1. Introduction
Cold air is frequently advected along the lee side of high terrain and channeled equatorward, creating a cold surge. Cold surges are shallow regions of cold, stable air that extend approximately 500–1000 km, or one Rossby radius of deformation away from high terrain (Pierrehumbert and Wyman 1985). Cold surges typically have a coherent signature from two days up to an entire week and are often associated with strong low-level equatorward flow exceeding 20 m s−1, decreases in temperature of up to 20°–30°C, and increases in pressure of up to 15–30-hPa over a 24-h period (Garreaud 2001).
Cold surges typically occur with both a dynamic component and an orographic component (Lupo et al. 2001). As a synoptic-scale anticyclone moves into proximity of high terrain a pressure ridge (and associated cold air) can extend along the lee side of this terrain. The Coriolis force acts on the air along the high terrain, turning it to the left (right) in the Southern (Northern) Hemisphere, effectively tying the surge to the terrain (e.g., Forbes et al. 1987; Bell and Bosart 1988; Colle and Mass 1995). This stable, cold air cannot ascend over the high terrain and instead continues equatorward (Manins and Sawford 1982).
While some researchers have hypothesized that cold surges act dynamically, like Kelvin, shelf, topographic Rossby waves, and gravity currents (Schultz et al. 1997), Colle and Mass (1995) and Garreaud (2001) have found that cold surges are dominated by meridional advection of both temperature and anticyclonic vorticity, similar to a cold-air damming event along the East Coast of the United States (Bell and Bosart 1988). A cold-air damming event is characterized by a U-shaped pressure ridge (upside-down U in the Southern Hemisphere), and strong meridional cold-air advection to the lee of high terrain (Bell and Bosart 1988; Colle and Mass 1995).
Cold surges have been studied along many major mountain chains including the Andes (e.g., Garreaud 2001; Lupo et al. 2001), Appalachians (e.g., Forbes et al. 1987; Bell and Bosart 1988), Himalayas (e.g., Sumi 1985; Tilley 1990), and Rockies (e.g., Hartjenstein and Bleck 1991; Colle and Mass 1995). However, only recently have initial investigations into cold surges along the African highlands taken place. Metz et al. (2013) compared and contrasted extreme cold events (ECEs) along the Andes and African highlands. An ECE was defined as the top 1% of coldest 925-hPa temperatures at a particular latitude along the Andes or African highlands cold-surge “pathways” (Fig. 1). ECEs tend to be weaker along the African highlands likely because of the shorter terrain and the moderating influence of the Indian Ocean below (Metz et al. 2013). Synoptic-scale composites of ECEs along both the Andes and African highlands show that ECEs are associated with an amplified ridge–trough pattern and a surface anticyclone that is located in the broad poleward entrance region of a strong 250-hPa jet streak (Metz et al. 2013).
The 20-yr mean annual SSTs (1982–2001; blue shading in degrees Celsius), 25-yr mean annual SLP (1977–2001; black dashed lines every 2 hPa), 25-yr frequency of 925-hPa temperature < 16°C (1977–2001; multicolor shading in percent calculated from 18 262 analysis times), ERA-40 surface elevation (brown–green shading in meters), and points along extreme cold event preferential pathways (black circles and diamonds) (Metz et al. 2013).
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
This paper will expand on the work done by Metz et al. (2013) by examining coherent extreme cold-surge events along the African-highlands pathway (Fig. 1) to determine their frequency and associated characteristics. In particular, this paper will utilize a thickness metric, similar to Lupo et al. (2001), to identify the distribution of the most intense cold surges, in an absolute sense, over a 5-yr climatology. The remainder of this paper will be organized as follows. Section 2 will identify the data and methods utilized to identify and analyze extreme cold-surge events. Section 3 will discuss the results of a short 5-yr climatology of cold surges along the African highlands spanning 2008–12. Section 4 will synthesize the pertinent results and place them into the context of previous research by comparing the climatology of African-highlands cold surges with a climatology of Andes cold surges previously identified by Lupo et al. (2001).
2. Data and methods
A 5-yr climatology of cold surges along the African highlands was created using the 6-hourly 0.5° GFS analysis grids (NOAA/National Centers for Environmental Prediction 2000). This climatology spans 2008–12 and follows a similar methodology as Lupo et al. (2001), who characterized cold surges along the Andes by utilizing 1000–850-hPa thickness maps. For the purposes of this research 1000–850-hPa maps were also used for each cold-surge event along the African highlands to facilitate comparison, and each event was classified as type 1, 2, or 3 [similar to weak, moderate, and strong in Lupo et al. (2001)] depending on the equatorward extent of various thickness contours. Given that the southern extent of the African continent is at a different latitude than South America, some modifications to the Lupo et al. (2001) criteria were made. However, the resulting African-highlands criteria are very similar to those used along the Andes (cf. Lupo et al. 2001) and are robust as each case identified using the 1000–850-hPa thickness maps that is included in this climatology is associated with a terrain-tied cold surge. For a cold surge along the African highlands to be classified as type 1 the 135-dam thickness contour had to cross 35°S between 25° and 40°E (Fig. 2; green shading). A type 2 cold surge occurred when the 135-dam contour crossed 30°S between 30° and 40°E (Fig. 2; blue shading). Finally, a type 3 cold surge occurred when the 135-dam contour crossed 25°S, between 30° and 45°E (Fig. 2; orange shading), or the 138-dam contour crossed 15°S, between 30° and 45°E (Fig. 2; red shading).
Boxes illustrating the criteria used for determining each type of cold surge. Type 1 cold surges feature the 135-dam thickness contour that crossed 35°S between 25° and 40°E (green shading), type 2 cold surges feature the 135-dam contour that crossed 30°S between 30° and 40°E (blue shading), and type 3 cold surges feature the 135-dam contour that crossed 25°S between 30° and 45°E (orange shading) or the 138-dam contour that crossed 15°S between 30° and 45°E (red shading).
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
This paper identifies the most extreme cold surges from a low-level thickness perspective and facilitates comparison with the work completed by Lupo et al. (2001). These extreme cold-surge events are likely to have the most substantial agricultural and societal impacts over the largest range of latitudes. These surges can even approach the equator, potentially leading to the generation of tropical waves (e.g., Liebmann et al. 2009). However, it is important to note that a cold-surge signal can occur to the lee of the African highlands on many days throughout the entire year, and the approach considered herein only examines the most extreme cold surges, which often occur in the cold season. The 135-dam contour was chosen for the type 1 cold-surge events because this value represents 1000–850-hPa thickness values that were approximately one standard deviation below the annual mean 1000–850-hPa thickness at 35°S. Thickness metrics for type 2 and 3 cold surges follow from this benchmark in a similar method to Lupo et al. (2001). The longitudes for each classification were chosen after a thorough examination of synoptic maps to determine the typical horizontal extent of these surges and follow the African-highlands extreme cold event pathway identified in Metz et al. (2013; Fig. 1). If a cold surge featured days that fell under multiple classifications, the surge was classified based on its maximum equatorward extent.
For each of the cold surges identified, the duration, maximum total and north–south (υ) component of the 925-hPa wind speed, and maximum latitudinal extent of 925-hPa cold advection were quantified. A cold surge began at the first GFS analysis period when there was 925-hPa cold advection along the lee of the African highlands (shaded region in Fig. 2) and ended when there was no longer cold advection along this African-highlands pathway. Each cold surge considered in this study was required to last for longer than a 24-h period. After cold advection ended, 925-hPa temperatures tended to modify very quickly given the low latitudes of this cold air. The cold-surge duration (in days) reflects the total length of a cold surge, where, for example, a two-day surge is longer than 24 h and no longer than 48 h. All other variables mentioned above were also calculated within the shaded region of Fig. 2.
3. Climatological results
The 2008–12 climatology features a total of 186 cold-surge cases spanning all months of the year. Of these 186 cases, 84 are type 1, 27 are type 2, and 75 are type 3 cold surges. Furthermore, 145 of these 186 cases are canonical cold-surge events where an anticyclone tracks around the southern extent of the African continent advecting cold Antarctic air equatorward. The other 41 events feature a cyclone of varying intensities located to the east of Africa near the southwest coast of Madagascar. The western side of the cyclonic circulation aids in cold advection initially associated with an anticyclone to the lee of the African highlands.
a. Monthly and seasonal frequency
The frequency of cold-surge events peaks in September when 19% (35 events) of events occur followed closely by August when 14% (26 events) of events occur (Fig. 3a). Although September contains the most cold-surge events, the type 3 events are confined to the Southern Hemisphere winter of June, July, and August. Of the 72 events during these three months 83% (60 events) of them are type 3 (Fig. 3a). This seasonal lag in type 1 and 2 events is similar to what was found in Metz et al. (2013) for ECEs along the African highlands. Approximately 81% (151 events) of all cold-surge events occur in the winter and spring seasons, with a few more in spring because of the abundance of type 1 and 2 surges (Fig. 3b). The large number of type 1 and 2 cold surges in the spring may be due to the fact that a majority of the African-highlands pathway is over water. The thermal modification of the air by the Indian Ocean below could lengthen the cold-surge season. This seasonal lag is also seen in Appalachian cold-air damming events that maximize in March and occur near the Atlantic Ocean (Bell and Bosart 1988). However, the most type 3 cold surges (greatest equatorward extent) occur in June, July, and August when the source of Antarctic air is coldest.
(a) The monthly frequency of cold surges along the African highlands, and (b) the seasonal distribution of cold surges along the African highlands. Type 1, 2, and 3 cold surges are indicated by green, blue, and red, respectively. The percentage of cases in each bin is indicated on the top of each bar. Note that the y axes in (a) and (b) have different scales.
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
b. Duration of cold surge
African-highlands cold surges most typically last 3 days, as 33% (62 events) of all cold surges exhibit this duration (Fig. 4). Type 1 and 2 cold-surge events most frequently last between 2 and 4 days, while cold surges that extend farther equatorward are generally associated with longer durations as 66% (25 out of 38) of all cold surges lasting 5–8 days are type 3 (Fig. 4). Generally, as event duration increases the number of total cold surges decreases, indicating that cold surges along the African highlands typically feature a relatively short event duration (Fig. 4).
The distribution of cold-surge duration along the African highlands. Type 1, 2, and 3 cold surges are indicated by green, blue, and red, respectively The percentage of cases in each bin is indicated on the top of each bar.
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
c. Maximum 925-hPa wind speed
The most frequent maximum total wind speed associated with African-highlands cold surges is 35.0–39.9 kt (1 kt = 0.51 m s−1; 41 events), and 42% (79 events) of events exhibit a maximum 925-hPa total wind speed between 35.0 and 44.9 kt (Fig. 5a). Higher maximum wind speeds do not seem to be preferentially associated with events that move farther equatorward, as only 44% (31 of 70) of events associated with maximum wind speeds greater than or equal to 45.0 kt are type 3, whereas 38% (44 of 116) of events associated with maximum wind speeds less than or equal to 44.9 kt are also type 3 (Fig. 5a). The Somali jet is located off the eastern coast of Africa, particularly equatorward of 15.0°S. This jet can contribute to strong winds along the lee of the African highlands regardless of cold-surge classification (Findlater 1969a,b).
(a) The distribution of maximum 925-hPa wind speed associated with cold surges along the African highlands, and (b) the distribution of the maximum 925-hPa υ component of the wind associated with cold surges along the African highlands. Type 1, 2, and 3 cold surges are indicated by green, blue, and red, respectively. The percentage of cases in each bin is indicated on the top of each bar.
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
As with the distribution of 925-hPa total wind speed, the maximum υ component of the 925-hPa wind exhibits the greatest frequency of events between 35.0 and 39.9 kt (62 events). A majority (99 events) of events were associated with maximum υ-component wind speeds between 30.0 and 39.9 kt (Fig. 5b). When compared with the 925-hPa total wind speed distribution, the υ-component distribution is narrower and shifts toward slightly smaller values (Fig. 5).
d. Maximum latitudinal extent
Cold surge events most frequently extend between 15.0° and 24.9°S (114 events). Of the 186 cold-surge events, 35 (19%) featured 925-hPa cold advection that progressed to within 4.9°S of the equator, and 91% (32 of 35) of these events are classified as type 3 (Fig. 6). While previous research has shown that cold surges along the Andes can reach and even cross the equator (e.g., Parmenter 1976; Liebmann et al. 2009), the African highlands only maximize at a height approximately half that of the Andes. Thus, it is surprising that so many African-highlands cold surges extend to near the equator. There are two possible reasons for the relatively large number of surges that approach the equator. First, the shape of the eastern coastline of Africa is such that it projects eastward at about 15.0°S. North of this feature, the coast is oriented in a general north–south fashion. Thus, if the cold advection can progress equatorward of 15.0°S, then it may be able to move more easily toward the equator. Second, the Somali jet is positioned immediately off the eastern coast of Africa north of 15.0°S (Findlater 1969a,b) and reaches its most western extent from June to September (Krishnamurti et al. 1976). Thus, if cold air can progress equatorward of 15.0°S, it can intersect with the climatological position of the Somali jet, and this jet can aid in the transport of this cold air toward the equator. The presence of the Somali jet may inhibit cold surges from having their maximum latitudinal extent between 5.0° and 14.9°S (Fig. 6).
The distribution of the maximum equatorward extent of cold surges along the African highlands. Type 1, 2, and 3 cold surges are indicated by green, blue, and red, respectively. The percentage of cases in each bin is indicated on the top of each bar.
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
4. Summary and discussion
In this study, a 5-yr climatology of the most extreme cold-surge events along the African highlands was completed. These extreme cold-surge events are likely to have the largest agricultural and societal impacts over a wide range of latitudes. Cold surges along the African highlands occur most frequently in the month of September, with type 3 surges (extending farthest equatorward) being confined to the winter months of June, July, and August. Of the 186 cold-surge events identified, 84 are type 1, 27 are type 2, and 75 are type 3. The relatively small number of type 2 cold surges could be a result of the cold-surge definition provided herein. However, this definition was chosen for consistency with the Lupo et al. (2001) study. Other results of this study show that cold surges most frequently last for three days, with type 3 cold surges typically lasting longer. The maximum magnitude of the 925-hPa wind speed associated with a cold surge does not predict its equatorward extent. Furthermore, extreme cold surges along the African highlands most frequently extend to between 15.0° and 24.9°S, with 19% of all cold-surge events progressing to within 4.9°S of the equator.
When comparing the distribution of African highlands cold surges with that found by Lupo et al. (2001) for cold surges along the Andes Mountains (both studies utilized a 5-yr climatology and followed a similar methodology), it is seen that the African-highlands cold-surge season peaks in September whereas the Andes season peaks in July (Fig. 7a). This difference may be due to the influence of the Indian Ocean on African-highlands cold surges. It is important to note that the cold-surge climatology presented herein for African-highlands cold surges spans a different 5-yr period than the Lupo et al. (2001) climatology. However, these differences in cold-season peaks between African-highlands and Andes cold surges are similar to differences in the distributions of ECEs along these two pathways shown by Metz et al. (2013). When comparing the distribution of all Andes cold surges with the distribution of only type 3 surges along the African highlands, it is seen that the shapes of the distributions are very similar, with an approximate normal distribution from May through September (Fig. 7b).
(a) The distribution of the total number of Andes cold surges from a 5-yr climatology [from Lupo et al. (2001), n = 256, blue bars] compared with the total number of African-highlands cold surges from this 5-yr climatology (n = 186, orange bars), and (b) the distribution of the total number of Andes cold surges from a 5-yr climatology [from Lupo et al. (2001), n = 256, blue bars] compared with the type 3 African-highlands cold surges from this 5-yr climatology (n = 75, red bars).
Citation: Journal of Applied Meteorology and Climatology 56, 6; 10.1175/JAMC-D-15-0191.1
Lupo et al. (2001) found that there were 256 total cold-surge events along the Andes, whereas this study found 186 cold-surge events along the African highlands (a 27% reduction). This decreased frequency could result from a few possible reasons. First, the Andes are approximately 2 times as high as the African highlands. The higher altitude of the Andes allows the mountains to be more effective at blocking cold, stable air that can subsequently be transported equatorward. Second, the southernmost point of continental South America is approximately 20° of latitude closer to the South Pole than the southernmost point of continental Africa. This difference in southern extent allows for anticyclones to advect cold air from Antarctica into proximity of the Andes with potentially less modification by the water below. Third, a majority of the African-highlands cold-surge pathway is located over water, whereas the Andes cold-surge pathway is strictly over land. The moderating influence of the ocean likely plays a role in the reduced cold-surge frequency along the African highlands (Fig. 7a).
Future work on African-highlands cold surges will involve creating composites of cold-surge environments for both the canonical cold surges events and those that feature a cyclone off the southwest coast of Madagascar (see section 3). Representative case-study events from each composite will be identified to more clearly examine the synoptic-scale characteristics associated with typical African-highlands cold surges. Finally, this work utilizes a thickness metric to identify the distribution of the most extreme cold surges over the 5-yr period considered and enables comparison with Lupo et al. (2001). This absolute thickness metric identifies the months of June–November as the period of most extreme cold surges. An alternate perspective on this issue might be to construct a standardized anomaly metric to examine the distribution of cold surges throughout the entire year, many of which are likely to be less intense (from a thickness perspective) than those experienced in during the cold-season months. This alternative perspective would likely result in the identification of additional cold-surge events and might change the monthly distribution of cold surges identified in this paper given that weaker events would also be considered.
Acknowledgments
The authors thank Dr. Heather Archambault, Dr. Eric Hoffman, and Dr. Neil Laird for their helpful suggestions and comments throughout the entirety of this project. Two anonymous reviewers greatly improved the quality of this manuscript. This project was partially completed by the first author as a part of the Hobart and William Smith Colleges undergraduate summer research program. Support for this project was provided by the Hobart and William Smith Colleges Provost Office and National Science Foundation Grant AGS-1258548.
REFERENCES
Bell, G. D., and L. F. Bosart, 1988: Appalachian cold-air damming. Mon. Wea. Rev., 116, 137–161, doi:10.1175/1520-0493(1988)116<0137:ACAD>2.0.CO;2.
Colle, B. A., and C. Mass, 1995: The structure and evolution of cold surges east of the Rocky Mountains. Mon. Wea. Rev., 123, 2577–2610, doi:10.1175/1520-0493(1995)123<2577:TSAEOC>2.0.CO;2.
Findlater, J., 1969a: A major low-level air current near the Indian Ocean during the northern summer. Quart. J. Roy. Meteor. Soc., 95, 362–380, doi:10.1002/qj.49709540409.
Findlater, J., 1969b: Interhemispheric transport of the air in the lower troposphere over the western Indian Ocean. Quart. J. Roy. Meteor. Soc., 95, 400–403, doi:10.1002/qj.49709540412.
Forbes, G. S., D. W. Thomson, and R. A. Anthes, 1987: Synoptic and mesoscale aspects of an Appalachian ice storm associated with cold-air damming. Mon. Wea. Rev., 115, 564–591, doi:10.1175/1520-0493(1987)115<0564:SAMAOA>2.0.CO;2.
Garreaud, R. D., 2001: Subtropical cold surges: Regional aspects and global distribution. Int. J. Climatol., 21, 1181–1197, doi:10.1002/joc.687.
Hartjenstein, G., and G. Bleck, 1991: Factors affecting cold-air outbreaks east of the Rocky Mountains. Mon. Wea. Rev., 119, 2280–2292, doi:10.1175/1520-0493(1991)119<2280:FACAOE>2.0.CO;2.
Krishnamurti, T. N., J. Molinari, and H. L. Pan, 1976: Numerical simulation of the Somali jet. J. Atmos. Sci., 33, 2350–2362, doi:10.1175/1520-0469(1976)033<2350:NSOTSJ>2.0.CO;2.
Liebmann, B., G. N. Kiladis, L. M. V. Carvalho, C. Jones, C. S. Vera, I. Bladé, and D. Allured, 2009: Origin of convectively coupled Kelvin waves over South America. J. Climate, 22, 300–315, doi:10.1175/2008JCLI2340.1.
Lupo, A. R., J. J. Nocera, L. F. Bosart, E. G. Hoffman, and D. J. Knight, 2001: South American cold surges: Types, composites, and case studies. Mon. Wea. Rev., 129, 1021–1041, doi:10.1175/1520-0493(2001)129<1021:SACSTC>2.0.CO;2.
Manins, P. C., and B. L. Sawford, 1982: Mesoscale observations of upstream blocking. Quart. J. Roy. Meteor. Soc., 108, 427–434, doi:10.1002/qj.49710845608.
Metz, N. D., H. M. Archambault, A. F. Srock, T. J. Galarneau, and L. F. Bosart, 2013: A comparison of South American and African preferential pathways for extreme cold events. Mon. Wea. Rev., 141, 2066–2086, doi:10.1175/MWR-D-12-00202.1.
NOAA/National Centers for Environmental Prediction, 2000: NCEP FNL Operational Model Global Tropospheric Analyses, continuing from July 1999 (updated daily). National Center for Atmospheric Research Computational and Information Systems Laboratory Research Data Archive, accessed 14 June 2013, doi:10.5065/D6M043C6.
Parmenter, F. C., 1976: A Southern Hemisphere cold front passage at the equator. Bull. Amer. Meteor. Soc., 57, 1435–1440, doi:10.1175/1520-0477(1976)057<1435:ASHCFP>2.0.CO;2.
Pierrehumbert, R. T., and B. Wyman, 1985: Upstream effects of mesoscale mountains. J. Atmos. Sci., 42, 977–1003, doi:10.1175/1520-0469(1985)042<0977:UEOMM>2.0.CO;2.
Schultz, D., W. Bracken, L. Bosart, G. Hakim, M. Bedrick, M. Dickinson, and K. Tyle, 1997: The 1993 superstorm cold surge: Frontal structure, gap flow, and extratropical impact. Mon. Wea. Rev., 125, 5–39, doi:10.1175/1520-0493(1997)125<0005:TSCSFS>2.0.CO;2.
Sumi, A., 1985: A study of cold surges around the Tibetan Plateau by using numerical models. J. Meteor. Soc. Japan, 63, 377–395.
Tilley, J. S., 1990: On the application of edge wave theory to terrain-bounded cold surges: A numerical study. Ph.D. thesis 130, The Pennsylvania State University and National Center for Atmospheric Research, 353 pp. [Available from UCAR communications, P.O. Box 3000, Boulder, CO 80307-3000.]
