• Atlas, R., N. Wolfson, and J. Terry, 1993: The effect of SST and soil moisture on GLA model simulations of the 1988 U.S. summer drought. J. Climate,6, 2034–2048.

  • Black, R. X., 1997: Deducing anomalous wave source regions during the life cycles of persistent flow anomalies. J. Atmos. Sci.,54, 895–907.

  • Chang, F. C., and J. M. Wallace, 1987: Meteorological conditions during heat waves and droughts in the United States Great Plains. Mon. Wea. Rev.,115, 1253–1269.

  • Ferranti, L., T. N. Palmer, F. Molteni, and E. Klinker, 1990: Tropical–extratropical interaction associated with the 30–60 day oscillation and its impact on medium and extended range prediction. J. Atmos. Sci.,47, 2177–2199.

  • Gruber, A., and A. F. Krueger, 1984: The status of the NOAA outgoing longwave radiation data set. Bull. Amer. Meteor. Soc.,65, 958–962.

  • Hoskins, B. J., and T. Ambrizzi, 1993: Rossby wave propagation on a realistic longitudinally varying flow. J. Atmos. Sci.,50, 1661–1671.

  • Janowiak, J. E., 1988: The global climate for March–May 1988: The end of the 1986–87 Pacific warm episode and the onset of widespread drought in the United States. J. Climate,1, 1019–1040.

  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Karoly, D. J., R. A. Plumb, and M. Ting, 1989: Examples of horizontal propagation of quasi-stationary waves. J. Atmos. Sci.,46, 2802–2811.

  • Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc.,77, 1275–1277.

  • Lyon, B., and R. M. Dole, 1995: A diagnostic comparison of the 1980 and 1988 U.S. summer heat wave–droughts. J. Climate,8, 1658–1675.

  • Mo, K. C., J. R. Zimmerman, E. Kalnay, and M. Kanamitsu, 1991: A GCM study of the 1988 United States drought. Mon. Wea. Rev.,119, 1512–1532.

  • Namias, J., 1991: Spring and summer 1988 drought over the contiguous United States—Causes and prediction. J. Climate,4, 54–65.

  • Newman, M., and P. D. Sardeshmukh, 1998: The impact of the annual cycle on the North Pacific/North American response to remote low-frequency forcing. J. Atmos. Sci.,55, 1336–1353.

  • Nitta, T., 1987: Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation. J. Meteor. Soc. Japan,65, 373–390.

  • NOAA, 1988: Drought advisory 88/12. Summary of conditions and impacts. [Available from Climate Prediction Center, NOAA, World Weather Building, Washington, DC 20233.].

  • Palmer, T. N., and C. Brankovic, 1989: The 1988 United States drought linked to anomalous sea surface temperature. Nature,338, 54–57.

  • Plumb, R. A., 1985: On the three-dimensional propagation of stationary waves. J. Atmos. Sci.,42, 217–229.

  • Rasmusson, E. M., 1968: Atmospheric water vapor transport and the water balance of North America. Part II. Large-scale water balance investigations. Mon. Wea. Rev.,96, 720–734.

  • Riebsame, W. E., S. A. Changnon Jr., and T. R. Karl, 1991: Drought and Natural Resources Management in the United States—Impacts and Implications of the 1987–1989 Drought. Westview Press, 174 pp.

  • Ropelewski, C. F., 1988: The global climate for June–August 1988:A swing to the positive phase of the Southern Oscillation, drought in the United States, and abundant rain in the monsoon areas. J. Climate,1, 1153–1174.

  • Sardeshmukh, P. D., and B. J. Hoskins, 1988: The generation of global rotational flow by steady idealized tropical divergence. J. Atmos. Sci.,45, 1228–1251.

  • Shukla, J., and Y. Mintz, 1982: The influence of land-surface evaporation on the earth’s climate. Science,215, 1498–1501.

  • Spencer, R. W., 1993: Global oceanic precipitation from the MSU during 1979–91 and comparisons to other climatologies. J. Climate,6, 1301–1326.

  • Trenberth, K. E., and G. W. Branstator, 1992: Issues in establishing causes of the 1988 drought over North America. J. Climate,5, 159–172.

  • ——, ——, and P. A. Arkin, 1988: Origins of the 1988 North American drought. Science,242, 1640–1645.

  • Wolfson, N., R. Atlas, and Y. C. Sud, 1987: Numerical experiments related to the summer 1980 U.S. heat wave. Mon. Wea. Rev.,115, 1346–1357.

  • View in gallery

    Histogram of monthly precipitation anomalies, determined as departures from monthly means of the years 1950–92. Units are in in. month−1. The monthly mean value is printed as “climo.” The value for 1988 is printed atop the appropriate column. Precipitation is averaged in a region bounded by 34°–49°N, 81°–101°W.

  • View in gallery

    AMJ mean and April, May, and June monthly mean streamfunction anomalies at 300 mb. The contour interval is 2.5 × 106 m2 s−1. The zero contour is omitted for clarity; negative values are shaded. In the top panel, the two boxes bound the areas used for the projection coefficient time series displayed in Fig. 3, with the inner box indicating region 1 and the outer box indicating region 2.

  • View in gallery

    Evolution of the projection coefficient. The coefficient is determined by projecting the observed 300-mb 10-day low-pass streamfunction field upon the AMJ-mean streamfunction anomaly, in either region 1 (thin line) or region 2 (thick line).

  • View in gallery

    Time–longitude plot of observed 300-mb 10-day low-pass streamfunction anomalies averaged over the latitude belt from 30° to 60°N for 1 April 1988–30 June 1988. The contour interval is 4 × 106 m2 s−1. The zero contour is omitted for clarity; negative values are shaded and represented with thinner lines.

  • View in gallery

    Time vs latitude plot of observed 10-day low-pass OLR anomalies for the period 1 April–30 June 1988. The contour interval is 10 W m−2. The zero contour is omitted for clarity; negative values are shaded.

  • View in gallery

    Observed 10-day low-pass streamfunction anomalies at 300 mb, spaced three days apart during the period of initial growth of the early June anticyclone. The contour interval is 5 × 106 m2 s−1. The zero contour is omitted for clarity; negative values are shaded.

  • View in gallery

    Observed 10-day low-pass OLR and residual forcing anomalies at 300 mb, spaced three days apart. The forcing anomalies are indicated by contours; contour interval is 1.5 × 10−10 s−1. The zero contour is omitted for clarity; negative values are indicated by dashed contours. The OLR anomalies (W m−2) are indicated by shading.

  • View in gallery

    Modeled streamfunction anomalies at 300 mb on 1 June 1988, for model runs described in text. (a) Forced with global observed residual forcing. (b) No forcing. Plotting conventions are the same as Fig. 6.

  • View in gallery

    Modeled streamfunction anomalies at 300 mb, spaced three days apart, during the period of initial growth of the early June anticyclone. Forcing is applied only between 90°E and 180. Plotting conventions are the same as Fig. 6.

  • View in gallery

    Plumb flux and streamfunction anomalies at 300 mb averaged over the period 26 May–1 June 1988. (a) Observed. (b) Model results (corresponding to the model run shown in Fig. 9). Plumb flux vectors with a value less than 3 m2 s−2 are not shown; the maximum vector is 45 m2s−2. The zero contour is omitted for clarity; negative values are represented with dashed lines.

  • View in gallery

    Modeled streamfunction anomalies at 300 mb on 1 June 1988, for model runs similar to that shown in Fig. 9 except using different base states described in text. (a) Base state is one month earlier. (b) Base state is one month later. Plotting conventions are the same as Fig. 6.

  • View in gallery

    Observed (left) and modeled (right) 10-day low-pass streamfunction anomalies at 300 mb, spaced three days apart, during the period of initial growth of the late June anticyclone. Forcing is applied only between 90°E and 180°. Plotting conventions are the same as Fig. 6.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 271 269 6
PDF Downloads 112 112 6

Rossby Wave Propagation and the Rapid Development of Upper-Level Anomalous Anticyclones during the 1988 U.S. Drought

View More View Less
  • 1 NOAA/CIRES Climate Diagnostics Center, University of Colorado, Boulder, Colorado
© Get Permissions
Full access

Abstract

The upper-tropospheric circulation is investigated for the three months of April, May, and June 1988 during which the Great Plains region of the United States experienced one of its most severe droughts in history. It is found that during this period the April–June (AMJ) seasonal-mean anomaly was not representative of the variability of 10-day low-pass anomalies. Rather, over North America large fluctuations on monthly and shorter timescales occurred, with the dominant streamfunction anomalies not strongly anticyclonic until June. In fact, the AMJ anomaly was dominated by two episodes of rapidly developing, intense anomalous anticyclones in early and late June.

Examination of the daily 10-day low-pass streamfunction anomalies at 300 mb suggests that propagating Rossby waves originating in the west Pacific played a dominant role in the initiation of these intense anomalous anticyclones. Numerical experiments with a linear, time-dependent, barotropic model also support this hypothesis. These results suggest that the AMJ anomaly, which has been characterized as a wave train seemingly forced in the east Pacific, may not provide a useful picture of the circulation associated with the drought. Instead, the drought may be better studied not as a single seasonal event, but rather as a succession of events that together produced a serious hydrological deficit.

Corresponding author address: Dr. Matthew Newman, CIRES, University of Colorado, Campus Box 449, Boulder, CO 80309–0449.

Email: men@cdc.noaa.gov

Abstract

The upper-tropospheric circulation is investigated for the three months of April, May, and June 1988 during which the Great Plains region of the United States experienced one of its most severe droughts in history. It is found that during this period the April–June (AMJ) seasonal-mean anomaly was not representative of the variability of 10-day low-pass anomalies. Rather, over North America large fluctuations on monthly and shorter timescales occurred, with the dominant streamfunction anomalies not strongly anticyclonic until June. In fact, the AMJ anomaly was dominated by two episodes of rapidly developing, intense anomalous anticyclones in early and late June.

Examination of the daily 10-day low-pass streamfunction anomalies at 300 mb suggests that propagating Rossby waves originating in the west Pacific played a dominant role in the initiation of these intense anomalous anticyclones. Numerical experiments with a linear, time-dependent, barotropic model also support this hypothesis. These results suggest that the AMJ anomaly, which has been characterized as a wave train seemingly forced in the east Pacific, may not provide a useful picture of the circulation associated with the drought. Instead, the drought may be better studied not as a single seasonal event, but rather as a succession of events that together produced a serious hydrological deficit.

Corresponding author address: Dr. Matthew Newman, CIRES, University of Colorado, Campus Box 449, Boulder, CO 80309–0449.

Email: men@cdc.noaa.gov

1. Introduction

Much of the contiguous United States experienced severe dry conditions during the spring and summer of 1988. Overall damage to the society and environment made the 1988 drought one of the nation’s greatest disasters of the twentieth century (NOAA 1988).

In some regions, the 1988 drought actually started in late 1987 and peaked in the summer of 1988 (e.g., Riebsame et al. 1991). In the Great Plains, anomalously low precipitation existed throughout the spring and early summer. This can be seen in Fig. 1, which shows the distribution of monthly precipitation anomalies for the period 1950–92, averaged in the Great Plains region. The value of the anomaly (in inches) for each month of 1988 is also indicated. The precipitation deficit was greatest in late spring; both May and June had their lowest precipitation in over 40 years. Large-scale anomalous atmospheric and oceanic circulations were observed prior to and during the peak of the drought (e.g. Janowiak 1988; Ropelewski 1988; Trenberth et al. 1988;Trenberth and Branstator 1992). In the atmosphere very strong anomalous upper-level anticyclones appeared over the central North Pacific and the Great Plains. In the ocean colder than normal sea surface temperature (SST) occurred in the eastern tropical Pacific and central north Pacific.

The causes of the drought have been a subject of debate. Trenberth et al. (1988) suggested that the primary cause of the drought was the strong anomalous upper-level anticyclone over the Great Plains in the April–June-mean (AMJ-mean) map. They further claimed that this anomalous upper-level anticyclone was part of a stationary wave train forced by steady atmospheric heating associated with the SST anomalies in the eastern tropical Pacific. Their conclusion was based on a steady-state, linear planetary wave atmospheric model simulation of the observed wave train in the 300-mb AMJ-mean geopotential height anomaly map.

Namias (1991), on the other hand, argued that the initiation of the drought was rooted in extratropical climate variations of the ocean and atmosphere, for the dry conditions appeared in the spring of 1988 as early as March before the significant drop of sea surface temperature in the eastern tropical Pacific in May. He also showed that past summer droughts in the Great Plains had little direct relation to negative SST anomalies in the eastern tropical Pacific. He further postulated that interactions with dry soil that developed in the plains during April and May probably played a role in strengthening the anticyclone aloft. In agreement with the results of Namias (1991), Lyon and Dole (1995) found that during spring of 1988 stationary waves emanated from two apparent source regions, one over the central North Pacific to the north of the Hawaiian Islands and a second over the far western Pacific. They found no significant anomalous wave activity propagating out of the eastern tropical Pacific. The anomalous stationary wave activity flux becomes very weak by early July. They concluded that remote forcing played a predominant role in initiating the drought, but that anomalous local boundary conditions played a more significant role in the drought’s persistence.

General circulation model studies of the 1988 U.S. drought are also inconclusive. Palmer and Brankovic (1989) claimed that the SST anomalies in the eastern tropical Pacific were important for the development of the drought. Mo et al. (1991) demonstrated, on the other hand, that the SST anomalies in the eastern tropical Pacific only played a minor role. They emphasized that atmospheric conditions in May were critical to the development of the strong anomalous anticyclone in June. Atlas et al. (1993) stressed the importance of anomalous North American soil moisture in the development of the drought.

Severe summer droughts in the central United States are historically accompanied by large-scale anomalous upper-level anticyclones (e.g., Chang and Wallace 1987). Strong upper-level anticyclones can make the surface drier by steering baroclinic activity onto a more poleward track. Dry surface conditions, on the other hand, could reduce evapotranspiration, increase surface temperatures, and, therefore, induce and/or strengthen upper-level anticyclones. Reduced evapotranspiration also tends to suppress precipitation over the midlatitude land surfaces due to the reduced local water vapor source (e.g., Rasmusson 1968). It is thus possible that positive feedbacks between the land surface processes and the dynamics of the atmosphere can make an existing drought self-perpetuating (e.g., Shukla and Mintz 1982; Wolfson et al. 1987; Atlas et al. 1993).

Understanding the cause of the large-scale upper-tropospheric circulation anomalies is thus crucial to understanding and forecasting summer droughts in the Great Plains. Trenberth et al. (1988) emphasized the role of the AMJ-averaged upper-tropospheric circulation anomalies in causing the 1988 U.S. drought. Lyon and Dole (1995) suggested, however, that this seasonal mean does not adequately represent the state of the atmosphere for the three months. Theoretical calculations (Newman and Sardeshmukh 1998) also show that rapid changes in the climatological wind field during spring could make the atmosphere sensitive to a much different pattern of anomalous forcing in early spring than late spring, and that circulation anomalies over North America are more sensitive to anomalous forcing in the west Pacific during late spring than at any other time of year. Thus, steady remote forcing might not produce a persistent springtime anomaly over North America.

In this paper, the large-scale upper-tropospheric circulation anomalies associated with the 1988 drought will be examined using the observed 10-day lowpass streamfunction anomalies at 300 mb. The cause of the circulation anomalies over North America will be inferred from the evolution of the global circulation anomalies and outgoing longwave radiation (OLR) anomalies. It will be further diagnosed with a linear, time-dependent barotropic model applied at 300 mb.

2. Observational results

The primary data used in this study are the 17-yr (1979–95) daily analyzed vorticity and divergence fields at 300 mb from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) Reanalysis (Kalnay et al. 1996) and the daily interpolated OLR (Gruber and Krueger 1984; Liebmann and Smith 1996). We will be analyzing the variability of the streamfunction and OLR fields during the spring of 1988 at both monthly and shorter timescales. Monthly or seasonal anomalies are defined as the departure from the 17-yr (1979–95) monthly or seasonal climatology. Low-frequency anomalies are computed as follows. The original daily data were truncated to T21 resolution and harmonic analysis was performed for each year. The annually varying climatology is then determined as the 17-yr average of the annual mean plus the first three harmonics. Daily anomalies are defined as the departure of the original daily data from this climatology. A 61-point 10-day low-pass Lanczos filter is applied to the daily anomalies to filter the high-frequency eddies.

Figure 2 shows maps of the AMJ-mean and the April, May, and June monthly mean streamfunction anomalies at 300 mb. In the AMJ-mean map, the anomalous circulation was anticyclonic over Canada and much of the United States. This anomalous anticyclone looks like part of a stationary wave train extending from the central North Pacific to the Caribbean [see also Fig. 5 in Trenberth et al. (1988) and Fig. 2 in Lyon and Dole (1995)]. Trenberth et al. (1988) suggested that this stationary wave train was forced primarily by steady heating anomalies due to a northward shift in the intertropical convergence zone (ITCZ) brought about by the SST anomalies observed in the east tropical Pacific.

However, the individual April, May, and June monthly mean anomalies differ considerably from the AMJ-mean anomalies as well as from one another, as previously noted by Lyon and Dole (1995) for the geopotential height anomalies at 300 mb. In fact, a comparison of the AMJ-mean anomaly with the individual monthly means shows that different features of the wave train in the AMJ-mean map appear to occur during different months. Most notably, the anticyclones over central Canada and the Caribbean are primarily due to the June anomaly, whereas the cyclones over the southeast United States and the northeast Pacific occur predominantly during May. Only the central Pacific anticyclone shows some persistence from month to month. This suggests that interpreting the AMJ-mean anomaly as a steady stationary wave train could be problematic and misleading.

The daily 10-day low-pass streamfunction field at 300 mb is projected onto the AMJ-mean streamfunction anomaly field to examine further the steadiness of the wave train over North America in Fig. 2a. A projection coefficient p0 is defined as
i1520-0442-11-10-2491-e1
where ψ′ denotes the daily 10-day low-pass streamfunction and ψ0 the AMJ-mean streamfunction anomaly. It is clear from Eq. (1) that p0 is related to pattern correlation between ψ′ and ψ0 over the area λ1 < λ < λ2 and ϕ1 < ϕ < ϕ2. A projection coefficient around unity means that the amplitude and spatial patterns of ψ′ and ψ0 are very similar. A positive p0 much larger than unity indicates that the low-pass anomaly is similar to that of the AMJ mean, but much more intense. Similarly, a large negative p0 suggests that the low-pass anomaly is nearly opposite to that in the AMJ-mean map, that is, the anomalous circulation over central North America is cyclonic.

Figure 3 shows the evolution of the projection coefficient, p0, within each of the two regions indicated by the boxes in Fig. 2a from 1 April to 30 June 1988. Region 1 (the small box in Fig. 2a, bounded by λ1 = 120°W, λ2 = 75°W, ϕ1 = 35°N and ϕ2 = 75°N) focuses on the variability of the anomalous anticyclone over the Great Plains. The larger region 2 (the big box in Fig. 2a, bounded by λ1 = 130°W, λ2 = 60°W, ϕ1 = 15°N and ϕ2 = 75°N) encompasses the variability of the wave train over North America. The projection coefficient, and, therefore, the anomalous circulation over North America, exhibited large fluctuations on monthly and shorter timescales during the three months of April, May, and June of 1988. Over the Great Plains the anomalous circulation in April and most of May was very different from that in June. During much of April and May the anomalous circulation over the Great Plains was either cyclonic or moderately anticyclonic. Positive values of p0 during this time are associated with a pronounced cyclone over the southeast United States. From late May, the anomalous circulation in region 1 rapidly changed from weakly cyclonic to strongly anticyclonic. The intensity of this anomalous anticyclone reached its maximum in early June and decreased dramatically afterward. Around 20 June a second strong anomalous anticyclone intensified, remaining through early July. Clearly, the anomalous anticyclone over the Great Plains in the AMJ-mean map (Fig. 2a) was primarily due to the two very strong anomalous anticyclonic events in the first 10 days and last 10 days of June.

Figure 4 shows a time–longitude plot of the low-pass streamfunction anomalies averaged between 30° and 60°N to illustrate the temporal variability of the circulation in other longitudes. Over North America (e.g., longitudes 60°–135°W) the circulation exhibited substantial intraseasonal variation throughout the spring, particularly during May. The two strong anomalous anticyclones in early and late June are evident. In April and May, not only were there no persistent anomalous anticyclones over North America, but a pronounced negative streamfunction anomaly existed during mid-May. There is an appearance of energy propagation across the Pacific in the first half of May and in early June, culminating in large anomalies over North America. It is also interesting to note that the zonal mean anomaly is negative during much of May and positive throughout most of June.

The observed OLR anomalies also exhibited significant intraseasonal variation during the three months of April, May, and June 1988. Figure 5 shows time–latitude plots of 10-day low-pass OLR anomalies over the east tropical Pacific (averaged between 120° and 140°W) and the west tropical Pacific (averaged between 120° and140°E). The ITCZ in the east tropical Pacific shifted northward in mid-May by about 5° of latitude. However, the OLR anomaly in the ITCZ region was not steady. Also, it peaked in mid-May (at all longitudes, not shown), well before the development of the prominent North American anticyclones during June. Over the west Pacific, the OLR anomalies are of generally larger amplitude. Throughout May and early June, there is a suggestion of some periodicity north of the equator, with a timescale of about 15 days, which appears to be related to pronounced intraseasonal oscillations in equatorial tropical convection between about 60°E and 180° at this time (not shown). Note that a very strong OLR anomaly appeared over the west tropical Pacific around May 25 and lasted for about two weeks. This feature occurred about a week prior to the time that the anomalous anticyclone over North America reached its maximum. The unsteadiness of the OLR anomalies over the east tropical Pacific shown in the top panel of Fig. 5 implies that it is unlikely that the associated anomalous atmospheric heating is the primary forcing of the stationary wave train in the AMJ-mean anomalous circulation map.

The strong pulse of OLR anomaly in late May over the west Pacific, occurring just prior to the rapid development of the anomalous anticyclone in late May and early June over the Great Plains, suggests the possible importance of cross-Pacific propagation of Rossby waves forced by heating anomalies over the west Pacific. Additional evidence is obtained from synoptic maps of low-pass streamfunction anomalies in late May and early June. Figure 6 shows the 10-day low-pass streamfunction anomalies on 23, 26, 29 May and 1 June 1988. On 23 May the streamfunction anomaly was positive over Southeast Asia and quite weak over the North Pacific and North America. In the following days, the anomalous high over Southeast Asia weakened and a strong anomalous anticyclone developed over the central North Pacific while the streamfunction anomaly remained negative over North America. By 26 May the series of centers extends almost zonally across the Pacific and North America. From that date onward, an anomalous anticyclone developed very rapidly over North America and the anomalous anticyclone over the central North Pacific weakened. The positive streamfunction anomaly over the Great Plains reached its maximum in early June and started to decrease. By 12 June (not shown) there was no significant streamfunction anomaly over the United States. The second anticyclone appears to have developed in a similar manner, with a center developing first over Southeast Asia around 10 June, over the central Pacific around 15 June, and finally over the Great Plains around 20 June, lasting into the early part of July.

3. Model results

The evolution of the 10-day low-pass streamfunction anomalies in Fig. 6 suggests that Rossby waves originating in Southeast Asia and the west Pacific may have played an important role in the rapid development of the anomalous anticyclones over North America in late May and again in mid-June 1988. To test this hypothesis, barotropic model runs were carried out in which a Rossby wavemaker, with temporal and spatial scales similar to those observed, was located over the west Pacific. The development of the modeled wave train over North America can then be compared to observations.

The barotropic model, including a “free surface correction” that parameterizes some baroclinic processes (Ferranti et al. 1990), takes the form
i1520-0442-11-10-2491-e2
where overbars refer to annually varying climatological means, primes refer to anomalies, ζ is the absolute vorticity, v is the total wind, vψ is the rotational wind, τD is the linear damping timescale, KH is the coefficient of a weak scale-dependent biharmonic diffusion, R is the Rossby radius of deformation (1/4 of the earth’s radius), and F′ is the anomalous forcing. Note that F′ represents the combined effects of all the terms neglected in the linear barotropic model, such as baroclinicity, nonlinearity, boundary effects including topography, and high-frequency eddy feedbacks. For the results reported in this paper, τD = 10 days and KH is chosen so that it produces a damping with a timescale of 4 days at the highest wavenumber. Since this is a linear system, the spatial patterns of modeled streamfunction anomalies shown below are not sensitive to the damping rate, but increased damping does reduce the amplitude. Equation (2) is solved numerically using the spectral transform technique with a triangular truncation of maximum zonal wavenumber 42 (T42 truncation). The model is applied at the 300-mb level, which is generally considered a good compromise between the equivalent barotropic level and the level of maximum forcing (Hoskins and Ambrizzi 1993). During the course of a model integration the climatological base state and anomalous forcing continuously evolve according to observations.

If F′ vanishes, Eq. (2) describes the linear dynamics of free barotropic Rossby waves propagating on a zonally and meridionally varying climatological ambient flow. Similarly, when F′ is specified, (2) describes the linear dynamics of forced barotropic Rossby waves. Conversely, we could calculate F′ as the residual of the left-hand-side terms evaluated using the observed reanalysis data. We have done so for the spring of 1988. Maps of this residual “forcing” for 26 and 29 May and 1 June 1988 are shown in Fig. 7. Also included in the figure are the observed OLR lowpass anomalies, denoted by gray-scale shading. The most interesting feature in Fig. 7 is that the residual forcing over the west Pacific is much larger than that over North America, although the streamfunction anomalies are much larger over the east Pacific–North America region (see Figs. 6b and 6c). In the OLR field, there is a belt of strong negative OLR anomaly in the west Pacific, but not in the east Pacific. The anomalous convection over the Philippines peaked around 30 May and was associated with regions of suppressed convection to both the north and south. This banded structure in anomalous convection appears similar to the “Pacific–Japan” pattern identified in monthly mean cloud-amount data for June–August, which may be related to La Niña (Nitta 1987). The band of enhanced OLR also extends northeastward, coincident with a large precipitation anomaly in the Microwave Sounding Unit (Spencer 1993) precipitation data (not shown). In the east equatorial Pacific there are only very weak OLR anomalies.

OLR anomalies are often considered proxies of anomalous heating, and it would be useful to gauge the impact of these west Pacific heating anomalies upon the development of the observed North American circulation anomaly. A natural method would be to force the barotropic model with the Rossby wave source (Sardeshmukh and Hoskins 1988) determined from observed anomalous divergence. This can pose a practical problem, however, because whereas in the Tropics divergence anomalies may be directly related to heating anomalies, in the extratropics divergence is often a diagnostic quantity that arises through the quasigeostrophic omega equation. Also, heating anomalies will affect other terms that comprise the residual forcing in (2), such as the nonlinear transient eddy flux term. Thus, there is no easy way to precisely quantify the contribution of the heating anomalies to the forcing of the barotropic model. For the purpose of the current paper we will not attempt to distinguish between sources of forcing, since we are merely interested in generating a realistic Rossby wave train in the Eastern Hemisphere and then determining whether it propagates with significant amplitude into the North America region. Thus, it is sufficient to treat the residual forcing as a “black box” of forcing for the 10-day low-pass eddies.

As a consistency check on our calculations, we first ran the model forced by the “exact” global residual forcing. The model was initialized with the observed streamfunction anomaly on 23 May and then integrated for 9 days from 23 May to 1 June. Figure 8a shows the modeled streamfunction anomaly on 1 June, which compares very well with the observed anomaly (bottom panel of Fig. 6). Minor differences between the two can be attributed mainly to evaluating the residual forcing from daily observations and then linearly interpolating this forcing to each time step in the model. Overall, the observed circulation anomaly is reasonably well reproduced when forcing the model with the complete observed residual forcing field.

Figure 8b shows model anomalous streamfunction on 1 June in an experiment in which the residual forcing is everywhere set to zero. The streamfunction anomaly in Fig. 8b is much smaller than in Fig. 8a, although the patterns in the two figures are somewhat similar. Thus, the wave-train over the Pacific and North America in Fig. 8a is not due to normal-mode instability of the climatological base state, but rather to the anomalous forcing.

To produce a Rossby wavemaker over Southeast Asia and the west Pacific, the model is forced by residual forcing between 90°E and 180 only. The initial condition is again the global streamfunction anomaly on 23 May, but results are virtually the same if the initial condition is also restricted to the 90°E–180° region. Results are shown in Fig. 9 for model streamfunction anomalies on 23, 26, 29 May and 1 June. Comparison with Fig. 6 shows that the model captures the gross features of the wave train over the North Pacific–North America region. The primary discrepancy is that the length scale of the modeled wave train is somewhat longer than is observed, which results in a pattern correlation of 0.3 in region 1 between the modeled and observed anomaly. Thus, by 1 June the center of the main anticyclone in the model (Fig. 9d) is east of the observed anticyclone (Fig. 6d) by about 15° longitude. This difference in the location of the anomalous anticyclone is due to the neglect of the residual forcing over North America, which acts to retard the propagation of Rossby waves in this region. Indeed, it can be deduced from Fig. 7b that the residual forcing over North America acts to force an anomalous anticyclone over the western half of the continent and an anomalous cyclone over the eastern half of the continent, as might be expected of the ignored baroclinic processes.

The model results can also be compared with the observed development of the early June anticyclone by using the Plumb flux. Figure 10 shows the Plumb flux computed from the anomalous stationary wave averaged between 26 May and 1 June, superimposed upon the averaged anomalous streamfunction [the formula for FSA of Black (1997) was used to calculate the anomalous Plumb flux in this figure]. The Plumb flux has been used as a diagnostic of the energy propagation of stationary Rossby waves (e.g., Plumb 1985; Karoly et al. 1989; Lyon and Dole 1995; Black 1997). Thus, Fig. 10 could be interpreted as indicating that much of the observed cross-Pacific wave energy flux (top panel) is well captured by the model (bottom panel). Of course, it should be noted that there is ambiguity inherent in the interpretation of the Plumb flux. For example, multiple sources of wave activity will in general not produce a pattern of Plumb flux that is the same as the sum of the Plumb flux due to each individual source. Nevertheless, to the extent that the Plumb flux is representative of the total wave field, these figures demonstrate the similarity between the model results and the observed early June episode.

The model-simulated rapid development of the anomalous anticyclone over North America was not sensitive to the initial starting date. Successful simulations were also made in experiments with the model initialized a few days either before or after 23 May 1988. The model results were more sensitive to the date of the annual cycle of the base state. To demonstrate this latter point, we repeated the model run shown in Fig. 9, using the same initial condition and evolving forcing but with a base state that started 23 April. As before, the base state evolved daily. A second experiment was conducted for which the base state started 23 June. Results on 1 June (i.e., on day 9 of the simulation) are shown in Fig. 11 and can be contrasted with the last panel of Fig. 9. It is clear that the largest response occurred for the late May base state and that the pattern of the response also changed. This is consistent with the results of Newman and Sardeshmukh (1998), who showed that rapid changes in the climatological wind field during spring could make the atmosphere sensitive to a different pattern of anomalous forcing each month. Finally, we found that the importance of the temporal evolution of the forcing arose at least in part because forcing that produced a strong North American response did not persist. In fact, a stronger response was produced for a model run where we repeated the experiment of Fig. 9, but held the forcing steady with its 25 May pattern.

The development of the late June anticyclone was also investigated using the model. Figure 12 shows the observed and modeled streamfunction anomalies on 15, 18, 21, and 24 June. Again, observed residual forcing between 90°E and 180 was used as F′. The development of the wavetrain and the anomalous anticyclone over North America in the model was similar to observations. In fact, the wave train position and length scale on 24 June were better simulated than was the case for the early June event, leading to a region 1 pattern correlation of 0.7 between the modeled and observed anomaly. Thus, the model successfully reproduced the essence of cross-Pacific propagation of Rossby waves forced over the west Pacific.

4. Concluding remarks

The evolution of the upper-tropospheric circulation anomalies associated with the severe 1988 U.S. drought is investigated. The April–June seasonal anomaly is not representative of the low-frequency variability during the drought. Rather, several rapidly intensifying and decaying, large-amplitude, anomalous cyclones and anticyclones passed over the Great Plains, with the two strongest anomalous anticyclones developing near the end of May and in mid–late June. It was these two latter events that dominated the AMJ seasonal anomaly over the Great Plains. Examination of daily low-pass streamfunction anomaly maps and numerical experiments with a linear barotropic model strongly suggest that Rossby waves forced in Southeast Asia and the western North Pacific were important to the rapid development of the anomalous circulation over North America in late May and mid June.

Our results are consistent with those of Newman and Sardeshmukh (1998). They demonstrated that height anomalies over North America would be most sensitive to anomalous forcing in the west Pacific during the late spring, the time during which the Southeast Asian monsoon intensifies. Their linear regression of observed 10-day low-pass 200-mb global geopotential height and divergence fields for late spring of the years 1980–95 showed a significant connection between anomalous divergence over the Philippines and anomalous heights over central North America about 7 days later.

Clearly, the current modeling work is very preliminary. The development of the anomalous anticyclones over North America could be different in a more complex model that explicitly includes baroclinicity, topography, and hydrology. For example, the dry surface conditions that existed at the end of May undoubtedly had an impact that could not be considered in the barotropic model. Nevertheless, the fact that the observed evolution of the anomalous anticyclones in June 1988 can be approximately simulated by such a simple model strongly suggests that cross-Pacific propagation of Rossby waves forced over Southeast Asia and the western North Pacific should be considered in understanding the development of the anomalous circulation associated with the 1988 drought.

In this paper, we focused on the atmospheric dynamics in June, the month when the drought was most severe. It is worth recalling from Fig. 1 that extremely low monthly mean precipitation was also observed in the Great Plains in April and May, even though the circulation anomaly aloft was not strongly anticyclonic during either month. Clearly, the strong anomalous anticyclones in June 1988 contributed to the record low precipitation and record high temperatures for the month, markedly enhancing the drought’s severity, but they certainly could not be the cause of the drought in earlier months. In sum, our findings imply that the 1988 drought may be better studied not as one single, steady seasonal event, but rather as a succession of events that together produced a serious hydrological deficit. The extent to which these events are deterministically influenced by boundary forcing (both due to anomalous SST and soil moisture feedbacks), and the extent to which they arise from natural variability of the Tropics/extratropics, remains to be explored.

Acknowledgments

The authors thank Drs. Randall Dole, Martin Hoerling, Walter Robinson, Prashant Sardeshmukh, and Klaus Weickmann for interesting and helpful discussions. Thanks also to Kristie Paine for help with the precipitation data, and Catherine Smith and two anonymous reviewers for comments on an earlier version of this manuscript. This research was supported by the NOAA Climate and Global Change Program.

REFERENCES

  • Atlas, R., N. Wolfson, and J. Terry, 1993: The effect of SST and soil moisture on GLA model simulations of the 1988 U.S. summer drought. J. Climate,6, 2034–2048.

  • Black, R. X., 1997: Deducing anomalous wave source regions during the life cycles of persistent flow anomalies. J. Atmos. Sci.,54, 895–907.

  • Chang, F. C., and J. M. Wallace, 1987: Meteorological conditions during heat waves and droughts in the United States Great Plains. Mon. Wea. Rev.,115, 1253–1269.

  • Ferranti, L., T. N. Palmer, F. Molteni, and E. Klinker, 1990: Tropical–extratropical interaction associated with the 30–60 day oscillation and its impact on medium and extended range prediction. J. Atmos. Sci.,47, 2177–2199.

  • Gruber, A., and A. F. Krueger, 1984: The status of the NOAA outgoing longwave radiation data set. Bull. Amer. Meteor. Soc.,65, 958–962.

  • Hoskins, B. J., and T. Ambrizzi, 1993: Rossby wave propagation on a realistic longitudinally varying flow. J. Atmos. Sci.,50, 1661–1671.

  • Janowiak, J. E., 1988: The global climate for March–May 1988: The end of the 1986–87 Pacific warm episode and the onset of widespread drought in the United States. J. Climate,1, 1019–1040.

  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Karoly, D. J., R. A. Plumb, and M. Ting, 1989: Examples of horizontal propagation of quasi-stationary waves. J. Atmos. Sci.,46, 2802–2811.

  • Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc.,77, 1275–1277.

  • Lyon, B., and R. M. Dole, 1995: A diagnostic comparison of the 1980 and 1988 U.S. summer heat wave–droughts. J. Climate,8, 1658–1675.

  • Mo, K. C., J. R. Zimmerman, E. Kalnay, and M. Kanamitsu, 1991: A GCM study of the 1988 United States drought. Mon. Wea. Rev.,119, 1512–1532.

  • Namias, J., 1991: Spring and summer 1988 drought over the contiguous United States—Causes and prediction. J. Climate,4, 54–65.

  • Newman, M., and P. D. Sardeshmukh, 1998: The impact of the annual cycle on the North Pacific/North American response to remote low-frequency forcing. J. Atmos. Sci.,55, 1336–1353.

  • Nitta, T., 1987: Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation. J. Meteor. Soc. Japan,65, 373–390.

  • NOAA, 1988: Drought advisory 88/12. Summary of conditions and impacts. [Available from Climate Prediction Center, NOAA, World Weather Building, Washington, DC 20233.].

  • Palmer, T. N., and C. Brankovic, 1989: The 1988 United States drought linked to anomalous sea surface temperature. Nature,338, 54–57.

  • Plumb, R. A., 1985: On the three-dimensional propagation of stationary waves. J. Atmos. Sci.,42, 217–229.

  • Rasmusson, E. M., 1968: Atmospheric water vapor transport and the water balance of North America. Part II. Large-scale water balance investigations. Mon. Wea. Rev.,96, 720–734.

  • Riebsame, W. E., S. A. Changnon Jr., and T. R. Karl, 1991: Drought and Natural Resources Management in the United States—Impacts and Implications of the 1987–1989 Drought. Westview Press, 174 pp.

  • Ropelewski, C. F., 1988: The global climate for June–August 1988:A swing to the positive phase of the Southern Oscillation, drought in the United States, and abundant rain in the monsoon areas. J. Climate,1, 1153–1174.

  • Sardeshmukh, P. D., and B. J. Hoskins, 1988: The generation of global rotational flow by steady idealized tropical divergence. J. Atmos. Sci.,45, 1228–1251.

  • Shukla, J., and Y. Mintz, 1982: The influence of land-surface evaporation on the earth’s climate. Science,215, 1498–1501.

  • Spencer, R. W., 1993: Global oceanic precipitation from the MSU during 1979–91 and comparisons to other climatologies. J. Climate,6, 1301–1326.

  • Trenberth, K. E., and G. W. Branstator, 1992: Issues in establishing causes of the 1988 drought over North America. J. Climate,5, 159–172.

  • ——, ——, and P. A. Arkin, 1988: Origins of the 1988 North American drought. Science,242, 1640–1645.

  • Wolfson, N., R. Atlas, and Y. C. Sud, 1987: Numerical experiments related to the summer 1980 U.S. heat wave. Mon. Wea. Rev.,115, 1346–1357.

Fig. 1.
Fig. 1.

Histogram of monthly precipitation anomalies, determined as departures from monthly means of the years 1950–92. Units are in in. month−1. The monthly mean value is printed as “climo.” The value for 1988 is printed atop the appropriate column. Precipitation is averaged in a region bounded by 34°–49°N, 81°–101°W.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 2.
Fig. 2.

AMJ mean and April, May, and June monthly mean streamfunction anomalies at 300 mb. The contour interval is 2.5 × 106 m2 s−1. The zero contour is omitted for clarity; negative values are shaded. In the top panel, the two boxes bound the areas used for the projection coefficient time series displayed in Fig. 3, with the inner box indicating region 1 and the outer box indicating region 2.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 3.
Fig. 3.

Evolution of the projection coefficient. The coefficient is determined by projecting the observed 300-mb 10-day low-pass streamfunction field upon the AMJ-mean streamfunction anomaly, in either region 1 (thin line) or region 2 (thick line).

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 4.
Fig. 4.

Time–longitude plot of observed 300-mb 10-day low-pass streamfunction anomalies averaged over the latitude belt from 30° to 60°N for 1 April 1988–30 June 1988. The contour interval is 4 × 106 m2 s−1. The zero contour is omitted for clarity; negative values are shaded and represented with thinner lines.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 5.
Fig. 5.

Time vs latitude plot of observed 10-day low-pass OLR anomalies for the period 1 April–30 June 1988. The contour interval is 10 W m−2. The zero contour is omitted for clarity; negative values are shaded.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 6.
Fig. 6.

Observed 10-day low-pass streamfunction anomalies at 300 mb, spaced three days apart during the period of initial growth of the early June anticyclone. The contour interval is 5 × 106 m2 s−1. The zero contour is omitted for clarity; negative values are shaded.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 7.
Fig. 7.

Observed 10-day low-pass OLR and residual forcing anomalies at 300 mb, spaced three days apart. The forcing anomalies are indicated by contours; contour interval is 1.5 × 10−10 s−1. The zero contour is omitted for clarity; negative values are indicated by dashed contours. The OLR anomalies (W m−2) are indicated by shading.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 8.
Fig. 8.

Modeled streamfunction anomalies at 300 mb on 1 June 1988, for model runs described in text. (a) Forced with global observed residual forcing. (b) No forcing. Plotting conventions are the same as Fig. 6.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 9.
Fig. 9.

Modeled streamfunction anomalies at 300 mb, spaced three days apart, during the period of initial growth of the early June anticyclone. Forcing is applied only between 90°E and 180. Plotting conventions are the same as Fig. 6.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 10.
Fig. 10.

Plumb flux and streamfunction anomalies at 300 mb averaged over the period 26 May–1 June 1988. (a) Observed. (b) Model results (corresponding to the model run shown in Fig. 9). Plumb flux vectors with a value less than 3 m2 s−2 are not shown; the maximum vector is 45 m2s−2. The zero contour is omitted for clarity; negative values are represented with dashed lines.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 11.
Fig. 11.

Modeled streamfunction anomalies at 300 mb on 1 June 1988, for model runs similar to that shown in Fig. 9 except using different base states described in text. (a) Base state is one month earlier. (b) Base state is one month later. Plotting conventions are the same as Fig. 6.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Fig. 12.
Fig. 12.

Observed (left) and modeled (right) 10-day low-pass streamfunction anomalies at 300 mb, spaced three days apart, during the period of initial growth of the late June anticyclone. Forcing is applied only between 90°E and 180°. Plotting conventions are the same as Fig. 6.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2491:RWPATR>2.0.CO;2

Save