1.  Introduction

a.  Teleconnection Patterns

For over a century an integral part of climatological investigation has been aimed at understanding the spatial and temporal patterns of climate. Large-scale oscillatory climate patterns, commonly referred to as ‘teleconnection patterns', are modes of variability that greatly influence changes in many other earth systems including weather and ocean processes. Teleconnections are frequently associated with anomalous thermal and moisture conditions over extensive areas that can drastically change with pattern regime shifts. As a result, the biological, climatological, and societal impacts of an event can be highly variable, and at times, quite significant.

Hildebrandsson (1897) and Lockyer (1906) were perhaps the first to detect teleconnection signatures while investigating large-scale pressure variability between remote areas of the world. Walker (1923, 1924) followed with the documentation of three oscillations that were described as inherently important to include in forecasting the world's weather. However, it wasn't until Angström (1935) that the term, teleconnection, was officially coined. At the time, the designation broadly referred to any climatic fluctuation with perceivable patterns. Bjerknes (1966, 1969) provided evidence for a refined definition of atmosphere-ocean interactions where "pulses" of surface forcings, such as sea surface temperature variability, trigger large-scale atmospheric responses.

Today, there is general consensus among the scientific community that teleconnection patterns are low frequency modes of climatic oscillation, sometimes referred to as "preferred states of flow" (NOAA-Cooperative Institute for Research in Environmental Sciences' Climate Diagnostics Center [NOAA-CIRES CDC] 2005), with observable large-scale patterns of anomalous atmospheric pressure and circulation variability (NOAA/ National Weather Service's Climate Prediction Center [NOAA/NWS CPC] 2005). Teleconnections describe statistical relationships of atmospheric and/or oceanic parameters obtained between at least two remote locations (representative of ideal centers of maximum variability that are typically based on observations at the surface though sometimes at a prescribed height) that link the climates of distant places and frequently the areas between (American Meteorological Society 2005; Livezey and Smith 1999). Therefore, the knowledge of conditions at one activity center can be used to understand and often predict the climate of other areas. Teleconnections are considered propagating systems of field variable correlations that occur in opposing regime states (usually referred to as positive and negative phases) with periods of neutral activity, and recur on intra-annual, inter-annual, and inter-decadal time scales throughout the historical record. Spatially, teleconnections occur on regional to planetary scales (Barry and Carleton 2001). The mechanisms that control these oscillatory patterns are generally known to stem from interactions of the atmosphere and ocean, though exact triggers remain a controversial topic of scientific debate.  

Knowledge of teleconnections, their impacts, and the ability to predict individual events have considerably improved over the past few decades. Skillful forecasts are generally available one to two seasons in advance for the El Niño/ Southern Oscillation (ENSO), the most prominent teleconnection pattern impacting the tropics (Collins et al. 2002). The secondary, extra-tropical influences of ENSO are global and at times more influential to areas than more localized patterns (Halpert and Ropelewski 1992; Mo and Livezey 1986). Consequently, it has been the most widely studied teleconnection and in recent years overwhelming media and government interest in ENSO has developed. This is, undoubtedly, related to the great impact of ENSO on not only ecosystems but also human infrastructure. In the strong La Niña (sometimes called the cool phase) ENSO winter of 1998/99 the Washington state Mount Baker ski industry thrived as snowfall totals reached 28.96 meters (1142 inches), setting a world record (Mantua 2000). Alternatively, the global economic loss attributed to the 1982/83 El Niño (the warm state of ENSO) was over $8 billion from flooding in Bolivia, Ecuador, and Peru, hurricanes in Tahiti and Hawaii, and droughts/ fires in Africa, India, Indonesia, and Australia (University Corporation for Atmospheric Research [UCAR]/NOAA 1994).

In recent years other notable teleconnections have also become better understood globally, specifically extra-tropical patterns. Approximately 13 primary patterns are now credited with having impacts on northern mid latitude climate. This includes the North Atlantic Oscillation (NAO), that is of great importance to the climate of eastern North America, Greenland, and Europe along with ocean processes north of 60ºN (Barnston and Livezey 1987; Hurrell 1995; Lamb and Peppler 1987; van Loon and Rogers 1978). The ability to forecast many of these teleconnections is improving. However, most (like the NAO) exhibit more intra-annual variability than ENSO and are only skillfully predicted for approximately 15 days out ( NOAA/NWS CPC 2006a; USDEC, NOAA, NWS, NCEP 2006).

b.  Knowledge and Present Challenges

There is an abundant amount of scientific literature explaining the seasonal characteristics and impacts of teleconnection patterns. As a result, it has been well established that ENSO and the NAO have strong winter signatures in North America. In the United States, typical El Niño winters correspond to a system of enhanced meridional atmospheric flow and generally warmer than average temperatures across most of the northern half of the country while southeastern regions are anomalously cool. These seasons are also associated with drier conditions in the midwest and northeast, with amplified storm tracks and wet weather across the south. In a typical La Niña winter the opposite conditions can be expected across the country with pronounced zonal circulation patterns (Barnston et al. 1999; NOAA/NWS CPC 2004). Positive regime states of the NAO generally produce intensified Atlantic westerlies and storm tracks across eastern Canada to northern Europe, resulting in eastern United States winters that are anomalously warm with fewer than average snowfall events. These systems have a more variable direction in negative phase NAO years, often producing colder, snowier winters across the eastern United States (Hurrell 1996; Kapala et al. 1998; Rogers 1997; Visbeck 2005). For a better understanding, Figure 1 displays the correlations of January-March surface air temperature and precipitation rate in the United States to indices for ENSO and the NAO teleconnection patterns. These maps were generated through the NOAA-CIRES CDC (NOAA-CIRES CDC 2006a).

While it is beneficial to have a general understanding of the seasonal conditions associated with a particular teleconnection event, like those described above, to improve present forecasting skill a more detailed comprehension of seasonal climate variability is needed. Current knowledge of teleconnection phase related climate variability is chiefly based on statistical averages. Most temperature and precipitation forecasts for a three-month season are presented to the public as 1) averages for that season and 2) as departures from already established average seasonal temperature and precipitation conditions, usually with some percent chance of occurrence. Predicting the climate variability within a season that is connected to teleconnections is difficult because few of these relationships are known (Gershunov et al. 2000; Higgins et al. 2000; Smith and Sardeshmukh 2000). However, this information is crucially important for improving seasonal outlooks. We may be able to anticipate, for example, that winter temperatures in Washington State will be anomalously warm during strong El Niño regimes, but temperatures during these periods may also be highly variable month to month or even week to week. Seasonal averages in these instances would provide very little to an accurate winter forecast.

Another challenge for improving seasonal outlooks requires additional insight into neutral teleconnection phases and their subsequent regional impacts. Considerably less research has been devoted to these periods despite nearly all teleconnections over the last 50 years having a greater frequency of neutral regimes than other phases. At the moment neutral regimes, if even included in seasonal forecasts, are the most difficult to predict and are often described as times of near normal conditions. It is very likely that neutral regimes are associated with inter- and intra-seasonal climate variability that diverges from normal conditions but to know for sure this should be further investigated. Fortunately, the literature on ENSO neutral periods is growing, though for other teleconnections like the NAO studies devoted to neutral regimes are scarce.

A final challenge for teleconnection pattern analysis is to identify the impacts of multiple teleconnection phase interactions on climate and anomalous seasonal variability. As an individual regime of one pattern may dominate at the same time as that of another pattern (i.e., La Niña and positive NAO or El Niño and neutral NAO), the combined influences on regional climate could significantly vary from the influences of each phase type in isolation. Anomalous weather conditions could be intensified or reduced depending on the collective phases. As this topic has not been fully explored, developing an understanding of combination teleconnection phase impacts is a worthwhile goal for better predictions.

c. Goals and Significance of Proposed Research

The overall goal of this dissertation is to provide a more detailed explanation of teleconnection pattern impacts on seasonal weather than is currently available. Specifically, this research will be an assessment of intra-seasonal climate variability associated with dominant teleconnection pattern activity in the United States. Though the inclusion of all teleconnection modes is beyond the scope of this investigation, two very prominent teleconnections, ENSO and the NAO, will be thoroughly evaluated. All individual (positive, negative, and neutral) and combination phases of these patterns over the past 50 years will be assessed for influences to the intra-seasonal variability of thermal and moisture characteristics across the country. To do so, monthly maximum and minimum temperatures and total monthly precipitation variables will be statistically analyzed during winter at 110 spatially consistent United States locations. The winter season is chosen because, in the United States, both the ENSO and NAO signatures are most pronounced at this time of year allowing for the best separation of true teleconnection signals from other natural variability.

This dissertation should provide information to the climate community that improves the understandings of not only ENSO and the NAO, but also how each impacts North America. It is hypothesized that the temperature and precipitation dynamics of individual teleconnection regimes are not consistent across a three-month interval, and may also be highly variable across the country. It is also hypothesized that these relationships can be properly evaluated and become reasonably understood within this assessment. A benefit of studying temperature and precipitation characteristics is that it allows for additional insight into the larger-scale systems that produce those conditions (i.e., how storm tracks and related wind and pressure centers may be shifted within a season). This work may also provide a foundation for further investigation in other seasons and with other teleconnections. Intuitively, this would be worthwhile since such problems have yet to be addressed.

This research offers new insight and a uniquely comprehensive approach to climate and teleconnection pattern investigation. Intra-seasonal variability (versus just average temperature and precipitation conditions) that is associated with individual teleconnection events has never been as thoroughly assessed as this examination proposes. Additionally, the assessment of variability during times of combination teleconnection phases and neutral phases has only been briefly and provisionally studied in the scientific literature. For the first time, this topic will be carefully examined in this research. A key contribution of this work may be improved regionalized seasonal forecasts that better prepare many for approaching winter seasons. Advance warning of where and when the most severe winter weather will likely be experienced allows for preparations of roadways and transportation networks threatened by icy conditions, high wind speeds, and significant snowfall totals. Alternatively, better prediction of abundant snows may improve seasonal net gains and prove beneficial to mountain towns dependent on skiing and winter tourism.

2. Background

a. El Niño/ Southern Oscillation

The ENSO teleconnection describes a propagating system of thermally-driven mass circulation with east-west displacements in Walker cells over the tropical Pacific. In the La Niña (cool) state of ENSO, an enhanced Walker circulation corresponds to pronounced low-pressure centers, high sea surface temperatures, and convection with strengthened trade winds in the western Pacific. At the same time, high pressure, low sea surface temperatures, dry conditions, and a steep oceanic thermocline occur east of the dateline. The pressure, circulation, temperature, and precipitation patterns reverse when the Walker circulation weakens during an El Niño (warm) regime and the trades subside, relaxing the thermocline and increasing temperatures off the coast of South America (Barry and Carleton 2001; Bjerknes 1966, 1969; Doberitz 1968; Julian and Chervin 1978; Kidson 1975; Troup 1965; Walker 1923, 1924; Walker and Bliss 1932).

The ENSO signal is the strongest in the system of inter-annual atmosphere-ocean climate variability (Wang et al. 1999). La Niña and El Niño events appear to be "phase locked" with the annual cycle and individually persist for approximately one year, though it is not uncommon for a phase to last for several consecutive years (Wallace et al. 1998; UCAR 2005). Throughout this time the strength of a single event can vary. The ENSO literature frequently designates an ENSO episode as either a strong, moderate, or weak event.

Both phases recur on a period of approximately three to seven years, typically with intervals of neutral ENSO phase dominance between. Individual regimes tend to develop between the months of April and June, peak in late autumn or early winter, and decay during the spring (Smith et al. 1999). The distance from the tropical Pacific can indicate the amount of time lag an area has before experiencing any impacts. For example, in the United States maximum ENSO effects are often recorded in the winter months that follow onset (UCAR/NOAA 1994). For additional information on the anatomy of ENSO phase development, persistence, and decay, see Rasmusson and Carpenter (1982) and Harrison and Larkin (1996, 1998).

Without a single, known causal mechanism for the ENSO phenomenon many ENSO indices have been made available to trace the strength and persistence of the teleconnection. Regardless of the atmospheric and oceanic variables selected for each index, the temporal regime trends are often similar. Sir Gilbert Walker's Southern Oscillation Index, along with other principle indices, relies on sea-level pressure anomalies at locations in the Pacific to define ENSO modes (Walker 1923, 1924). This is usually expressed as the anomaly at Tahiti (17.5ºS, 149.6ºW) minus that at Darwin (12.4ºS, 130.9ºE). However, some like Wright et al. (1988) and van Loon and Madden (1981) include different locations such as Cocos Island (12ºS, 97ºE). Other indices are based on precipitation rates, air temperature, zonal and meridional surface wind components, and total fractional cloudiness from various sites in the Pacific (Wright 1984, 1985; NOAA-CIRES CDC 2005). Currently, sea surface temperature anomalies tend to be favored in the literature as the single best variable to use in formulating an ENSO index (Hanley et al. 2002). These anomalies are most often acquired for the Niño 3.4 region of the Pacific (5°N-5°S, 120-170°W) (NOAA/NWS CPC 2006b).          

b. ENSO Impacts on United States Winters

For many regions across the United States the impacts of ENSO on winter climate have been well defined. During an El Niño regime, the western half of the country is divided with nearly opposite thermal and moisture anomalies (Arkin 1982; Federov and Philander 2000; Horel and Wallace 1981; Wallace and Gutzler 1981; Webster 1981). Several consecutive months of a warmer eastern equatorial Pacific triggers a system of atmosphere-ocean feedbacks that strengthens the Aleutian Low in the north Pacific. This essentially causes a split of the jet stream into two main branches, a polar jet directed toward Alaska and a pacific jet displaced south of its normal location over the central United States (NOAA/NWS CPC 2004). The focus of storm tracks that can be considerably more severe in a strong event, are oriented across the southwestern tier of the country bringing cooler, wetter weather to these areas (Douglas and Englehart 1981, 1984; Norton et al. 1985; Schonher and Nicholson 1989) . As high pressure ridges form beneath the polar jet, the northwestern United States experiences anomalously high winter temperatures and altogether drier conditions (Mantua 2000; Miles et al. 2001).

The central portions of the country generally experience the same divide (Livezey et al. 1997). However, the areas situated along the north-south border of this divide, from approximately San Francisco, California to Cheyenne, Wyoming, are considerably more difficult to forecast in ENSO years (Schonher and Nicholson 1989). Weather patterns for the southeastern United States mimic the southwest, displaying anomalously cool, wet winters during an El Niño. These conditions can be even more pronounced than in the southwest as large troughs develop directly over the region. Some evidence indicates the northeastern tier can expect higher than average temperatures with below average precipitation during El Niño winters (Halpert and Roplewski 1992; Kiladis and Diaz 1989). Yet, the conditions of this area are historically more ambiguous and sporadic.

A La Niña regime brings the opposite winter weather patterns just described to the United States (Trenberth 1997). An important difference between ENSO regimes is that the climate signal for the cool phase is better documented in the literature. In other words, the trends are more reliable with less exception during a La Niña than an El Niño, and subsequently, result in generally better forecasts for these periods (Western Regional Climate Center 2006).

c. North Atlantic Oscillation

The NAO teleconnection describes the degree of negative correlation in sea level pressures, and those at heights to approximately 500-mb, of the subtropical Azores high and subpolar Icelandic low (Wallace and Gutzler 1981; Rogers 1984). In a positive phase of the NAO a strengthening of both the Azores high and Icelandic low is evident. In this state, the highest latitudes of the North Atlantic are associated with anomalously low pressures while pressures of the central Atlantic are anomalously high. In a negative phase the Azores high and Icelandic low weaken, allowing for the opposite pressure anomalies at these locations (Lamb and Peppler 1987).

Barnston and Livezey (1987) identify the NAO as a yearly leading mode of low frequency circulation variability throughout the northern hemisphere, based on assessments of monthly 700-mb height anomalies. The NAO is present in all months and is a dominant summer pattern, however, the NAO signature is most pronounced in winter months and weakest in autumn (Barnston and Livezey 1987; Rogers 1990). Mächel et al. (1998) detected the sea level pressure gradient between the Azores high and Icelandic low to be 9-mb/ 1000 km in winter versus a gradient of 5.5-mb/ 1000 km in autumn.

Several time series of NAO indices are defined in the scientific literature. Indices frequently differ in where the pressure anomaly ‘centers' of the NAO pattern are located. This is mainly due to evidence suggesting NAO pressure centers can shift, especially over decades, and are not static through time (Jones et al. 1997).  Therefore, selecting an NAO index for study should be dependent on the spatial and temporal parameters of the investigation.

An NAO index that is defined with commonly referenced pressure centers, developed by Rogers (1984), accounts for the difference between standardized mean surface pressure anomalies at Ponta Delgada, Azores and Akuyeri, Iceland. Other indices declare a northern anomaly center at various sites in Iceland, such as Stykkisholmur or Reykjavik, and use southern centers at Lisbon, Portugal and Gibraltar, Spain (Hurrell 1995). Currently, the NOAA/NWS CPC utilizes an NAO index that can account for the inherent spatial variability and seasonality of the pattern. The index is calculated using a Rotated Principal Component Analysis procedure applied to standardized monthly 500-mb anomalies that is restructured from a technique used in Barnston and Livezey (1987) (NOAA/NWS CPC 2006c).

Phases of the NAO are highly variable in duration, ranging from inter-seasonal to inter-decadal oscillations (Nakamura 1996). This can lead to extreme difficulty in forecasting the NAO. However, an observable trend in the record of winter season NAO activity generally displays negative phase characteristics from 1950 to approximately 1979. Since 1980, positive NAO phases have dominated the winter record though there is some indication in recent years that negative phase tendencies are again returning (Halpert and Bell 1997; Hurrell 1995).

d. NAO Impacts on United States Winters

Across the northern hemisphere, oscillatory patterns of temperature, relative humidity, precipitation, and wind can be observed alongside the seesaw-like pressure patterns of the NAO (Cayan 1992; van Loon and Rogers 1978). The variability is related to displacements in the North Atlantic jet stream, coupled with subsequent changes in the Gulf Stream, and the associated movement and intensity of storm tracks around pressure centers (in figure-8 style rotations) that strengthen and weaken with the NAO. For the contiguous United States a positive NAO regime tends to bring warmer winter weather and generally drier conditions to most of the country, especially in the north and east, with a northerly shift in the jet stream and more robust circulation around the Azores high and Icelandic low (Deser and Blackmon 1993). With positive phases being a dominant trend for the past few decades, the eastern United States has experienced generally mild winters compared to those of the 50s, 60s, and 70s. Areas in a band stretching from the eastern United States/ southeastern Canada to northern Europe are also likely to experience above average temperatures though winter storms increase eastward as they are directed away from North America toward Iceland (In a strongly positive event, or one of longer duration, these anomalies can extend into the northern and central reaches of Siberia.) (Hurrell et al. 2001).

Winter weather in the United States tends to be more severe and sporadic in a negative NAO state. This is especially noticeable east of the Mississippi River (Hurrell et al. 2001). Along with lower than average temperatures, more frequent and intense winter storms are common for these regimes as the jet stream is oriented south of its usual position (Hurrell and van Loon 1997). This happens as the driving pressure centers weaken and storm tracks that are directed around these regions in a positive phase are now driven toward them. Arctic air from Canada drifts south to the United States resulting in more cold air outbreaks and intensified wind patterns. Nor'easters are also more common during a negative NAO phase. The 2002/03 negative NAO winter brought record setting snowstorms up and down the east coast, especially in the Mid Atlantic states (NASA Earth Observatory 2003).

3.  Study Region and Data

a.  Meteorological Variables and Stations

Weather data for this examination are acquired from the NOAA National Climatic Data Center's (NOAA NCDC) United States Historical Climatology Network (USHCN) for stations within a study region of the contiguous United States (NOAA NCDC 2006a). The USHCN is comprised of stations from the U.S. Cooperative Observing Network with data that are tested for inhomogeneities and adjusted for non-climatic influences on the data (NOAA NCDC 2006b). The meteorological variables selected for intra-seasonal climate analysis are monthly (January to March) maximum temperature and minimum temperature (in hundredths of degrees Fahrenheit) and total precipitation (in hundredths of inches) from 1950-2002.

For a homogeneous representation across the country 110 of the 1221 available USHCN stations are used, as sites within the network have a notably denser distribution in the eastern United States (Figure 2). A stratified, systematic, unaligned pattern is followed as closely as possible, given initial station distribution and lower quality station eliminations, to select sites for this research. This approach is beneficial because it allows a network to be defined by stations that are uniformly distributed and evenly spaced yet not rigidly aligned (Davis 2002). A review of the station metadata files indicates the 110 stations are high quality and contain less than three percent missing data for each variable.

b.  Teleconnection Patterns and Indices

To examine the influence of ENSO and NAO events on United States intra-seasonal climate variability, monthly values of SST anomalies in the Niño 3.4 region of the Pacific, commonly referred to as the Niño 3.4 Index, are acquired from the NOAA/NWS CPC for all winters from 1950-2002 (NOAA/NWS CPC 2006b). In addition, monthly index values of the NAO, obtained by standardizing 500-mb height anomalies of the 20°-90°N region in the Northern Hemisphere, are gathered for the same temporal period of record from the NOAA/NWS CPC (NOAA/NWS CPC 2006c). Winters are defined annually as the period of 1 January through 31 March, and are ultimately selected for analysis over other seasons as the ENSO and NAO teleconnection signatures are often most pronounced in North America at this time of year (Smith and Sardeshmukh 2000).

Every month of record is then classified by its dominant ENSO (La Niña, El Niño, or neutral ENSO) and NAO (positive NAO, negative NAO, or neutral NAO) regime. For the ENSO teleconnection, the following standard phase limits are utilized based on monthly Niño 3.4 Index values: a month with a SST anomaly < -0.5 is classified as La Niña, a month with a SST anomaly > 0.5 is defined as El Niño, and a month with a SST anomaly ≤ 0.5 and ≥ -0.5 is a neutral ENSO month. Similarly, NAO teleconnection regime boundaries are established from index values as follows: months with index values exceeding 0.5 are designated positive NAO, below -0.5 are negative NAO, and ≤ 0.5 and ≥ -0.5 are neutral NAO events.

Regime classification by these criteria results in 49 La Niña months, 38 El Niño months, and 72 neutral ENSO months (Figure 3). Further, 56 positive NAO months are defined, 47 negative NAO months, and 56 neutral NAO months (Figure 4). For verification of teleconnection event classifications, anomaly composite maps of United States winter (January to March) air temperature and precipitation rate are created through the NOAA-CIRES CDC (NOAA-CIRES CDC 2006b) for individual ENSO and NAO phases (Figures 5, 6). Years included in the composites are limited to those with at least two winter months classed with the same phase. Though this assessment aims to assess intra-seasonal variability, the seasonal temperature and precipitation impacts are well known for these teleconnections, with the exception of neutral periods, and comparisons to the composites sufficiently validate the datasets (Barnston and Livezey 1987; Horel and Wallace 1981; Hurrell 1995; Lamb and Peppler 1987; Mantua 2000; NOAA/ NWS CPC 2004; Rogers 1984; Ropelewski and Halpert 1986; Wallace and Gutzler 1981; Yarnal and Leathers 1988).

Finally, the secondary goal of this analysis is to study intra-seasonal climate variability as a result of combination teleconnection phase (i.e., La Niña and positive NAO, El Niño and neutral NAO, etc.) impacts. Therefore, the final partitioning of the datasets is for all months where combination ENSO and NAO phases occurred. Table 1 indicates the number of months over the 53 years that fit into each phase combination grouping. 

4.  Preliminary Assessment

a.  Principal Components Analysis

To reduce the dimensionality of the original datasets and obtain regionalizations of the temperature and precipitation characteristics across the United States, principal components analyses are performed. A principal components analysis is a useful method for demonstrating the extent to which conditions vary together across space. The statistical technique is first applied to the time series of all stations for each weather variable (maximum temperature, minimum temperature, and total precipitation) in 1) January months, 2) February months, 3) March months, and 4) all winters, to identify the primary modes of natural variability in the data. This results in 12 principal components analyses. Only the leading correlation matrix components from each analysis are retained for further evaluation.

Choosing the correlation matrix components to be retained is based on an assessment of the corresponding scree graphs of latent roots (finding the "cutoff" where the line connecting latent root points changes from steep to relatively flat). In general, cutoff latent root values for components range between 0.7 and 1.0, loosely following both Kaiser and Joliffe rules (Dunteman 1989), and retained components account for 80% to 90% of the variance.

After these principal component correlation matrices are obtained, the same statistical methodologies are applied to months of only ENSO or NAO teleconnection phase activity. The spatial and temporal trends of temperature and precipitation in these component analyses can be compared to the primary modes of variability just identified to see if there are mode changes and whether or not the trends may be related to the teleconnection patterns. Therefore, for each regime (including the neutral phases) 12 principal components analyses are additionally performed.

The map patterns (loadings) gathered from each principal components analysis are then added to a GIS and mapped for each United States station. Within the GIS these data are spatially interpolated across the country with an inverse distance weighting interpolation procedure to better identify spatial trends in the data. The nonlinear interpolation technique is best suited for this research as loadings are not expected to respond uniformly across the entire study region (DeMers 2000). For the meteorological variables and winter months, unique maps are created for each principal component (in all years and in the distinct teleconnection regimes). As an example, Figure 7 displays maps of the three leading components for 1950-2002 February maximum temperature. The results depicted in the maps can be compared across components, United States regions, and winter months.

 b. Score Time Series Analysis

Finally, the score time series of the leading principal components are acquired and graphed to highlight the temporal variability of each meteorological component. The scores, generally ranging between -3 and 3, illustrate the strength of a unique mode of variability, where strong score differentiation between phases of a teleconnection pattern likely indicates that a relationship between the pattern and that particular mode of variability exists (McCabe et al. 2004). For example, it would be quite reasonable to assume ENSO is tied to January minimum temperature variability if the average score for the time series obtained in a La Niña phase test equals -0.95 and in an El Niño phase test equals 0.95, as this displays strong phase differentiation. Though cause and effect relationships cannot be purely determined by assessing the scores, the procedure is useful for identifying where teleconnection links are probable and the areas in need of further exploration.

Here, average scores are assessed for individual meteorological components and winter months uniquely by ENSO and NAO phases. Table 2 shows the average scores for February maximum temperature, minimum temperature, and total precipitation from 1950-2002 obtained for the six leading components. From the table, the mode of variability in components 3, for both maximum temperature and total precipitation, may be linked to ENSO whereas component 6 minimum temperature variability is the most likely to be ENSO influenced.

5.  Future Research Plans

The preliminary assessment just described provides a foundation for future research. The first task of this dissertation involves a broad, exploratory examination of the principal component and score time series analyses already performed for each winter month and the total season. For example, the component loading maps displayed in figure 7 show a leading component (figure 7a) that is unlikely to be controlled by a teleconnection. This map shows maximum temperatures varying together across the entire country which is a pattern likely demonstrative of natural or anthropogenic trends. However, figure 7c displays maximum temperature variability that corresponds to what one might expect to find in an ENSO year. As this variability pattern is only for the month of February it can be compared to the other winter months, and those found for the NAO, to see if a controlling teleconnection impact can be identified for United States maximum temperatures. Then, unique phases of that teleconnection can be assessed in conjunction with the scores for additional insight into what intra-seasonal variability relationships to ENSO and NAO are present. It may be important in this part of the research to also perform rotated principal component analyses in attempt to best highlight maximum variability trends.

The second task of the dissertation will test the validity of what is found in the previous step by comparing results with compositing analyses of ENSO and NAO signals. It is anticipated that all phases of each teleconnection will have to be evaluated in this manner by producing maps of the variables during months of a "strong" regime. An interesting facet of this research task involves examining the spatial variability of the potential teleconnection impacts across the country.

A third task of the dissertation will be to assess correlations between the scores and the variables. One way this can be achieved is by overlaying the ENSO and NAO indices with the score time series of individual variable components (Figures 8, 9). In evaluating where these time series peak and valley together, even more information can be gathered about the specific relationships of the two teleconnections to winter temperature and precipitation variability. For future analysis, these datasets may require the application of some preliminary filtering techniques.

Finally, the tasks already described will be performed again for all meteorological variables during combination teleconnection events (i.e., La Niña and NAO+). Though this task is left for the end, it is considered a very important part of the investigation. However, much of the information that can be gathered earlier should help determine which combinations (for which variables) will be best to examine and with how much detail.

References

American Meteorological Society 2005. Glossary of Meteorology: teleconnection.http://amsglossary.allenpress.com/glossary/search?p=1&query=teleconnection&submit=Search. Accessed December 14, 2005.

Angström, A., 1935: Teleconnections of climate changes in present time. Geogr. Annal., 17, 242-258.

Arkin, P.A., 1982: The relationship between interannual variability in the 200mb tropical wind field and the Southern Oscillation. Mon. Wea. Rev., 115, 1391-1404.

Barnston, A.G., and R.E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115, 1083-1126.

Barnston, A.G., Leetma, A., Kousky, V.E., Livezey, R.E., O'Lenic, E.A., Van den Dool, H., Wagner, A.J., and D.A. Unger, 1999: NCP forecasts of the El Niño of 1997-98 and its impacts. Bull. Amer. Meteor. Soc., 80,

    1829-1852.

Barry, R.G., and A.M. Carleton, 2001: Synoptic and Dynamic Climatology. Routledge, London, USA, and Canada, 7^th Ed., pp. 358-438.

Bjerknes, J., 1966: A possible response of the atmospheric Hadley circulation to equatorial anomalies of ocean temperature. Tellus, 18, 820-829.

Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97, 165-172.

Cayan, D. 1992: Latent and sensible heat flux anomalies over the northern oceans: The connection to monthly atmospheric circulation. J. Climate, 5, 354-369.

Collins, M., Frame, D., Sinha, B, and C. Wilson, 2002: How far ahead could we predict El Niño? Geopyhs. Res. Letts., 29, 1492.

Davis, J.C., 2002: Statistics and Data Analysis in Geology. John Wiley & Sons, Inc., USA, 3^rd Ed., pp. 299-307.

DeMers, M.M., 2000: Fundamentals of Geographic Information Systems. John Wiley & Sons, Inc., USA, 2^nd Ed., pp. 258-273.

Deser, C., and M.L. Blackmon, 1993: Surface climate variations over the North Atlantic Ocean during winter: 1900-1989. J. Climate, 6, 1743-1753.

Doberitz, R., 1968: Cross spectrum analysis of rainfall and sea temperature at the equatorial Pacific Ocean. B. Meteor. Abhhandlung., 8, 1-61.

Douglas, A.V., and P.J. Englehart, 1981: On a statistical relationship between autumn rainfall in the central equatorial Pacific and subsequent winter precipitation in

    Florida. Mon. Wea. Rev., 109, 2377-2382.

Douglas, A.V., and P.J. Englehart, 1984: Factors leading to the heavy precipitation regimes of 1982-83 in the United States. In Proc. Eight Annual Climate Diagnostics Workshop, Downsview, Ontario, pp.42-54, {NTIS   

    PB84-192418}.

Dunteman, G.H., 1989: Principal Components Analysis, Series: Quantitative Applications in the Social Sciences. Sage Publications, Inc., California, pp. 22-23.

Federov, A.V., and S.G. Philander, 2000: Is El Niño changing? Science, 288, 1997-2001.

Gershunov, A., Barnett, T.P., Cayan, D., Tubbs, T., and L. Goddard, 2000: Predicting ENSO impacts on intraseasonal precipitation statistics in California: The 1997-1998 event. J. Hydrometeorology, 1, 201-210.

Halpert, M.S., and C.F. Ropelewski, 1992: Surface temperature patterns associated with Southern Oscillation. J. Climate, 5,577-593.

Halpert, M.S., and G.D. Bell, 1997: Climate assessment for 1996. Bull. Amer. Meteor. Soc., 78, S1-S49.

Hanley, D.E., Bourassa, M.A., O'Brien, J.J., Smith, S.R., and E.R. Spade, 2002: A quantitative evaluation of ENSO indices. J. Climate, 8, 1249-1258.

Harrison, D.E., and N.K. Larkin, 1996: The COADS sea level pressure signal: a near- global El Niño composite and time series view. J. Climate, 9, 3025-3055.

Harrison, D.E., and N.K. Larkin, 1998: El Niño Southern Oscillation sea surface temperature and wind anomalies. Rev. Geophys., 36, 353-399.

Higgins, R.W., Leetma, A., Xue, Y., and A. Barnston, 2000: Dominant factors influencing predictability of U.S. precipitation and surface air temperature. J. Climate, 13, 3994-4017.

Hildebransson, H.H., 1897: Quelques recherches sur le centers d'action de l'atmosphere: ecarts et moyennes barometriques mensuelles. Kongl. Svenska Vetenscaps-Akad. Handlinger 29, Stockholm, 36 pp.

Horel, J.D., and J.M. Wallace, 1981: Planetary atmospheric phenomena associated with the Southern Oscillation. Mon. Wea. Rev., 109, 813-829.

Hurrell, J.W., 1995: Decadal trends in the North Atlantic Oscillation: regional temperature and precipitation. Science, 269, 676-679.

Hurrell, J.W., 1996: Influence of variations in extratropical wintertime teleconnections on Northern Hemisphere temperatures. Geophys. Res. Lett., 23, 665-668.

Hurrell, J.W., and H. van Loon, 1997: Decadal variations in climate associated with the North Atlantic Oscillation. Climatic Change, 36, 301-326.

Hurrell, J.W., Kushnir, Y., and M. Visbeck, 2001: The North Atlantic Oscillation. Science, 291, 603-605.

Jones, P.D., Jonssen, T., and D. Wheeler, 1997: Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and south-west Iceland. Int. J. Climatol., 17, 1433-1450.

Julian, P.R., and R.M. Chervin, 1978: A study of the southern oscillation and Walker circulation phenomenon. Mon. Wea. Rev., 106, 1433-1451.

Kapala, A., Mächel, H., and H. Flohn, 1998: Behavior of the centers of action above the Atlantic since 1881. II. Association with regional climate anomalies. Int. J. Climatol., 18, 23-26.

Kidson, J.W., 1975: Eigenvector analysis of monthly mean surface data. Mon. Wea. Rev. 103, 187-196.

Kiladis, G.W., and H.F. Diaz, 1989: Global climatic anomalies associated with extreme in the Southern Oscillation. J. Climate, 2, 1069-1090.

Lamb, P.J., and R.A. Peppler, 1987: North Atlantic Oscillation: concept and an application. Bull. Amer. Meteor. Soc., 68, 1218-1225.

Livezey, R.E., Masutani, M., Leetma, A., Rui, H., Ji, M., and A. Kumar, 1997: Teleconnection response of the Pacific-North American region atmosphere to large central equatorial Pacific sst anomalies. J. Climate, 10,  

    1787-1820.

Livezey, R.E., and T.M. Smith, 1999: Covariability of aspects of North American climate with global sea surface temperatures on interannual to interdecadal timescales. J. Climate, 12, 289-302.

Lockyer, W.J.S., 1906: Barometric variations of long duration over large areas. Proc. Roy. Soc., A78, 43-60.

Mächel, H., Kapala, A., and H. Flohn, 1998: Behavior of the centers of action above the Atlantic since 1881. I. Characteristics of seasonal and interannual variability. Int. J. Climatol., 18, 1-22.

Miles, E.L., Mantua, N., and P. Mote, 2001: ENSO impacts on the Pacific Northwest: An Integrated Assessment, unpublished manuscript, Seattle, WA, USA.

Mantua, N., 2000: La Niña impacts in the Pacific Northwest, JISAO, University of Washington, unpublished manuscript, Seattle, WA, USA.

McCabe, G.J., Palecki, M.A., and J.L. Betancourt, 2004: Pacific and Atlantic Ocean influences on multidecadal drought frequency in the United States. PNAS, 101, 4136-4141.

Mo, K.C., and R.E. Livezey, 1986: Tropical-extratropical geopotential height teleconnections during the Northern Hemisphere winter. Mon. Wea. Rev., 114, 2488-2515.

Nakamura, H. 1996: Year-to-year and interdecadal variability in the activity of intraseasonal fluctuations in the Northern Hemisphere circulation. Theor. Appl. Climatol., 55, 19-32.

NASA Earth Observatory 2003. Searching for Atlantic Rhythms: Winter Weather and the North Atlantic Oscillation. http://earthobservatory.nasa.gov/Study/NAO_200307. Accessed August 25, 2006.

NOAA-CIRES CDC 2005. Atmospheric Teleconnection Patterns. http://www.cdc.noaa. gov/Teleconnections/. Accessed December 14, 2005.

NOAA-CIRES CDC 2005. Multivariate ENSO Index (MEI). http.//www/cdc.noaa.gov/people/klaus.wolter/MEI/mei.html. Accessed February 9, 2005.

NOAA-CIRES CDC 2006a. Linear Correlations in Atmospheric Seasonal/ monthly Averages. http://www.cdc.noaa.gov/Correlation/. Accessed October 10, 2006.

NOAA-CIRES CDC 2006b. Monthly/ Seasonal Climate Composites. http://www.cdc. Noaa.gov/cgi-bin/Composites/printpage.pl. Accessed March 9, 2006.

NOAA NCDC 2006a. United States Historical Climatology Network (USHCN). http://www.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html. Accessed January 24,2006.

NOAA NCDC 2006b. United States Historical Climatology Network Daily Temperature, Precipitation, and Snow Data for1871-1997. ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/daily/README. Accessed January 24, 2006.

NOAA/NWS CPC 2004. The ENSO Cycle. http://cpc.ncep.noaa.gov/products/aalysis_Monitoring/lamina/. Accessed February 26, 2004.

NOAA/NWS CPC 2005. Northern Hemisphere Teleconnection Patterns: Introduction. http://www.cpc.noaa.gov/teledoc/teleintro.shtml. Accessed December 14, 2005.

NOAA/NWS CPC 2006a. NAO-North Atlantic Oscillation. http://www.cpc.noaa.gov/Products/predictions/90day/seasglossar.html#nao. Accessed October 9, 2006.

NOAA/NWS CPC 2006b. Sea Surface Temperature (SST) Monthly. http://www.cpc.ncep.noaa.gov/data/indices/sstoi.indices. Accessed January 5, 2006.

NOAA/NWS CPC 2006c. Northern Hemisphere Teleconnection Indices. ftp://ftpprd.ncep.noaa.gov/pub/cpc/wd52dg/data/indices/tele_index.nh. Accessed January 5, 2006.

Norton, J., McLain, L., Brainard, R., and D. Husby, 1985: The 1982-83 El Niño event off Baja and Alta, California, and its ocean climate context. In El Niño North: Niño effects in the eastern sub-arctic Pacific Ocean. Eds. By

    Wooster and Fluharty,{Available from the Washington Sea Grant Program, Colleges of Ocean and Fishery Sciences, University of Washington}.

Rasmusson, E.M., and T.H. Carpenter, 1982: Variations in tropical sea surface temperature and surface wind fields associated with the Southern oscillation/ El Niño. Mon. Wea. Rev., 110, 354-384.

Rogers, J.C., 1984: The association between the North Atlantic Oscillation and the Southern Oscillation in the Northern Hemisphere. Mon. Wea. Rev., 112, 1999- 2015.

Rogers, J.C., 1990: Patterns of low-frequency monthly sea-level pressure variability (1899-1986) and associated wave cyclone frequencies. J. Clim., 3, 1364-1379.

Rogers, J.C., 1997: North Atlantic storm track variability and its association to the North Atlantic Oscillation and climate variability of northern Europe. J. Climate,10, 1635-1647.

Ropelewski, C.F., and M.S. Halpert, 1986: North American precipitation and temperature patterns associated with the El Niño/ Southern Oscillation (ENSO). Mon. Wea. Rev., 114, 2352-2362.

Schonher, T., and S.E. Nicholson, 1989: The relationship between California rainfall and ENSO events. J. Climate, 2, 1258-1269.

Smith, C.A., and P. Sardeshmukh, 2000: The effect of ENSO on the intraseasonal variance of surface temperature in winter. Int. J. Climatol., 20, 1543-1557.  

Smith, R.S., Legler, D.M., Remigio, M.J., and J.J. O'Brien: 1999. Comparison of 1997-98 U.S. temperature and precipitation anomalies to historical ENSO warm phases. J. Climate, 12, 3507-3515.

Trenberth, K.E., 1997: The definition of El Niño. Bull. Amer. Meteor. Soc., 78, 2771-2777.

Troup, A.J., 1965: The "southern oscillation"., J. Roy. Meteor. Soc., 102, 490-506.

UCAR/NOAA, 1994: El Niño and Climate Prediction. In Reports to the nation on our changing planet, Spring 1994, Ed. by Wallace, J.M. and S. Vogel, {NA27GP0232-01}.

UCAR 2005. Children of the tropics: El Niño and La Niña. http://www.ucar.edu/communications/factsheets/elnino/. Accessed February 10, 2005.

US Department of Commerce [USDEC], NOAA, NWS, NCEP 2006. Prediction Skill of  Major Extratropical Teleconnection Patterns on Daily Time Scales Corrected by an Analog Approach. http://www.emc.ncep.noaa.gov/

    officenotes/newernotes/on 449.pdf. Accessed October 9, 2006.  

van Loon, H., and R.A. Madden, 1981: The Southern oscillation I: Global associations with pressure and temperature in northern winter. Mon. Wea. Rev., 109, 1150-1162.

van Loon, H., and J.C. Rogers, 1978: The seesaw in winter temperature between Greenland and northern Europe. Part I: general description. Mon. Wea. Rev., 106, 296-310.

Visbeck, M. 2005. North Atlantic Oscillation. http://www.ldeo.columbia.edu/NOA/. Accessed February 12, 2005.

Walker, G.T., 1923: Correlations in seasonal variations of weather, VIII. Mem. Indian Meteor. Dept., 24, 75-131.

Walker, G.T., 1924: Correlations in seasonal variations of weather, IX. Mem. Indian Meteor. Dept., 24, 275-332.

Walker, G.T., and E.W. Bliss, 1932: World Weather V. Mem. Roy. Meteor. Soc., 44, 53-84.

Wallace, J.M., and D.S. Gutzler, 1981: Teleconnections in the geopotential height field during the Northern Hemisphere winter. Mon. Wea. Rev., 109, 784-812.

Wallace, J.M., Rasmusson, E.M., Mitchell, T.P., Sarachik, E.S., and H. von Storch, 1998: On the structure and evolution of ENSO-related climate variability in the tropical Pacific: lessons from TOGA. J. Geophys. Res., 103,

    14241-14259.

Wang, H.J., Zhang, R.E., Cole, J., and F. Chavez, 1999: El Niño and the related phenomenon Southern Oscillation (ENSO): The largest signal in interannual

            climate variation. PNAS, 96, 11071-11072.

Webster, P.J., 1981: Mechanisms determining the atmospheric response to sea surface temperature anomalies. J. Atmos. Sci., 38, 554-571.

Western Regional Climate Center 2006. El Niño, La Niña, and the Western U.S., Alaska, and Hawaii. http://www.wrcc.dri.edu/enso/ensofaq.html. Accessed August 13, 2006.

Wright, P.B, 1984: Relationships between the indices of the Southern Oscillation. Mon. Wea. Rev., 112, 1913-1919.

Wright, P.B., 1985: The Southern Oscillation: an ocean-atmosphere feedback system? Bull. Amer. Meteor. Soc., 66, 398-412.

Wright, P.B., Wallace, J.M., Mitchell, T.P., and C. Deser, 1988: Correlation structure of the El Niño/ Southern Oscillation phenomenon. J. Climate, 1, 609-625.

Yarnal, B. and D.J. Leathers, 1988: Relationships between interdecadal and interannual climatic variations and their effect on Pennsylvania climate. Ann. Assoc. Amer. Geog., 78, 624-641.

** The figures referenced in this write-up will be added soon.