yr (1946–85) data analysis, Knox and Lawford (1990) pointed out that the “in situ” blocking over western Canada is almost invariably associated with dry spring and summer months over central Alberta. Nevertheless, none of these studies quantitatively studied moisture transport features.
Other large-scale climatic phenomena have been proposed as being responsible for the initiation of the chain of events that leads to droughts over the Great Plains and Canadian Prairies. For example, past droughts have been studied as a hydroclimatic response to strong El Niño–Southern Oscillation (ENSO) or Pacific–North American (PNA) events (e.g., Trenberth et al. 1988; Knox and Lawford 1990; Trenberth and Branstator 1992; Maybank et al. 1995; Trenberth and Guillemot 1996; Chen and Newman 1998; Nemanishen 1998; Bonsal et al. 1999; Garnett 2002).
Although different mechanisms might be responsible for changes in the large-scale atmospheric conditions (note that the severe 2000/01 drought did not occur within an ENSO episode), precipitation on the regional scale is ultimately governed by three strongly coupled factors: (i) the atmospheric transport of moisture into the region, (ii) the amount of surface evapotranspiration within the region, and (iii) the conversion of the atmospheric moisture into precipitation.
The purpose of this paper is to study the 2000/01 drought from the perspective of these three driving factors for precipitation over the western and central Canadian Prairies with the primary focus being on the atmospheric moisture transport that affects the region. It is of interest to note that atmospheric moisture features have never been the focus of previous drought studies over this region although Trenberth and Guillemot (1996) discussed moisture transport over the United States in a comparison between the 1988 drought and the 1993 flood.
This study employs the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis data (Kalnay et al. 1996; Kistler et al. 2001) to investigate the following moisture transport related issues in association with the 2000/01 severe drought: 1) What is the relative importance of moisture from the Pacific Ocean and the Gulf of Mexico to the precipitation over the western and central prairies? 2) How was the moisture transport from these source regions in 2000/01 different from other years? 3) How did the anomalous moisture transport affect the winter and summer precipitation over the region? 4) How were these moisture transport features linked with large-scale atmospheric circulation characteristics?
The data and methodology used in this study are described in the next section. Surface and atmospheric moisture features associated with the drought will be presented in section 3. Moisture sources and transport features are discussed in section 4. Circulation and other dynamical processes controlling the atmospheric moisture transport and precipitation are given in section 5, and the concluding remarks are presented in section 6.
2. Data and methodology
We will focus on the hydroclimatic conditions over the western and central Canadian Prairies for the 2000/01 agricultural year (hereafter, 2000/01 AY refers to as the period from 1 September 2000 to 31 August 2001). Surface conditions characterizing the development of the drought are examined by analyzing the precipitation anomalies and dugout plots from the Canadian Prairie Farm Rehabilitation Administration (PFRA). Dugouts are man-made ditches that are used on farms for collecting water. In generating the precipitation plots, PFRA used daily precipitation data from the Timely Climate Monitoring Network (TCMN) of Environment Canada. The dugout plots are based on reports (T. Hadwen 2002, personal communication) from the PFRA district offices. (More information concerning the PFRA plots is available online at http://www.agr.gc.ca/pfra/drought/aboutmp_e.htm.) The high-resolution gridded precipitation data from the Meteorological Service of Canada (Mekis and Hogg 1999) are used for the correlation analysis between moisture flux and precipitation. Monthly mean precipitable water from the NCEP–NCAR reanalysis (Kalnay et al. 1996; Kistler et al. 2001) is used to examine the atmospheric moisture content.
Anomalies of these variables were obtained by calculating the difference between the monthly mean values during this specific year and their corresponding 54-yr normal. The Student's t test (Chervin and Schneider 1976) scores were applied to the anomalies to deduce their statistical significance. One of the advantages of using this method is that all areas and times are standardized with respect to each other by consideration of the variance (Karl and Riebsame 1984).
The significance tests will be conducted on two spatial scales: the continental scale and the regional scale. As shown by the thick lines in Fig. 1, the “continental boundaries” are defined along the 122.5°W longitude from 30° to 55°N latitude for zonal moisture transport and along the 30°N latitude from 122.5° to 87.5°W longitude for meridional moisture transport. These boundaries were chosen according to the climatic airflow pattern over the North America. They were designed to account for, as much as possible, the impact of the moisture influxes from the two main moisture sources (the Pacific Ocean and the Gulf of Mexico) on the Canadian Prairies. The “regional boundaries” for the western and central Canadian Prairies are defined along 115°W longitude from 47.5° to 52.5°N latitude for zonal moisture transport and along the 47.5°N latitude from 115° to 100°W longitude for meridional moisture transport. These regional boundaries are chosen according to the NCEP–NCAR reanalysis grid points that are closest to the western and southern boundaries of the western and central Canadian Prairies. The choice of the regional boundaries is also a result of consideration of the following factors. 1) The southern regional boundary was chosen to account for the moisture influxes to the western and central prairies where the drought occurred. 2) The western regional boundary was chosen according to general circulation patterns over this area. Usually, southwesterly airflows over the region prevent moisture north of 52.5°N from being advected into the prairies. Statistical analysis results also support this choice of the western regional boundary in that precipitation over the western and central prairies is best correlated to the moisture influxes across this chosen western regional boundary (47.5°–52.5°N).
Other NCEP–NCAR reanalysis fields such as monthly surface temperature and vertical motion are also examined in studying the large-scale atmospheric conditions characterizing the drought period.
We appreciate that two important factors could affect our results. First, vertical integration by using pressure level data could introduce more errors than by using sigma-level data. However, we do not have access to the NCEP–NCAR reanalysis on sigma levels. Second, mass balance correction might be important for the calculation of moisture fluxes over complex terrain. However, a comparison shows only slight differences over the study area between our result (using pressure-level data and without mass balance correction) and the mass balance–corrected moisture fluxes by using sigma-level data from Trenberth (more information available online at http://www.cgd.ucar.edu/cas/catalog/tn430/ncep/t42f/index.html). The difference in moisture fluxes is within about ±10%. These points imply that the above two issues are not significant over our study area and the data are quite acceptable for this article's objectives.
3. Surface and atmospheric moisture features during the drought
In meteorology, a drought is basically an extended dry period due to below-normal precipitation (Glickman 2000). We will first examine the development of the extreme dry conditions over the western and central Canadian Prairies for the 2000/01 AY by analyzing the temporal and spatial variations of precipitation anomalies, surface snow cover, dugout levels, and precipitable water over the region. The moisture transport anomalies and associated atmospheric circulation features affecting the development of the extended dry period will be examined in later sections.
a. Surface conditions
Figure 2a shows the annual mean of the precipitation percentiles over the Canadian Prairie provinces for the 2000/01 AY. Most parts of southern Alberta and Saskatchewan were in an extremely dry condition during this AY. In particular, several portions of southern and central Alberta suffered record dry conditions. While the focus of this study will be these drought conditions over the western and central prairies, it is also important to note that record wet conditions occurred over some portions of the eastern prairies.
The normalized winter (November 2000/March 2001) precipitation (Fig. 2b) shows a very similar dry–wet condition pattern to that of the annual averaged precipitation percentile. Extremely dry conditions were present over the western and some portions of the central prairies although wet conditions were present over some portions of the eastern prairies. This illustrates that the overall variation of precipitation was actually established early-on and that it persisted over the entire AY. Often, conditions over the Canadian Prairies are an extension of conditions occurring over the U.S. Great Plains. This year is not an exception. Dry conditions were also pronounced south of the Canadian Prairies (W. Skinner 2002, personal communication).
Snow cover observations were quite consistent with these precipitation features. Little snow fell over southern Alberta and Saskatchewan but extensive amounts were observed over southern Manitoba (A. Silis 2002, personal communication). The subsequent snowmelt from this snow cover pattern would have led to reduced soil moisture over the western and central prairies and, at least in the spring, to enhanced soil moisture over the eastern prairies. Feedback on atmospheric moisture from these conditions of soil moisture, through moisture recycling, could have resulted in less precipitation over the western and central prairies and thus the drought.
Water levels in the numerous dugouts over the Canadian Prairies were also quite consistent with the precipitation information. Figures 2c–2e show the distribution of dugout levels for selected periods before and within the 2000/01 AY. It is obvious that before this AY (Fig. 2c), the western and central Canadian Prairies were not dry. Dry conditions started to become evident over southern Alberta in September 2000, the beginning of the 2000/01 AY (Fig. 2d). These conditions continued to spread in area and became more severe throughout the year (Fig 2e).
b. Water vapor in the atmosphere
Figure 3 shows the precipitable water anomalies, revealed from the NCEP–NCAR reanalysis, during the winter (November 2000/March 2001), summer (May–August 2001), and the entire 2000/01 AY. Precipitable water is the vertically integrated atmospheric moisture content within an air column. Figure 3a shows that continent-wide reduced precipitable water occurred during the winter of this AY across southern Canada and the western and eastern United States. However, departure from the average moisture condition was significant (at over the 90% significance level) only over the western and central prairies and some portions of the northwestern and southeastern United States (Fig. 3b). This means that the atmosphere during the winter was much drier than normal over these regions. Even in the summer of 2001 (Fig. 3c), the atmosphere over the western and central prairies continued to be drier than normal although the moisture reduction was slightly less significant (at the 60%–70% significance levels; see Fig. 3d) than that during the winter. The 2000/01 AY annual mean of the precipitable water anomalies showed that the most significant atmospheric moisture reduction was centered over the western and central prairies (Fig. 3e) with a significance level above 60%. This further demonstrated that, in terms of atmospheric moisture content, dry conditions over the western and central prairies were established early-on and preferentially persisted during the entire AY.
c. Summary of drought features
Collectively, the atmospheric moisture content, precipitation, snow cover, and dugout information all illustrated that the 2000/01 AY was an extraordinary one. Dry conditions were established early-on over the western and central prairies, and wet conditions were established early-on over the eastern prairies. Both of these conditions persisted over the entire AY.
4. Moisture sources and transport
The Pacific Ocean and the Gulf of Mexico are the two primary moisture sources for the Canadian Prairies. The relative importance of these two moisture sources to the seasonal precipitation over the western and central Canadian Prairies will be discussed in this section. The temporal and spatial statistical significance of the anomalous moisture transport from these two sources will then be addressed with particular regard to the 2000/01 AY.
a. Relative importance of different moisture sources
To examine their relative contributions to precipitation over the western and central Canadian Prairies, temporal correlation coefficients between monthly precipitation amounts and moisture influxes from these two moisture sources were calculated for each month over the past 54 yr and plotted in Fig. 4. The average moisture influxes were calculated along the “continental moisture boundaries” (Fig. 1). Average precipitation over the western and central prairies was calculated using the 50 km × 50 km gridded data from the Meteorological Service of Canada (Mekis and Hogg 1999). Figure 4 clearly shows that moisture from the Pacific Ocean is more important to the winter precipitation over the western and central prairies. Moisture from the Gulf of Mexico is more important than that from the Pacific Ocean during other seasons, especially the summer when convection is common. Dirmeyer and Brubaker (1999) also found that moisture makes its way northward from the Gulf of Mexico in a stepwise fashion.
The same conclusion was also reached by examining the correlation coefficients between seasonally averaged precipitation and moisture transport. Winter precipitation is much better correlated to zonal moisture transport (with a correlation coefficient of 0.61) than to meridional moisture transport (with a correlation coefficient of 0.14). In contrast, summer precipitation is more strongly correlated to meridional moisture transport (with a correlation coefficient of 0.33) than to zonal moisture transport (with a correlation coefficient of only −0.08). This implies that most of the moisture for precipitation over the western and central Canadian Prairies is from the Pacific Ocean in winter and from the Gulf of Mexico in summer. Of course, good correlations between moisture fluxes and precipitation do not necessarily mean that the physical controls leading to stronger fluxes from the Gulf of Mexico are the same ones that influence summer precipitation over the western and central Canadian Prairies.
It is also worth pointing out that the correlation coefficients presented here are not high. This is not surprising because the development of precipitation is a very complicated meteorological process. It is controlled by many factors in addition to the moisture influx. Results presented here nevertheless illustrate the relative importance of the two moisture sources to precipitation over the western and central Canadian Prairies.
The seasonality of the relative importance of the two primary moisture sources suggests that the likelihood of severe drought over the western and central prairies may be much higher when there is significantly reduced zonal moisture transport into the continent during the winter followed by significantly reduced meridional moisture transport during spring and summer. To verify if this interpretation applies to the 2000/01 AY, the statistical significance of the anomalous moisture transports from these two sources will be examined next.
b. Statistical significance of moisture transport anomalies
Based on our findings concerning the relative importance of the two moisture sources, the discussion will now focus on two specific periods, namely, the winter and the summer. The Student's t test will be applied to deduce the statistical significance of the moisture transport anomalies from the two main moisture sources during the two periods.
Figure 5a shows the t-test scores of the anomalous continental-scale moisture transport from the Pacific Ocean and the Gulf of Mexico in the winter. Figure 5a clearly shows that, among the 54 yr, the 2000/01 AY suffered the most significant less-than-normal moisture transport from the Pacific Ocean during the winter (hereafter, we will use the adjectives “reduced” and “enhanced” to describe the “less than normal” and “more than normal” conditions, respectively). The moisture transport from the Gulf of Mexico during this period was not significantly enhanced and therefore did not compensate for the deficiency in the zonal transport.
It should also be noted that the significance levels for previous dry years (e.g., 1976/77, 1987/88, etc.) are lower than this year. This is consistent with Phillips (2002) who had reported that the 2000/01 drought was the worst over the prairies in the past 110 yr.
Figure 5b shows the t-test scores of moisture transport anomalies during the summer. During this period the moisture transport from the Pacific Ocean was slightly higher than normal whereas the moisture transport from the Gulf of Mexico was less than normal. As shown in the last section, the moisture transport from the Gulf of Mexico has a much larger influence than the moisture from the Pacific Ocean on summer precipitation over the western and central Canadian Prairies. The enhanced moisture transport from the Pacific Ocean did not offset the adverse effect of the reduced meridional moisture transport on summer precipitation over the region.
For regional-scale moisture transport into the western and central Canadian Prairies, the significance of zonal moisture transport anomalies during the winter of the 2000/01 AY (Fig. 5c) is similar to that for the continental zonal moisture transport. In particular, the deficit of zonal moisture transport into the prairies during the winter of this AY is statistically the second most significantly reduced among the 54 yr, just slightly less significant than the 1968/69 winter. The regional-scale meridional moisture transport from the Gulf of Mexico during the winter of 2000/01 was also less than normal. The combination of the zonal and meridional components at the regional scale resulted in the 2000/01 winter experiencing the most significantly reduced net moisture influx from the two primary moisture sources over the period of the NCEP–NCAR data.
During the summer of 2001, both zonal and meridional moisture transports at the regional scale (Fig. 5d) were slightly higher than normal. Apparently, this slightly enhanced moisture transport was too late or not sufficient to compensate for the significant lack of moisture during the winter.
According to the above analysis, the linkage is not always consistent between moisture fluxes at continental and regional scales. This is especially true for meridional moisture transport, which strongly depends on the strength of the Great Plains low-level jet. A more consistent linkage could be found between the meridional transports at the continental and regional boundaries when the Great Plains low-level jet is strong.
c. Spatial variations of moisture transport
To gain insights into the effects of these moisture transport anomalies on the precipitation over this region, the spatial structure of the moisture transport will be examined for the 2000/01 AY.
The averaged zonal moisture transport anomalies during the winter and the summer of this AY are plotted in Fig. 6 along with their t-test scores. It is clearly seen in Fig. 6a that, during the winter, a reduced zonal moisture transport band extended across the middle and southern portions of the North American continent from coast to coast. The north–south extension of this band at the West Coast was from the Gulf of California all the way to 55°N. Furthermore, these reduced anomalies were above the 90% significance level for the winter as shown in Fig. 6b. In other words, the zonal moisture transport from the Pacific to the continent and the western and central Canadian Prairies was significantly less than normal during the winter of the 2000/01 AY.
During the summer of 2001, Fig. 6c shows enhanced zonal moisture transport from the Pacific Ocean. However, the significantly enhanced band extended only from 42.5° to approximately 49°N (Fig. 6d). No significantly enhanced zonal moisture transport was observed over the Canadian Prairies.
The spatial distributions of the meridional moisture transport anomalies and their t-test scores are shown in Fig. 7. During the 2000/01 winter, the meridional moisture transport was significantly reduced over the region at the southwest of, and a small portion of, the Canadian Prairies (Figs. 7a,b). There was enhanced moisture transport from the Gulf of Mexico and the Gulf of California. However, only the transport from the Gulf of California was significantly enhanced. Instead of being transported directly to the north, the moisture from the Gulf of Mexico was directed to the northeast and bypassed the Canadian Prairies (Fig. 7a). During the summer of 2001 (Figs. 7c,d), there was significantly reduced moisture transport from the Gulf of Mexico into the continent. Significantly enhanced meridional moisture transport was observed from the Gulf of California but it did not extend far into the continent.
In summary, the major moisture transport features during the 2000/01 severe drought over the western and central Canadian Prairies are the following.
Zonal moisture transports into the continent and the Canadian Prairies were less than the amount observed in any other winter during the previous 54 yr. Zonal moisture transports were slightly above average into both the continent and the prairies during the summer.
Meridional moisture transport into the continent was near normal during the winter and reduced during the summer. Transport into the western and central prairies was reduced during the winter and slightly enhanced during the summer.
But, how were these moisture transport features generated and what are the dynamical explanations for these moisture transport features during this specific AY? These questions will be addressed in the next section.
5. Related circulation features
In this section, large-scale circulation features will be related to the moisture transport characteristics discussed earlier. Circulation-related vertical motion, moisture convergence, and surface temperature are also analyzed.
a. Low-level circulation patterns
Previous studies usually analyzed the 500- and 300-hPa circulation patterns primarily for teleconnection purposes (e.g., Trenberth et al. 1988; Knox and Lawford 1990; Trenberth and Branstator 1992; Trenberth and Guillemot 1996; Chen and Newman 1998; Maybank et al. 1995). However, the focus of this study is regional moisture transport and associated dynamical circulation features. The geopotential height anomalies averaged from 1000 to 700 hPa (referred to as “low level” hereafter) will be used for this analysis because atmospheric moisture transport occurs mainly at lower levels.
Figure 8 shows some large-scale circulation-related features during the 2000/01 winter. These include (a) the low-level geopotential height anomalies, (b) the moisture convergence anomalies, (c) the 500-hPa vertical motion anomalies, and (d) the surface temperature anomalies. In this section, the discussion will be focused on the low-level circulation patterns.
During the winter of this AY, the low-level averaged geopotential height anomalies (Fig. 8a) show an anomalous high over western North America. There is a good agreement between the strong perturbation easterly flow along the southern edge of the anomalous high and the significantly reduced zonal moisture transport (recall Fig. 6a). This circulation pattern limited the flow of moisture from the Pacific Ocean into the continent and the western and central Canadian Prairies and resulted in the significantly reduced zonal moisture transport from the Pacific Ocean during this winter. The wide extension of the anomalous high in the east–west direction is also not favorable for the western and central Canadian Prairies to benefit from the moisture from the Gulf of Mexico and the Gulf of California.
During the summer of 2001, the low-level circulation pattern presented an anomalous high over the southwestern United States and an anomalous low over northwestern Canada (Fig. 9a). The strong perturbation of the westerly flow along the northern edge of this anomalous high was responsible for the significantly enhanced zonal moisture transport band just south of the Canadian Prairies (Fig. 6c). The associated, enhanced northerly flow over the Great Plains inhibited the moisture from the Gulf of Mexico from being transported into the western and central Canadian Prairies (see Figs. 7c,d).
It should be noted that the anomalous high during the winter of 2000/01 AY (maximum of 35 m) was much stronger than that during the summer (maximum of 12 m). In fact, the annually averaged low-level geopotential height anomalies for the entire AY (not shown) presented the same pattern as that during the winter. This implies that the low-level circulation pattern shown in Fig. 8a was the dominant pattern during this severe drought that was responsible for the significantly reduced zonal moisture transport from the Pacific Ocean during the winter.
So far, we have shown that the significantly reduced zonal moisture transport to the western and central Canadian Prairies during the 2000/01 AY was associated with an anomalous high pressure system preferentially occurring during the winter. But, why was this year extraordinary? The location and strength of the anomalous high pressure are two key factors to consider for its blocking effects. To address this issue, the interannual variability of area-averaged winter low-level geopotential height over the “blocking area” of the western and central prairies is plotted in Fig. 10. This blocking area is bounded by the 45° and 57.5°N and the 125° and 102.5°W. Based on our preliminary analysis, this area was chosen according to its associated circulation pattern that typically leads to blocking effects on moisture transport into the western and central prairies. Figure 10 shows that the high pressure over the blocking area during the 2000/01 winter was the strongest of any year and thus its blocking effects for the western and central prairies were the strongest. It was this strongest high pressure system that resulted in the record-low zonal moisture transport during the 2000/01 winter.
An alternative explanation can be found in the splitting of the upper-level jet stream during this specific AY. The jet stream location indicates the long-wave pattern and is usually a good indicator of moisture transport strength. The strength and angle of the incoming flow at the West Coast are directly related to the moisture influx from the Pacific Ocean. An analysis of the 54-yr normal of the wind vectors and their magnitude show that the normal jet stream at 300 hPa passes directly over the prairies from July to October (not shown). Even though the jet stream does not pass over the prairies in other months, strong airflow usually does.
The jet stream structure was unusual during the winter of the 2000/01 AY. The 300-hPa jet stream split into two branches and bypassed the prairies in October and November of 2000 and January and March of 2001 (not shown). The splitting of the jet stream resulted in a reduced zonal component of the incoming flow at the West Coast. Consequently, this factor reduced the zonal moisture transport during the winter of this drought year.
b. Vertical motion, moisture convergence, and surface temperature
Vertical motion is very important for the conversion of atmospheric moisture into clouds and precipitation. Updrafts and moisture convergence associated with low pressure systems are often triggers for precipitation. In contrast, sinking motions in high pressure systems usually result in moisture divergence as well as hot and dry surface air in summer.
The vertical motion anomalies at 500 hPa during the winter and summer of this AY are shown in Figs. 8c and 9c, respectively. During the winter, consistent with the low-level circulation pattern (Fig. 8a), there was an extensive area with enhanced sinking over the western North American continent (Fig. 8c). The entire Canadian Prairies was within this enhanced sinking area. This persistent enhanced sinking resulted in reduced moisture convergence (Fig. 8b) over the southwestern and south-central prairies. The persistent anomalous high pressure system was associated with the slightly lower-than-normal surface temperature during the winter over the western and central prairies (Fig. 8d).
During the summer of 2001, the vertical motion presented a dipole pattern over the Canadian Prairies (Fig. 9c); that is, there was enhanced sinking over the western prairies but enhanced rising over the eastern prairies. This dipole structure in vertical motion led to the southwestern portion of the prairies continuing to experience reduced moisture convergence (Fig. 9b) while enhanced moisture convergence occurred over the eastern prairies, which was associated with this area's enhanced precipitation. Over the western and central Canadian Prairies, the persistent anomalous high pressure along with the dry surface conditions produced the warm surface temperature anomalies in Fig. 9d. The dynamical reasons for the enhanced moisture convergence and precipitation over the eastern prairies were weak local troughs found in the monthly geopotential height anomalies over the region (not shown).
Our statistical analysis illustrated that the above features were also extraordinary over the western and central prairies for this AY. Observed from the 54-yr period of reanalysis, the descending vertical velocities were the strongest over this area during the winter (especially from January to March), the summer (especially May, July, and August), and the entire 2000/01 AY. The moisture convergence over the southwestern and south-central prairies was the lowest on record during the winter, and the summer surface temperature was in the top 10%.
6. Concluding remarks
A severe and persistent drought occurred over the western and central Canadian Prairies during the 2000/01 agricultural year. The focus of this study was on the water vapor and atmospheric conditions associated with this major drought. Use was made of the 54-yr NCEP–NCAR reanalysis and other surface observational information. The study has led to several observations and conclusions as outlined below.
The precipitation over the western and central Canadian Prairies is correlated with moisture influxes from the two primary moisture sources (the Pacific Ocean and the Gulf of Mexico). During winter, moisture transport from the Pacific Ocean is more important to precipitation production over this region than that from the Gulf of Mexico. In other seasons, especially in summer, moisture from the Gulf of Mexico is more important than that from the Pacific Ocean.
Moisture transport was examined for the winter and summer periods of the 2000/01 AY. The analysis revealed that the western and central Canadian Prairies suffered the most significantly reduced zonal moisture transport from the Pacific Ocean during the winter of this AY. As well, there was no significantly enhanced meridional moisture influx from the Gulf of Mexico to compensate for the lack of zonal transport. During the summer of 2001, the meridional moisture transport from the Gulf of Mexico was reduced.
The significantly reduced moisture transport to this region largely arose due to the blocking effect of an anomalous high pressure over western North America. This “high” blocked the moisture from entering the western and central Canadian Prairies and forced the moisture to bypass the area. Some of the diverted moisture flux could have converged over the eastern prairies and produced extremely wet conditions there.
Consistent with the large-scale circulation patterns, a split jet stream structure was observed during the winter and enhanced sinking occurred over the western and central prairies during the entire 2000/01 AY. The splitting of the 300-hPa jet stream weakened the zonal components of the incoming airflow and significantly reduced the moisture influx from the Pacific Ocean during the winter of this AY. During the summer of 2001, the anomalous high pressure system was linked with warmer surface temperatures that worsened the drought.
In summary, this study has shown that water vapor transport into, integrated water vapor above, and precipitation over the western and central Canadian Prairies were all extraordinarily low during the 2000/01 agricultural year. These features occurred in part due to extraordinarily strong high pressures that began in the winter and persisted over much of the year. These high pressures blocked the zonal moisture influx into the western and central prairies during the winter, they were associated with extraordinarily strong subsidence during the entire AY, and they led to extremely warm summer conditions over the region. Collectively, these exceptional, interrelated features contributed to the initiation and maintenance of the 2000/01 severe drought over the western and central Canadian Prairies. This study has been focused on, from a natural variability point of view, analyses of some of the physical processes producing this major drought. More work needs to be carried out to verify if this severe drought may be evidence that the global water cycle is accelerating as a result of human-induced climate change.
Acknowledgments
The authors would like to thank the Canadian Prairie Farm Rehabilitation Administration (PFRA) for allowing us to use the precipitation and dugout plots. We also want to thank the anonymous reviewers for their constructive comments on improving the manuscript. This study is supported by Environment Canada, the Natural Sciences and Engineering Research Council of Canada, and the Panel on Energy Research and Development (PERD).
REFERENCES
Bonsal, B. R., A. K. Chakravarti, and R. G. Lawford, 1993: Teleconnections between North Pacific SST anomalies and growing season extended dry spells on the Canadian prairies. Int. J. Climatol., 13 , 865–878.
Bonsal, B. R., X. Zhang, and W. D. Hogg, 1999: Canadian Prairie growing season precipitation variability and associated atmospheric circulation. Climate Res., 11 , 191–208.
Chang, F-C., and E. A. Smith, 2001: Hydrological and dynamical characteristics of summertime droughts over the Great Plains. J. Climate, 14 , 2296–2316.
Chen, P., and M. Newman, 1998: Rossby wave propagation and the rapid development of upper-level anomalous anticyclones during the 1988 U.S. drought. J. Climate, 11 , 2491–2504.
Chervin, R. M., and S. H. Schneider, 1976: On determining the statistical significance of climate experiments with general circulation models. J. Atmos. Sci., 33 , 405–412.
Dirmeyer, P. A., and K. L. Brubaker, 1999: Contrasting evaporative moisture sources during the drought of 1988 and the flood of 1993. J. Geophys. Res., 104 (D16) 19383–19397.
Garnett, R., 2002: The Canadian Prairie drought of 2001: A four billion dollar shortfall. CMOS Bull., 30 , 37–39.
Glickman, T. S., Ed.,. . 2000: Glossary of Meteorology. 2d ed. Amer. Meteor. Soc., 855 pp.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437–478.
Karl, T. R., and W. E. Riebsame, 1984: The identification of 10- to 20-year temperature and precipitation fluctuations in the contiguous United States. J. Climate Appl. Meteor., 23 , 950–966.
Kistler, R., and Coauthors, 2001: The NCEP/NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82 , 247–267.
Knox, J. L., and R. G. Lawford, 1990: The relationship between Canadian Prairie dry and wet months and circulation anomalies in the mid-troposphere. Atmos.–Ocean, 28 , 189–215.
Maybank, J., B. Bonsal, K. Jones, R. Lawford, E. G. O'Brien, E. A. Ripley, and E. Wheaton, 1995: Drought as a disaster. Atmos.–Ocean, 33 , 195–222.
Mekis, E., and W. D. Hogg, 1999: Rehabilitation and analysis of Canadian daily precipitation time series. Atmos.–Ocean, 37 , 53–85.
Nemanishen, W., 1998: Drought in the Palliser Triangle. Prairie Farm Rehabilitation Administration Drought Committee, PFRA, Calgary, Alberta, Canada, 62 pp. [Available online at http://www.agr.gc.ca/pfra/pub/drprimer.pdf.].
Nkemdirim, L., and L. Weber, 1999: Comparison between the droughts of the 1930s and the 1980s in the southern prairies of Canada. J. Climate, 12 , 2434–2450.
Phillips, D. W., 2002: The top ten Canadian weather stories for 2001. CMOS Bull., 30 , 19–23.
Trenberth, K. E., and G. W. Branstator, 1992: Issues in establishing causes of the 1988 drought over North America. J. Climate, 5 , 159–172.
Trenberth, K. E., and C. J. Guillemot, 1996: Physical processes involved in the 1988 drought and 1993 flood in North America. J. Climate, 9 , 1288–1298.
Trenberth, K. E., G. W. Branstator, and P. A. Arkin, 1988: Origins of the 1988 North American drought. Science, 242 , 1640–1645.
Geographical location of the study area. The thin lines outline the location of the Canadian Prairies. The tree line is shown across the Canadian Prairies. The continental boundaries and regional boundaries used in this study are also shown with the thick lines
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Surface conditions over the Canadian Prairie provinces during the 2000/01 agricultural year. (a), (b) The annual total precipitation percentiles and percent of average precipitation for the winter, respectively. (c), (d), (e) The dugout levels, respectively, before, near the beginning of, and near the end of the agricultural year
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
(Continued)
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Precipitable water anomalies (mm) and corresponding t-test scores for the (a), (b) 2000/01 winter; (c), (d) summer of 2001; and (e), (f) the entire AY. Shaded areas are those with significance levels over (b) 90%, (d) 70%, and (f) 60%. The location of the Canadian Prairies is also shown in these figures
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Correlation coefficients between the monthly precipitation over the western and central Canadian Prairies and the moisture transport (zonal and meridional) across the continental boundaries. The sample length for each month is 54 yr (1948–2001). The moisture fluxes were calculated from the NCEP reanalysis data and the precipitation data are from the Meteorological Service of Canada (Mekis and Hogg 1999)
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Time series of the t-test scores of the zonal (solid line) and meridional (dashed line) moisture transport anomalies across the continental boundaries for (a) winter (Nov–Mar) and (b) summer (May–Aug); and across the regional boundaries for (c) winter and (d) summer. The arrow in each panel identifies the 2000/01 AY
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Averaged zonal moisture transport anomalies (kg m−1 s−1) during (a) the winter and (c) the summer of the 2000/01 AY; and (b), (d) the corresponding t-test scores. The shaded areas in (b) and (d) are those with over 90% significance level. The location of the Canadian Prairies is also shown in these figures
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Same as in Fig. 6, but for the meridional moisture transport (kg m−1 s−1)
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Circulation features during the winter of the 2000/01 AY. (a) Low-level geopotential height anomalies (m), (b) vertically integrated moisture convergence anomalies (mm day−1), (c) vertical motion anomalies (Pa s−1) on 500 hPa, and (d) surface temperature anomalies (K). The location of the Canadian Prairies is also shown in these figures
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Same as in Fig. 8, but for the summer of 2001
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
Time series of the winter averaged low-level geopotential height (m) over the blocking area for the western and central Canadian Prairies
Citation: Journal of Climate 17, 2; 10.1175/1520-0442(2004)017<0305:MTAOHF>2.0.CO;2
