• Anctil, F., , and Coulibaly P. , 2004: Wavelet analysis of the interannual variability in southern Quebec streamflow. J. Climate, 17 , 163173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Barros, V., , Chamorro L. , , Coronel G. , , and Baez J. , 2004: The major discharge events in the Paraguay River: Magnitudes, source regions, and climate forcings. J. Hydrometeor., 5 , 11611170.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., , Hodges K. I. , , and Hagemann S. , 2004: Sensitivity of the ERA40 reanalysis to the observing system: Determination of the global atmospheric circulation from reduced observations. Tellus, 56A , 456471.

    • Search Google Scholar
    • Export Citation
  • Beniston, M., , and Stephenson D. B. , 2004: Extreme climatic events and their evolution under changing climatic conditions. Global Planet. Change, 44 , 19.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradbury, J. A., , Dingman S. L. , , and Keim B. D. , 2002a: New England drought and relations with large scale atmospheric circulation patterns. J. Amer. Water Resour. Assoc., 38 , 12871299.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradbury, J. A., , Keim B. D. , , and Wake C. P. , 2002b: U.S. East Coast trough indices at 500 hPa and New England winter climate variability. J. Climate, 15 , 35093517.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradbury, J. A., , Keim B. D. , , and Wake C. P. , 2003: The influence of regional storm tracking and teleconnections on winter precipitation in the northeastern United States. Ann. Assoc. Amer. Geogr., 93 , 544556.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Briffa, K. R., , and Jones P. D. , 1993: Global surface air temperature variations during the twentieth century: Part 2: Implications for large-scale high-frequency palaeoclimatic studies. Holocene, 3 , 7788.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., 1976: Winter cyclone frequencies over the eastern United States and adjacent North Atlantic, 1964–1973. Bull. Amer. Meteor. Soc., 57 , 548553.

    • Search Google Scholar
    • Export Citation
  • Dai, A., , Fung I. Y. , , and DelGenio A. D. , 1997: Surface observed global land precipitation variations during 1900–88. J. Climate, 10 , 29432962.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Déry, S. J., , and Wood E. F. , 2004: Teleconnection between the Arctic Oscillation and Hudson Bay river discharge. Geophys. Res. Lett., 31 .L18205, doi:10.1029/2004GL020729.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , and Blackmon M. L. , 1993: Surface climate variations over the North Atlantic Ocean during winter—1900–1989. J. Climate, 6 , 17431753.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dudley, R. W., , and Hodgkins G. A. , 2002: Trends in streamflow, river ice and snowpack for coastal river basins in Maine during the 20th century. U.S. Geological Survey Water-Resources Investigations Rep. 02-4245, 26 pp.

  • Hannah, D. M., , and McGregor G. R. , 1997: Evaluating the impact of climate on snow- and ice-melt dynamics in the Taillon basin, French Pyrenees. J. Glaciol., 43 , 563568.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hannah, D. M., , Smith B. P. G. , , Gurnell A. M. , , and McGregor G. R. , 2000: An approach to hydrograph classification. Hydrol. Processes, 14 , 317338.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hartley, S., , and Dingman S. L. , 1993: Effects of climatic variability on winter–spring runoff in New England river basins. Phys. Geogr., 14 , 379393.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hartley, S., , and Keables M. J. , 1998: Synoptic associations of winter climate and snowfall variability in New England, USA, 1950-1992. Int. J. Climatol., 18 , 281298.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hodgkins, G. A., , Dudley R. W. , , and Huntington T. G. , 2003: Changes in the timing of high river flows in New England over the 20th Century. J. Hydrol., 278 , 244252.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hodgkins, G. A., , Dudley R. W. , , and Huntington T. G. , 2005: Summer low flows in New England during the 20th century. J. Amer. Water Resour. Assoc., 41 , 403412.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Houghton, J. T., , Ding Y. , , Griggs D. J. , , Noguer M. , , van der Linden P. J. , , Dai X. , , Maskell K. , , and Johnson C. A. , 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.

    • Search Google Scholar
    • Export Citation
  • Houseago, R. E., , McGregor G. R. , , King J. C. , , and Harangozo S. A. , 1998: Climate anomaly wave-train patterns linking southern low and high latitudes during South Pacific warm and cold events. Int. J. Climatol., 18 , 11811193.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huntington, T. G., , Hodgkins G. A. , , and Dudley R. W. , 2003: Historical trend in river ice thickness and coherence in hydroclimatological trends in Maine. Climatic Change, 61 , 217236.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., 1995: Decadal trends in the North-Atlantic Oscillation—Regional temperatures and precipitation. Science, 269 , 676679.

  • Junge, M. M., , and Stephenson D. B. , 2003: Mediated and direct effects of the North Atlantic Ocean on winter temperatures in northwest Europe. Int. J. Climatol., 23 , 245261.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Karl, T. R., , Williams C. N. Jr., , Quinlan F. T. , , and Boden T. A. , 1990: United States Historical Climatology Network (HCN) serial temperature and precipitation data. Environmental Science Division Publication 3404, Carbon Dioxide Information and Analysis Center, Oak Ridge National Laboratory, Oak Ridge, TN, 389 pp.

  • Keim, B. D., cited. 2006: A climate primer for New England. [Available online at http://airmap.unh.edu/background/ClimatePrimer.html.].

  • Keim, B. D., , Meeker L. D. , , and Slater J. F. , 2005: Manual synoptic classification for the East Coast of New England (USA) with an application to PM2.5 concentration. Climate Res., 28 , 143154.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kingston, D. G., , Lawler D. M. , , and McGregor G. R. , 2006: Linkages between atmospheric circulation, climate and streamflow in the northern North Atlantic: Research prospects. Prog. Phys. Geogr., 30 , 143174.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Krasovskaia, I., , Arnell N. , , and Gottschalk L. , 1994: Flow regimes in northern and western Europe: Development and application of procedures for classifying flow regimes. FRIEND: Flow Regimes from International Experimental and Network Data, P. Seuna, et al., Eds., IAHS Publication 221, 185–192.

  • Krepper, C. M., , García N. O. , , and Jones P. D. , 2003: Interannual variability in the Uruguay River basin. Int. J. Climatol., 23 , 103115.

  • Lawler, D. M., , McGregor G. R. , , and Phillips I. D. , 2003: Influence of atmospheric circulation changes and regional climate variability on river flow and suspended sediment fluxes in southern Iceland. Hydrol. Processes, 17 , 31953223.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lins, H. F., 1997: Regional streamflow regimes and hydroclimatology of the United States. Water Resour. Res., 33 , 16551667.

  • Lins, H. F., , and Slack J. R. , 1999: Streamflow trends in the United States. Geophys. Res. Lett., 26 , 227230.

  • Livezey, R. E., , and Chen W. Y. , 1983: Statistical field significance and its determination by Monte Carlo techniques. Mon. Wea. Rev., 111 , 4659.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Magilligan, F. J., , and Graber B. E. , 1996: Hydroclimatological and geomorphic controls on the timing and spatial variability of floods in New England, USA. J. Hydrol., 178 , 159180.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molnár, P., , and Ramírez J. A. , 2001: Recent trends in precipitation and streamflow in the Rio Puerco basin. J. Climate, 14 , 23172328.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Peterson, T. C., and Coauthors, 1998: Homogeneity adjustments of in situ atmospheric climate data: A review. Int. J. Climatol., 18 , 14931517.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Phillips, I. D., , McGregor G. R. , , Wilson C. J. , , Bower D. , , and Hannah D. M. , 2003: Regional climate and atmospheric circulation controls on the discharge of two British rivers, 1974–97. Theor. Appl. Climatol., 76 , 141164.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rimbu, N., , Boroneant C. , , Buta C. , , and Dima M. , 2002: Decadal variability of the Danube River flow in the lower basin and its relation with the North Atlantic Oscillation. Int. J. Climatol., 22 , 11691179.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., , Rowell D. P. , , and Folland C. K. , 1999: Oceanic forcing of the wintertime North Atlantic Oscillation and European climate. Nature, 398 , 320323.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Seaber, P. R., , Kapinos F. P. , , and Knapp G. L. , 1987: Hydrologic unit maps. USGS Water Supply Paper 2294, 63 pp.

  • Serreze, M. C., , Carse F. , , Barry R. G. , , and Rogers J. C. , 1997: Icelandic low cyclone activity: Climatological features, linkages with the NAO, and relationships with recent changes in the Northern Hemisphere circulation. J. Climate, 10 , 453464.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., , Bromwich D. H. , , Clark M. P. , , Etringer A. J. , , Zhang T. , , and Lammers R. , 2003: Large-scale hydro-climatology of the terrestrial Arctic drainage system. J. Geophys. Res., 108 .8160, doi:10.1029/2001JD000919.

    • Search Google Scholar
    • Export Citation
  • Slack, J. R., , and Landwehr J. M. , 1992: Hydro-Climatic Data Network: A U.S. Geological Survey streamflow data set for the United States for the study of climate variations, 1874-1988. USGS Open-File Rep. 92-129, 193 pp.

    • Crossref
    • Export Citation
  • Stenseth, N. C., , Ottersen G. , , Hurrell J. W. , , Mysterud A. , , Lima M. , , Chan K. S. , , Yoccoz N. G. , , and Adlandsvik B. , 2003: Studying climate effects on ecology through the use of climate indices: The North Atlantic Oscillation, El Nino Southern Oscillation and beyond. Proc. Roy. Soc. London, 270B , 20872096.

    • Search Google Scholar
    • Export Citation
  • Stewart, I. T., , Cayan D. R. , , and Dettinger M. D. , 2005: Changes toward earlier streamflow timing across western North America. J. Climate, 18 , 11361155.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Struglia, M. V., , Mariotti A. , , and Filograsso A. , 2004: River discharge into the Mediterranean Sea: Climatology and aspects of the observed variability. J. Climate, 17 , 47404751.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Taylor, A. H., 1996: North-south shifts of the Gulf Stream: Ocean-atmosphere interactions in the North Atlantic. Int. J. Climatol., 16 , 559583.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trigo, R. M., , Pozo-Vasquez D. , , Osborn T. J. , , Castro-Diez Y. , , Gamis-Fortez S. , , and Esteban-Parra M. J. , 2004: North Atlantic Oscillation influence on precipitation, river flow and water resources in the Iberian Peninsula. Int. J. Climatol., 24 , 925944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 re-analysis. Quart. J. Roy. Meteor. Soc., 131 , 29613012.

  • U.S. Water Resources Council, 1970: Water resources regions and subregions for the national assessment of water and related land resources. Washington DC, 188 pp.

  • Wettstein, J. J., , and Mearns L. O. , 2002: The influence of the North Atlantic–Arctic Oscillation on mean, variance, and extremes of temperature in the northeastern United States and Canada. J. Climate, 15 , 35863600.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yarnal, B., 1993: Synoptic Climatology in Environmental Analysis. Belhaven Press, 195 pp.

  • Yarnal, B., , and Leathers D. J. , 1988: Relationships between interdecadal and interannual climatic variations and their effect on Pennsylvania climate. Ann. Assoc. Amer. Geogr., 78 , 624641.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yarnal, B., , and Frakes B. , 1997: Using synoptic climatology to define representative discharge events. Int. J. Climatol., 17 , 323341.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ye, H., , Yang D. , , Zhang T. , , Zhang X. , , Ladochy S. , , and Ellison M. , 2004: The impact of climatic conditions on seasonal river discharge in Siberia. J. Hydrometeor., 5 , 286295.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Hypothesized process chain linking the winter NAO and river flow in New England [modified from Kingston et al. (2006)].

  • View in gallery

    Location of New England and river flow gauging stations used.

  • View in gallery

    Mean (1958–2001) annual flow regime for New England rivers, and (1961–90) New England precipitation and temperature [derived from U.S. Historical Climatology Network (Karl et al. 1990)].

  • View in gallery

    Composite 1000-hPa geopotential height (m) and wind vector (m s−1) for February (top) high and (bottom) low flow. [Length of arrow indicates wind strength (see scale)]

  • View in gallery

    Difference (in m) between February 1000-hPa geopotential height composites (high minus low flow composite). [Shading indicates significant difference between the composites (with 8 degrees of freedom); p levels are (from light to dark shades) 0.05, 0.01, and 0.001]

  • View in gallery

    Composite 500-hPa geopotential height (m) and wind vector (m s−1) for February (top) high and (bottom) low flow. [Length of arrow indicates wind strength (see scale)]

  • View in gallery

    Difference (in s−1) between December 500-hPa relative vorticity composites (high minus low flow composite). (Shading as in Fig. 5)

  • View in gallery

    Difference (in °C) between February 1000-hPa temperature composites (high minus low flow composite). (Shading as in Fig. 5)

  • View in gallery

    Composite 1000-hPa temperature (in °C) for March (left) high and (right) low flow.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 40 41 11
PDF Downloads 26 26 4

Large-Scale Climatic Controls on New England River Flow

View More View Less
  • 1 School of Geography, Earth, and Environmental Sciences, University of Birmingham, Birmingham, United Kingdom
  • | 2 Department of Geography, King’s College London, London, United Kingdom
  • | 3 School of Geography, Earth, and Environmental Sciences, University of Birmingham, Birmingham, United Kingdom
© Get Permissions
Full access

Abstract

Understanding atmospheric drivers of river flow variability necessitates clear knowledge of the process chain linking climate and hydrology, yet the nature of such linkages remains poorly understood for the New England region of the northeastern United States. This research gap is addressed through a composite analysis of large-scale climatic controls on monthly high and low river flow in New England for 1958–2001, based on 40-yr ECMWF Re-Analysis (ERA-40) data. Analysis is focused on climate fields at the North Atlantic spatial scale, with particular attention given to the influence of the North Atlantic Oscillation (NAO).

High (low) river flow is shown to be characterized by greater (lower) geopotential height throughout the year, and from December to April, higher (lower) temperature. Wind speed is inversely associated with river flow in all months, with wind direction more southerly (northerly) under high (low) flow situations. Relative vorticity differences reveal more cyclonic circulation centered downwind of New England under low river flow conditions (compared to high flow) from December to April. Reversal of river flow associations with temperature and vorticity in May are linked to snowmelt dynamics. Although cursory analysis suggests a positive association between the NAO and New England river flow, closer inspection reveals this to be less straightforward. River flow is more closely linked to the East Coast trough (rather than the Icelandic low and Azores high), while air temperature anomalies resemble the NAO–sea surface temperature rather than NAO–air temperature pattern.

Corresponding author address: Daniel G. Kingston, School of Geography, Earth, and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom. Email: dgk366@bham.ac.uk

Abstract

Understanding atmospheric drivers of river flow variability necessitates clear knowledge of the process chain linking climate and hydrology, yet the nature of such linkages remains poorly understood for the New England region of the northeastern United States. This research gap is addressed through a composite analysis of large-scale climatic controls on monthly high and low river flow in New England for 1958–2001, based on 40-yr ECMWF Re-Analysis (ERA-40) data. Analysis is focused on climate fields at the North Atlantic spatial scale, with particular attention given to the influence of the North Atlantic Oscillation (NAO).

High (low) river flow is shown to be characterized by greater (lower) geopotential height throughout the year, and from December to April, higher (lower) temperature. Wind speed is inversely associated with river flow in all months, with wind direction more southerly (northerly) under high (low) flow situations. Relative vorticity differences reveal more cyclonic circulation centered downwind of New England under low river flow conditions (compared to high flow) from December to April. Reversal of river flow associations with temperature and vorticity in May are linked to snowmelt dynamics. Although cursory analysis suggests a positive association between the NAO and New England river flow, closer inspection reveals this to be less straightforward. River flow is more closely linked to the East Coast trough (rather than the Icelandic low and Azores high), while air temperature anomalies resemble the NAO–sea surface temperature rather than NAO–air temperature pattern.

Corresponding author address: Daniel G. Kingston, School of Geography, Earth, and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom. Email: dgk366@bham.ac.uk

1. Introduction

Understanding the climatic forcing of river flow represents a major research challenge of practical relevance, due to high socioeconomic and ecological dependence on water resources. This relevance is further enhanced in light of the pressing need to predict future water stress and risk within the context of climate change (Houghton et al. 2001). Hydrologists have long been aware of the influence of climate on river flow, although traditional analyses rarely extended beyond the drainage basin scale. As noted in Kingston et al. (2006), increased attention has been given to large-scale hydroclimatological linkages within the last decade, and a more integrated approach adopted to examine hydrological and climatological variability within a spatiotemporal framework of ocean–atmosphere–land surface interactions (i.e., the hydroclimatological system). Several recent studies have sought to quantify variation and interrelationships between atmospheric processes and river flow, including Bradbury et al. (2002b), Stewart et al. (2005), and Molnár and Ramírez (2001) in the United States; Anctil and Coulibaly (2004) and Déry and Wood (2004) in Canada; Krepper et al. (2003) in Uruguay; Lawler et al. (2003) in southwest Iceland; Phillips et al. (2003) in Britain; Struglia et al. (2004) across the Mediterranean; and Ye et al. (2004) in Siberia.

In developing predictive and explanatory models for the hydroclimatological system, researchers make the assumption that climatological variables are the primary forcing factor of river flow dynamics (Krasovskaia et al. 1994). Amongst the most commonly used climatological variables are indices of atmospheric circulation patterns. These indices are useful as they provide a surrogate measure of the general climatic situation and, therefore, a summary of complex internal atmospheric relationships (Stenseth et al. 2003). One such pattern is the North Atlantic Oscillation (NAO), indices of which have been shown to explain approximately one-third of surface temperature variation north of 20°N (Hurrell 1995) and 10% of the variation in North Atlantic precipitation (Dai et al. 1997).

The NAO describes the inverse correlation of atmospheric mass between the two semipermanent centers of action in the North Atlantic, the Icelandic low and the Azores high, and is the primary mode of atmospheric circulation variation in the North Atlantic region (Hurrell 1995). When both centers of action are active (weak), a greater (reduced) pressure gradient exists between these centers, and the NAO is described as being in a positive (negative), or high (low), phase (Hurrell 1995). In the New England region of the northeast United States (i.e., Maine, New Hampshire, Vermont, Connecticut, Massachusetts, and Rhode Island), a high NAO is associated with an enhanced southerly component to the prevailing westerly winds (Hurrell 1995; Yarnal and Leathers 1988) and a more northerly positioned but zonally orientated polar front jet (Bradbury et al. 2002b). Relationships between a negative NAO and anomalously cool regional sea surface temperatures (SSTs) have also been suggested (Bradbury et al. 2002b, 2003; Hartley and Keables 1998). The position and intensity of the East Coast pressure trough is linked to the state of the NAO, with a more easterly position and deeper trough evident in negative NAO winters (Bradbury et al. 2002a). Correspondingly, Wettstein and Mearns (2002) have documented a positive association between seasonal air temperature in New England and the NAO.

Although negative NAO winters are associated with increased snowfall due to incursion of polar air (Hartley and Keables 1998), fewer depressions also pass over the region due to storm tracks following a more offshore path (Bradbury et al. 2003). This complex relationship between the NAO and northeast U.S. precipitation is further highlighted by Yarnal and Leathers (1988): while weak (but statistically significant) positive correlation between the NAO and precipitation was found for Pennsylvania, the more northerly wind vector component associated with a negative NAO was shown to result in either above- or below-average regional precipitation, depending on the precise position of the East Coast trough. Perhaps reflecting this, Bradbury et al. (2002b) found no statistically significant (hereafter p = 0.05) correlation between the NAO and monthly New England rainfall.

A number of hydroclimatological trends over the last 30–40 yr have been observed across New England. These include a reduction in winter river ice thickness, increasing winter air temperatures, an earlier date of river ice break-up in spring (Huntington et al. 2003), earlier spring peak flow (Dudley and Hodgkins 2002; Hodgkins et al. 2003), a reduction in the winter snow: rain ratio (Hartley and Dingman 1993), and an increasing trend in annual river flow, particularly in minimum to median flow quantiles (Lins and Slack 1999). These investigations cite the NAO, and its increased tendency toward a positive state between the mid-1970s and mid-1990s, as a possible causal factor of these trends. Despite these apparent links, there is a paucity of quantitative investigations into the relationship between the NAO and New England river flow (Kingston et al. 2006). An exception is the work of Bradbury et al. (2002b), who described interrelationships between the NAO and discharge for 37 New England rivers at monthly, seasonal, and decadal time scales over the course of the twentieth century. Although mean December–March discharge was not significantly correlated with the NAO for any of the rivers examined, correlation between monthly river flow and the NAO was found to be weakly positive but nevertheless statistically significant. Only three rivers displayed statistically significant correlation with the NAO when decadally smoothed. This link was explained through correlation of the NAO and New England air temperature, with no significant relationship found between the NAO and precipitation.

Although previous studies therefore indicate an NAO influence on the hydroclimatology of New England, it is apparent that the cascade of processes linking climate and hydrology has not yet been systematically defined. This is exemplified by the apparent lack of correlation between the NAO and precipitation despite a statistically significant NAO link to river flow. Aside from the variable influence of the NAO on precipitation delivery (Yarnal and Leathers 1988), the differing precipitation versus river flow NAO association may be associated with the different temporal resolution of analysis and hydrological buffering within river basins (Hannah et al. 2000). This is likely to be further complicated by the compensation effect between winter precipitation delivery in the solid phase and subsequent spring melt causing a temporal dislocation of water input and output in New England.

Given the above research gap and the potential importance of elucidating a quantifiable NAO signal in river flow (i.e., for development of predictive relationships, water resource management, and interpretation of climate change scenarios), this paper aims to systematically investigate the process cascade linking large-scale climate and river flow. Particular emphasis is given to identifying NAO signatures in this process cascade. This is achieved through a composite analysis comparing North Atlantic climate fields between monthly high and low river flow conditions in New England, for 1958–2001. As an organizing framework for this investigation, a conceptual model of the process cascade linking New England river flow to the NAO is provided (Fig. 1). Based on the literature review of Kingston et al. (2006), the model hypothesizes a nested suite of processes from the NAO to river flow, including regional SSTs, the position and intensity of the East Coast pressure trough, wind direction, storm-track position, regional air temperature and precipitation, and river flow state.

The paper is structured as follows: Section 2 explains the datasets and methodology. Section 3 provides a brief description of the hydrological and climatological characteristics of the New England region. Composite analysis results are presented in section 4 and discussed in section 5, before conclusions are drawn in section 6.

2. Data and methods

Monthly river flow data for 1958–2001 were obtained from the Hydro-Climatic Data Network (HCDN; Slack and Landwehr 1992). River flow data are only included in the HCDN if they conform to rigorous criteria regarding accuracy and stability of the stage–discharge relationship, levels of land-use change, and anthropogenic influence on discharge. This is therefore a near-ideal dataset to use for investigation of climatic impacts on river flow (Slack and Landwehr 1992). Four selection criteria were employed to determine the stations sampled from the HCDN: 1) continuous record spanning at least 30 yr; 2) record finish date in the late 1990s, or preferably 2001 (so as to match the climatic data span of 1958–2001); 3) gauged basin area between 50 and 1000 km2; and 4) independent basins (i.e., not subbasins of other gauged areas). The latter criterion was achieved using the U.S. Geological Survey (USGS) hydrologic unit codes assigned to each station (Seaber et al. 1987). These scale through four categories from major geographic regions (of which New England is one) to subregions, accounting units and cataloging units (Seaber et al. 1987). No more than one gauging station was taken from each accounting unit, and only one or two gauging stations were used from most subregions [defined as a river system, a reach and associated tributaries, a closed basin(s), or coastal drainage area (Seaber et al. 1987)]. These selection criteria yielded 21 river flow gauging stations (Table 1; Fig. 2).

An environment-to-circulation approach using composite analysis (Yarnal 1993) was adopted for analyzing climatic drivers of high and low river flow. This involved identifying high and low river flow situations and then the common climatic conditions associated with each of these flow levels. Examples of previous use of this technique for hydroclimatological investigations include work on climatic forcing of Paraguay River discharge (Barros et al. 2004), climatic impacts on snow and ice-melt in the Pyrenees (Hannah and McGregor 1997) and synoptic conditions associated with above-average discharge in the Susquehanna River basin in Pennsylvania (Yarnal and Frakes 1997). Although such an approach reveals statistical associations, it does not yield direct cause-and-effect relationships. Nevertheless, by examining a suite of climate variables, this paper aims to demonstrate a sufficient number of statistical associations to significantly improve understanding of links in the climate–river flow process cascade for New England.

Records were split into monthly time series and standardized prior to analysis. Standardization was achieved by centering each record around its 1958–2001 mean and scaling by the standard deviation (i.e., z scores), and was necessary to control for differences in river flow magnitude arising from contrasts in drainage basin size. A spatial average (unweighted arithmetic mean) of standardized data was then taken for each monthly time step, creating a 1958–2001 all–New England river flow time series for each month of the year. Each monthly all–New England river flow series was ranked according to standardized discharge, and the top and bottom 8 yr for each month were selected as the high and low flow years, respectively (equating to the approximate 20th and 80th percentiles of each monthly time series). These then formed the basis for the subsequent composite analysis. The 20th and 80th percentiles were chosen as the best compromise between sampling only high and low river flow while also maintaining an acceptable sample size. Such a relative (rather than absolute) discharge threshold is considered to be the most objective and consistent criterion for selection of high and low flow years (Beniston and Stephenson 2004).

Although the New England region is designated a discrete hydrological unit by the U.S. Water Resource Council (U.S. Water Resources Council 1970), this is based on drainage divides rather than factors affecting hydrological variability (Lins 1997). Despite this, New England has been shown to broadly correspond to the northeast hydroclimatological region identified by Lins (1997). Correlation of individual records with the regional mean (used herein) also indicates sufficient spatial homogeneity for an average to be representative of monthly regional variation (mean correlation coefficients between stations and the regional mean vary from 0.64 in September to 0.82 in February). Although the presence of local variation within this region is acknowledged, these analyses suggest this is not sufficient to preclude the use of a regional mean for the purposes of this investigation.

Composite climate fields were derived from the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40; Uppala et al. 2005) for the region 25°–85°N, 100°W–20°E (at a resolution of 2.5°). Reanalysis datasets such as ERA-40 provide the advantage of a comprehensive range of climatological variables with global coverage, and have been used in a number of previous hydroclimatological studies (e.g., Serreze et al. 2003; Barros et al. 2004; Trigo et al. 2004). Fluxes and values within reanalysis datasets are determined by a numerical weather prediction data assimilation system rather than directly from observations; thus model physics and parameterizations have implications for the dataset (Bengtsson et al. 2004). Furthermore, changes in the observational data network used to drive the reanalysis (such as the introduction of satellite data since 1979) have caused temporal biases in the data output (Bengtsson et al. 2004). These effects are mediated here by the examination of anomalies (rather than trends), and the relatively dense observational network in the study area (Bengtsson et al. 2004).

Composites were created for four variables: geopotential height (at 1000 and 500 hPa), temperature (1000 hPa), wind vector (1000 and 500 hPa), and relative vorticity (500 hPa). Table 2 lists the hydroclimatological relevance of these variables, and thus justification for inclusion. Although important in river flow generation, precipitation has not been directly examined herein. As stated in section 1, precipitation has a complex and potentially indirect link to river flow at the monthly time scale (due to basin storage and transfer processes), further complicated in New England by snow- and melt-dominated flow regimes (Fig. 3). The well-documented considerable spatial and temporal variability in precipitation adds further difficulty for accurate quantification of climate–environment relationships over large geographic domains, compounded by measurement difficulties and assessment of their representativeness (e.g., Peterson et al. 1998). This is reflected in the poor reliability of ERA-40 precipitation data (Uppala et al. 2005). Furthermore, the more reliable merged gauge and satellite precipitation datasets do not cover the study period used herein. Given these issues, inclusion of precipitation is deemed beyond the scope of this paper, although this remains a goal for future research. The potential for precipitation occurrence is indicated by proxy through the other variables examined (Table 2).

Comparisons were made between the high and low flow composites for each month. The Student’s t test was used to determine the statistical significance of the differences between high and low flow situations on a grid cell basis. In addition to an examination of the standard deviation of the composites, use of the t test guards against one of the main criticisms of compositing, that it may result in the grouping of dissimilar situations, thus yielding an unrepresentative or misleading composite (Yarnal 1993; Houseago et al. 1998). This is because composites composed of widely different values will exhibit high standard deviation, and will thus be unlikely to be statistically significant from other composites when compared using a t test.

An issue relating to the use of t tests on gridded data is multiplicity, or the number of significant t tests that could be expected by chance. As outlined by Livezey and Chen (1983), testing for field significance involves determining 1) whether the number of locally significant t tests is greater than that which can be expected by chance (the 0.05 level is used here), and 2) the effect of spatial correlation between grid cells. Spatial correlation between grid cells reduces the effective sample size, and thus the threshold number of locally significant t tests required.

Taking into account 1) and then estimating 2) is considered sufficient for the purposes of this paper. Although Monte Carlo testing can be used to determine the precise effects of spatial correlation, this is not thought to be crucial here, and may even be slightly misleading. This is because while the number of positive individual significance tests that could be expected by chance is indicated, the probability of them occurring in a particular spatial pattern or at a particular location is not given (Taylor 1996). Field significance statistics are also highly dependent on the size of the area under investigation, as a higher number of significant t tests are needed to pass the threshold percentage of local significance when considering larger areas. These arguments are particularly pertinent here given that a key focus lies with differences over a relatively small area of the North Atlantic (New England), although within the context of a relatively large spatial domain. For such reasons, field significance will be used only as a guide to the interpretation of results, in association with previous work on climatic spatial correlation decay lengths (Briffa and Jones 1993), rather than a criterion for automatic acceptance or rejection of a particular field.

3. Study area

The annual flow regimes of the selected New England rivers consist of a marked peak between March and May, followed by minimum annual flows from July and September, before a secondary broad high flow period in late autumn and early winter, and slightly reduced flow in January and February (Fig. 3). The form and timing of the spring flow peak within the context of annual precipitation variation (Fig. 3) indicates the importance of snowfall storage and meltwater release in determining the annual hydrograph (Magilligan and Graber 1996; Hodgkins et al. 2003), and therefore the noteworthy influence of temperature–frozen water store interactions upon river flows.

The majority of study basins are heavily forested, with over 90% of the drainage area covered in many basins, and a mean for all basins of 85%. Basin sizes range from 74 to 997 km2. A mix of upland and lowland gauging stations was included in this study, with altitudes varying from 61 to 765 m. Table 1 provides further study basin characteristics. Averages for 1961–90 show that November is the month of highest precipitation in New England, with February the lowest, although precipitation is fairly evenly distributed throughout the year (Magilligan and Graber 1996) (Fig. 3). Mean monthly temperature for 1961–90 over the study region ranges from −6°C (January) to 20°C (July) (Fig. 3). This seasonal temperature pattern is due in part to New England’s location on the boundary between temperate extratropical and cold polar air masses, and the seasonal movement of this boundary. During winter, the polar front is frequently located to the south of the region, while milder and more humid conditions are experienced during the summer months following northward retreat of arctic air (Keim 2006).

4. Results

The years forming each composite and associated river flow statistics are presented in Table 3. A summary of the most salient climatic differences between high and low river flow are presented in Table 4, with further explanation below. Geopotential height at 1000 and 500 hPa for all months is greater (lower) in high (low) river flow situations over most of New England, with the area of maximum difference in the region just east of Newfoundland (Figs. 4, 5 and 6). The opposite geopotential height–river flow association occurs to the west of New England. A greater pressure gradient occurs over the region during low river flow, particularly in the winter half-year. Low flow during winter is also characterized by a more northerly orientation of isobars at the 1000-hPa level. In terms of magnitude of difference in geopotential height, there is a greater disparity between the high and low river flow composites in winter, although similar levels of statistical significance are achieved in summer months.

Statistically significant relative vorticity differences occur between high and low flow composites from December to May. Positive relative vorticity (i.e., cyclonic circulation) is observed in the vicinity of New England throughout. From December to April this is significantly stronger immediately east of New England in low flow composites (Fig. 7), with (reduced) statistical significance extending over New England in some cases. In May, the reverse vorticity–river flow association occurs, with more cyclonic circulation over New England in the high flow composite, but with maximum difference upwind of this region. No statistically significant differences occur from July to November.

Associated with the stronger pressure gradient, 1000- and 500-hPa wind speed is greater in the vicinity of New England for low river flow compared to high flow conditions. These differences in strength are primarily due to changes in zonal wind speed, with the meridional component much weaker. Variation in the direction of the meridional component has a more notable influence on wind vector, such that airflow over New England has a more northerly component under low river flow conditions and southerly component under high flow conditions (Figs. 4 and 6).

Changes in geopotential height, relative vorticity, and wind vector can be linked to variation in the position of atmospheric Rossby waves, and in particular the position of the trough over the east coast of North America. Under high flow conditions the trough occupies a more northwesterly position; it is further to the southwest during low flow periods (Figs. 4, 5 and 6). This indicates a more southerly position of the polar front for low river flow conditions.

Following the reduced northerly component to wind direction, more northwesterly positioned East Coast trough and more northerly polar front, air temperature is greater over New England in high river flow composites compared to low flow composites from December to April (Fig. 8). From December to February, this positive temperature anomaly is accompanied by a larger-scale pattern consisting of negative anomalies to the northwest and southwest (Fig. 8). In contrast, high river flow in May (and to a lesser extent June) is characterized by significantly lower temperature, with higher values in low flow composites. This pattern is associated with a greater land–sea temperature contrast in high flow situations. Temperature does not show any statistically significant differences between high and low river flow from July to November.

5. Discussion

a. Climatic processes linked to high and low river flow

The previous section detailed systematic differences in regional and Atlantic-scale climate between high and low flow situations for New England. Intra-annual variation in the relative importance of climatic factors was also revealed. The following discussion addresses potential large-scale climate mechanisms associated with the generation of high and low river flow.

The strong relationship between 1000-hPa temperature and river flow during winter is indicative of thermal control of 1) the snow:rain ratio, which affects whether precipitation influences river flow concurrently or if it enters storage as snow or ice (Hartley and Dingman 1993), and 2) subsequent storage release dynamics of the frozen water store. During spring, temperature becomes more important for the timing and magnitude of the annual melt peak for New England rivers (Fig. 3) (Hodgkins et al. 2003). This becomes particularly apparent when the latitudinal position of the 0°C isotherm is examined, which for March is located in the north of New England in the high flow composite, and in the south during low flow (Fig. 9).

The reversal of the temperature–river flow relationship in May and June can also be linked to snowmelt dynamics. Exploratory data analysis shows anomalous warmth in April precedes low flow in May and June, and has previously been linked to an earlier finish to the melt season (i.e., depletion of frozen water stores) (Hodgkins et al. 2005). Consequently, May and June experience reduced meltwater input to river flow. Huntington et al. (2003) show an analogous temporal dynamic for river–ice thickness variation in Maine. This hypothesis is further supported by an increased land–sea temperature contrast in May and June low flow composites: these show higher air temperatures over land rather than a cooler sea, consistent with the occurrence of a warm spring (given the greater thermal inertia of the sea). Increased evaporation may play a role in inverting temperature–river flow linkages, but is unlikely to be the sole cause as this relationship is not found in other (warmer) summer months, and similarly, because the relationship appears stronger in May rather than June.

The increased influence of cold, dry arctic air from the north in winter low river flow situations (as indicated by wind vector, geopotential height, and vorticity composites; Figs. 4 –7) corresponds well with the reduced temperatures also associated with low flow (Fig. 8). During summer, the more southerly winds for high flow indicate a greater maritime influence, or at least a reduced continental influence on the region. This can be associated with more humid conditions, lower evaporation, increased likelihood of precipitation (Keim et al. 2005; Keim 2006) and, in turn, increased river flow. Variation in the intensity of the zonal circulation also imparts an influence on the magnitude of monthly river flow such that intense westerly airflows, and thus a more continental influence, are associated with low river flow. In contrast, reduced westerly airflow may be associated with enhanced advection of moist maritime air over the region from the east and the occurrence of rainfall in summer and early autumn (Keim et al. 2005), driving higher river flows.

The more intense atmospheric circulation in low river flow composites is reflected in the pattern of regional storm tracks. There are two major paths followed by Atlantic cyclones along the east coast of the United States, either northward along the coast toward New England or following the northern limit of the Gulf Stream current eastward (Colucci 1976). Increased zonal wind speeds are associated with the latter (Serreze et al. 1997; Hartley and Keables 1998), resulting in reduced precipitation delivery over New England (Bradbury et al. 2003), and in turn, lower river flow. This is supported by the more cyclonic circulation centered east (i.e., downwind) of New England in low flow composites (Fig. 7). Additionally, the relatively low temperature during low flow conditions (Fig. 8) would make any precipitation more likely to fall as snow (as opposed to rain) and enter storage, rather than immediately contributing to river flow. It should be noted that when the temperature–flow relationship reverses in May and rainfall becomes the dominant form of precipitation, high river flow is associated with stronger cyclonic relative vorticity (upwind) than low flow.

b. NAO links to New England river flow

Cursory analysis suggests that climatic conditions over New England associated with high and low river flow support a positive relationship between the NAO and New England river flow. In agreement with the hypothesized process cascade outlined in Fig. 1, high flow has been shown to be associated with a weaker East Coast trough, a more northerly polar front, together with more southerly airflow (Figs. 4 –6) and generally higher temperatures (Fig. 8). Results are also suggestive of increased liquid-phase precipitation over the region in high river flow situations.

Closer inspection of the climate fields associated with high and low river flow reveal the influence of the NAO to be less straightforward. From December to March especially, the difference between the high and low flow geopotential height composites does not reveal a dipole between the Icelandic low and Azores high [the typical NAO signal (Hurrell 1995)], but rather a more longitudinal pattern associated with the strength and position of the Rossby wave pattern downstream of the Rocky Mountains (Fig. 5). Temperature differences between high and low river flow (Fig. 8) are also markedly different to the characteristic NAO–temperature pattern of positive anomalies over the northeast United States and northwest Europe and negative anomalies over southeast Canada and southern Europe (Hurrell 1995). Instead, the winter temperature anomalies associated with changes in river flow (Fig. 8) are highly suggestive of the SST tripole pattern associated with the NAO (Rodwell et al. 1999), with high (low) flow corresponding to the positive (negative) phase of this pattern. The SST tripole pattern has previously been shown to be associated with variation in Danube river flow in central Europe (Rimbu et al. 2002) and air temperature in northwest Europe (Junge and Stephenson 2003). Deser and Blackmon (1993) document a similar pattern as the first principal component of North Atlantic surface air temperature for November–March. Further research is needed to understand why air temperature displays the NAO–SST, rather than NAO–air temperature, anomaly pattern.

The lack of clarity in the influence of the NAO on river flow implies a complex process cascade linking atmospheric circulation and river flow variation in New England, and one that cannot satisfactorily be explained by a single atmospheric circulation pattern index. This complexity provides some context for the relatively weak NAO–river flow correlation found by Bradbury et al. (2002b). It also highlights the general limitations of relying on circulation pattern indices alone in investigations of climate–environment interactions, because despite their advantages they represent time-averaged dynamical relationships only, and as such may oversimplify or mask the mechanisms involved (Kingston et al. 2006). Hence it is vital that investigation of the process cascade linking climate and environmental variables (such as river flow) should be undertaken to explain mechanistic relationships before such indices are invoked as explanatory tools.

6. Conclusions

At the monthly time scale, river flow across New England has been shown to be sensitive to changes in the nature of large-scale climate fields over the northern North Atlantic, with coherent and physically interpretable patterns of variation emerging. In winter and early spring, geopotential height and temperature anomalies are likely to exert a key control on river flow through their effects on precipitation form and the occurrence of snowmelt. Differences in the wind field exist between high and low flow situations throughout the year, and indicate the importance of continental versus maritime air masses for river flow variation. The importance of using a monthly (versus seasonal average) temporal resolution is highlighted by the reversal of temperature and vorticity associations with river flow between April and May. These climatic differences indicate a limited positive association between the NAO and New England river flow but, more importantly, highlight the importance of developing an understanding of the processes linking large-scale climate and river flow variation as a physical basis for the use of circulation indices as descriptors of environmental variation.

Although we believe this paper contributes significantly to understanding the process cascade linking large-scale climate variation with high and low monthly river flow in New England, we recognize a number of caveats exist with respect to the findings. The precipitation link of this process cascade has not been explicitly analyzed; however, proxies for precipitation (Table 2) plausibly indicate its role in streamflow generation. The small sample size of the composites, the spatial and temporal resolution of the climatic data, and the averaging of discharge data across New England also need to be considered. Similarly, the influence of spatial gradients in climate (e.g., latitudinal or coast-to-interior) has not been explicitly addressed. Accordingly, the results are unlikely to describe variation in locally important controls on river flow generation. It must also be remembered that the statistical associations uncovered can only be used to infer cause-and-effect relationships. Nevertheless, the building-up of physically consistent statistical associations between river flow and a number of different climate variables does fundamentally underpin the conclusions drawn herein.

Future studies of New England climate–river flow relationships would benefit from direct analysis of the role of precipitation, although with careful consideration of the aforementioned potentially complex and indirect link between precipitation and river flow. Additional factors that may influence local hydroclimatological variation include SSTs, with the similarity of the river flow–temperature anomaly to the NAO–SST tripole pattern and the proximity of the study area to the Gulf Stream north wall and southern limit of the Labrador Current key candidates for further investigation. As well as concurrent relationships, antecedent land surface and SST conditions may be important at the monthly time scale (in addition to lower-frequency variations in ocean thermal state). System lags are likely to be particularly important for snowmelt dynamics, in terms of accumulation of snow, basin priming for initiation of the melt season and duration of the melt season (as illustrated by the inversion of temperature-river flow associations in May). This indicates the possibility for the development of predictive relationships for seasonal regional river flow using the climate fields examined herein (precipitation is relatively poorly predicted by seasonal forecast systems). These results therefore provide a valuable starting point for process-based investigations of this in New England and elsewhere, and are likely to be important for future studies focused on downscaling the impacts of variation in large-scale climate fields to river flow.

Acknowledgments

Daniel Kingston gratefully acknowledges receipt of a Ph.D. scholarship from the School of Geography, Earth, and Environmental Sciences, University of Birmingham, United Kingdom. Useful comments from three anonymous reviewers led to improvements in the manuscript. Anne Ankorn redrew Fig. 2.

REFERENCES

  • Anctil, F., , and Coulibaly P. , 2004: Wavelet analysis of the interannual variability in southern Quebec streamflow. J. Climate, 17 , 163173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Barros, V., , Chamorro L. , , Coronel G. , , and Baez J. , 2004: The major discharge events in the Paraguay River: Magnitudes, source regions, and climate forcings. J. Hydrometeor., 5 , 11611170.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., , Hodges K. I. , , and Hagemann S. , 2004: Sensitivity of the ERA40 reanalysis to the observing system: Determination of the global atmospheric circulation from reduced observations. Tellus, 56A , 456471.

    • Search Google Scholar
    • Export Citation
  • Beniston, M., , and Stephenson D. B. , 2004: Extreme climatic events and their evolution under changing climatic conditions. Global Planet. Change, 44 , 19.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradbury, J. A., , Dingman S. L. , , and Keim B. D. , 2002a: New England drought and relations with large scale atmospheric circulation patterns. J. Amer. Water Resour. Assoc., 38 , 12871299.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradbury, J. A., , Keim B. D. , , and Wake C. P. , 2002b: U.S. East Coast trough indices at 500 hPa and New England winter climate variability. J. Climate, 15 , 35093517.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradbury, J. A., , Keim B. D. , , and Wake C. P. , 2003: The influence of regional storm tracking and teleconnections on winter precipitation in the northeastern United States. Ann. Assoc. Amer. Geogr., 93 , 544556.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Briffa, K. R., , and Jones P. D. , 1993: Global surface air temperature variations during the twentieth century: Part 2: Implications for large-scale high-frequency palaeoclimatic studies. Holocene, 3 , 7788.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., 1976: Winter cyclone frequencies over the eastern United States and adjacent North Atlantic, 1964–1973. Bull. Amer. Meteor. Soc., 57 , 548553.

    • Search Google Scholar
    • Export Citation
  • Dai, A., , Fung I. Y. , , and DelGenio A. D. , 1997: Surface observed global land precipitation variations during 1900–88. J. Climate, 10 , 29432962.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Déry, S. J., , and Wood E. F. , 2004: Teleconnection between the Arctic Oscillation and Hudson Bay river discharge. Geophys. Res. Lett., 31 .L18205, doi:10.1029/2004GL020729.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , and Blackmon M. L. , 1993: Surface climate variations over the North Atlantic Ocean during winter—1900–1989. J. Climate, 6 , 17431753.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dudley, R. W., , and Hodgkins G. A. , 2002: Trends in streamflow, river ice and snowpack for coastal river basins in Maine during the 20th century. U.S. Geological Survey Water-Resources Investigations Rep. 02-4245, 26 pp.

  • Hannah, D. M., , and McGregor G. R. , 1997: Evaluating the impact of climate on snow- and ice-melt dynamics in the Taillon basin, French Pyrenees. J. Glaciol., 43 , 563568.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hannah, D. M., , Smith B. P. G. , , Gurnell A. M. , , and McGregor G. R. , 2000: An approach to hydrograph classification. Hydrol. Processes, 14 , 317338.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hartley, S., , and Dingman S. L. , 1993: Effects of climatic variability on winter–spring runoff in New England river basins. Phys. Geogr., 14 , 379393.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hartley, S., , and Keables M. J. , 1998: Synoptic associations of winter climate and snowfall variability in New England, USA, 1950-1992. Int. J. Climatol., 18 , 281298.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hodgkins, G. A., , Dudley R. W. , , and Huntington T. G. , 2003: Changes in the timing of high river flows in New England over the 20th Century. J. Hydrol., 278 , 244252.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hodgkins, G. A., , Dudley R. W. , , and Huntington T. G. , 2005: Summer low flows in New England during the 20th century. J. Amer. Water Resour. Assoc., 41 , 403412.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Houghton, J. T., , Ding Y. , , Griggs D. J. , , Noguer M. , , van der Linden P. J. , , Dai X. , , Maskell K. , , and Johnson C. A. , 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.

    • Search Google Scholar
    • Export Citation
  • Houseago, R. E., , McGregor G. R. , , King J. C. , , and Harangozo S. A. , 1998: Climate anomaly wave-train patterns linking southern low and high latitudes during South Pacific warm and cold events. Int. J. Climatol., 18 , 11811193.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huntington, T. G., , Hodgkins G. A. , , and Dudley R. W. , 2003: Historical trend in river ice thickness and coherence in hydroclimatological trends in Maine. Climatic Change, 61 , 217236.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., 1995: Decadal trends in the North-Atlantic Oscillation—Regional temperatures and precipitation. Science, 269 , 676679.

  • Junge, M. M., , and Stephenson D. B. , 2003: Mediated and direct effects of the North Atlantic Ocean on winter temperatures in northwest Europe. Int. J. Climatol., 23 , 245261.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Karl, T. R., , Williams C. N. Jr., , Quinlan F. T. , , and Boden T. A. , 1990: United States Historical Climatology Network (HCN) serial temperature and precipitation data. Environmental Science Division Publication 3404, Carbon Dioxide Information and Analysis Center, Oak Ridge National Laboratory, Oak Ridge, TN, 389 pp.

  • Keim, B. D., cited. 2006: A climate primer for New England. [Available online at http://airmap.unh.edu/background/ClimatePrimer.html.].

  • Keim, B. D., , Meeker L. D. , , and Slater J. F. , 2005: Manual synoptic classification for the East Coast of New England (USA) with an application to PM2.5 concentration. Climate Res., 28 , 143154.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kingston, D. G., , Lawler D. M. , , and McGregor G. R. , 2006: Linkages between atmospheric circulation, climate and streamflow in the northern North Atlantic: Research prospects. Prog. Phys. Geogr., 30 , 143174.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Krasovskaia, I., , Arnell N. , , and Gottschalk L. , 1994: Flow regimes in northern and western Europe: Development and application of procedures for classifying flow regimes. FRIEND: Flow Regimes from International Experimental and Network Data, P. Seuna, et al., Eds., IAHS Publication 221, 185–192.

  • Krepper, C. M., , García N. O. , , and Jones P. D. , 2003: Interannual variability in the Uruguay River basin. Int. J. Climatol., 23 , 103115.

  • Lawler, D. M., , McGregor G. R. , , and Phillips I. D. , 2003: Influence of atmospheric circulation changes and regional climate variability on river flow and suspended sediment fluxes in southern Iceland. Hydrol. Processes, 17 , 31953223.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lins, H. F., 1997: Regional streamflow regimes and hydroclimatology of the United States. Water Resour. Res., 33 , 16551667.

  • Lins, H. F., , and Slack J. R. , 1999: Streamflow trends in the United States. Geophys. Res. Lett., 26 , 227230.

  • Livezey, R. E., , and Chen W. Y. , 1983: Statistical field significance and its determination by Monte Carlo techniques. Mon. Wea. Rev., 111 , 4659.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Magilligan, F. J., , and Graber B. E. , 1996: Hydroclimatological and geomorphic controls on the timing and spatial variability of floods in New England, USA. J. Hydrol., 178 , 159180.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molnár, P., , and Ramírez J. A. , 2001: Recent trends in precipitation and streamflow in the Rio Puerco basin. J. Climate, 14 , 23172328.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Peterson, T. C., and Coauthors, 1998: Homogeneity adjustments of in situ atmospheric climate data: A review. Int. J. Climatol., 18 , 14931517.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Phillips, I. D., , McGregor G. R. , , Wilson C. J. , , Bower D. , , and Hannah D. M. , 2003: Regional climate and atmospheric circulation controls on the discharge of two British rivers, 1974–97. Theor. Appl. Climatol., 76 , 141164.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rimbu, N., , Boroneant C. , , Buta C. , , and Dima M. , 2002: Decadal variability of the Danube River flow in the lower basin and its relation with the North Atlantic Oscillation. Int. J. Climatol., 22 , 11691179.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., , Rowell D. P. , , and Folland C. K. , 1999: Oceanic forcing of the wintertime North Atlantic Oscillation and European climate. Nature, 398 , 320323.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Seaber, P. R., , Kapinos F. P. , , and Knapp G. L. , 1987: Hydrologic unit maps. USGS Water Supply Paper 2294, 63 pp.

  • Serreze, M. C., , Carse F. , , Barry R. G. , , and Rogers J. C. , 1997: Icelandic low cyclone activity: Climatological features, linkages with the NAO, and relationships with recent changes in the Northern Hemisphere circulation. J. Climate, 10 , 453464.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., , Bromwich D. H. , , Clark M. P. , , Etringer A. J. , , Zhang T. , , and Lammers R. , 2003: Large-scale hydro-climatology of the terrestrial Arctic drainage system. J. Geophys. Res., 108 .8160, doi:10.1029/2001JD000919.

    • Search Google Scholar
    • Export Citation
  • Slack, J. R., , and Landwehr J. M. , 1992: Hydro-Climatic Data Network: A U.S. Geological Survey streamflow data set for the United States for the study of climate variations, 1874-1988. USGS Open-File Rep. 92-129, 193 pp.

    • Crossref
    • Export Citation
  • Stenseth, N. C., , Ottersen G. , , Hurrell J. W. , , Mysterud A. , , Lima M. , , Chan K. S. , , Yoccoz N. G. , , and Adlandsvik B. , 2003: Studying climate effects on ecology through the use of climate indices: The North Atlantic Oscillation, El Nino Southern Oscillation and beyond. Proc. Roy. Soc. London, 270B , 20872096.

    • Search Google Scholar
    • Export Citation
  • Stewart, I. T., , Cayan D. R. , , and Dettinger M. D. , 2005: Changes toward earlier streamflow timing across western North America. J. Climate, 18 , 11361155.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Struglia, M. V., , Mariotti A. , , and Filograsso A. , 2004: River discharge into the Mediterranean Sea: Climatology and aspects of the observed variability. J. Climate, 17 , 47404751.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Taylor, A. H., 1996: North-south shifts of the Gulf Stream: Ocean-atmosphere interactions in the North Atlantic. Int. J. Climatol., 16 , 559583.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trigo, R. M., , Pozo-Vasquez D. , , Osborn T. J. , , Castro-Diez Y. , , Gamis-Fortez S. , , and Esteban-Parra M. J. , 2004: North Atlantic Oscillation influence on precipitation, river flow and water resources in the Iberian Peninsula. Int. J. Climatol., 24 , 925944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 re-analysis. Quart. J. Roy. Meteor. Soc., 131 , 29613012.

  • U.S. Water Resources Council, 1970: Water resources regions and subregions for the national assessment of water and related land resources. Washington DC, 188 pp.

  • Wettstein, J. J., , and Mearns L. O. , 2002: The influence of the North Atlantic–Arctic Oscillation on mean, variance, and extremes of temperature in the northeastern United States and Canada. J. Climate, 15 , 35863600.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yarnal, B., 1993: Synoptic Climatology in Environmental Analysis. Belhaven Press, 195 pp.

  • Yarnal, B., , and Leathers D. J. , 1988: Relationships between interdecadal and interannual climatic variations and their effect on Pennsylvania climate. Ann. Assoc. Amer. Geogr., 78 , 624641.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yarnal, B., , and Frakes B. , 1997: Using synoptic climatology to define representative discharge events. Int. J. Climatol., 17 , 323341.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ye, H., , Yang D. , , Zhang T. , , Zhang X. , , Ladochy S. , , and Ellison M. , 2004: The impact of climatic conditions on seasonal river discharge in Siberia. J. Hydrometeor., 5 , 286295.

    • Crossref
    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Hypothesized process chain linking the winter NAO and river flow in New England [modified from Kingston et al. (2006)].

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 2.
Fig. 2.

Location of New England and river flow gauging stations used.

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 3.
Fig. 3.

Mean (1958–2001) annual flow regime for New England rivers, and (1961–90) New England precipitation and temperature [derived from U.S. Historical Climatology Network (Karl et al. 1990)].

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 4.
Fig. 4.

Composite 1000-hPa geopotential height (m) and wind vector (m s−1) for February (top) high and (bottom) low flow. [Length of arrow indicates wind strength (see scale)]

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 5.
Fig. 5.

Difference (in m) between February 1000-hPa geopotential height composites (high minus low flow composite). [Shading indicates significant difference between the composites (with 8 degrees of freedom); p levels are (from light to dark shades) 0.05, 0.01, and 0.001]

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 6.
Fig. 6.

Composite 500-hPa geopotential height (m) and wind vector (m s−1) for February (top) high and (bottom) low flow. [Length of arrow indicates wind strength (see scale)]

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 7.
Fig. 7.

Difference (in s−1) between December 500-hPa relative vorticity composites (high minus low flow composite). (Shading as in Fig. 5)

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 8.
Fig. 8.

Difference (in °C) between February 1000-hPa temperature composites (high minus low flow composite). (Shading as in Fig. 5)

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Fig. 9.
Fig. 9.

Composite 1000-hPa temperature (in °C) for March (left) high and (right) low flow.

Citation: Journal of Hydrometeorology 8, 3; 10.1175/JHM584.1

Table 1.

Gauging stations used and basin characteristics (after Slack and Landwehr 1992).

Table 1.
Table 2.

Variables used and their hydroclimatological significance.

Table 2.
Table 3.

List of years making up the high and low streamflow composites for each month, and the mean and standard deviation of streamflow in each composite (units of discharge are z scores).

Table 3.
Table 4.

Summary of relationship between river flow and climate.

Table 4.
Save