• Beasley, W., 1985: Positive cloud-to-ground lightning observations. J. Geophys. Res., 90 ((D4),) 61316138.

  • Burrows, W. R., , and P. King, 2000: Neuro-statistical models for predicting lightning occurrence in Canada: Climatology and potential predictors. Proc. 2000 Int. Lightning Detection Conf., Tucson, AZ, Global Atmospherics, Inc., Paper 15. [Available from Global Atmospherics, Inc., 2705 E. Medina, Tucson, AZ 85706.].

    • Search Google Scholar
    • Export Citation
  • Burrows, W. R., , P. J. Lewis, , B. Kochtubajda, , B. Snyder, , and V. Turcotte, 2002: Lightning occurrence patterns over Canada and adjacent United States from lightning detection network observations. Atmos.–Ocean,40, 59–80.

    • Search Google Scholar
    • Export Citation
  • Cooray, V., 1997: Energy dissipation in lightning flashes. J. Geophys. Res., 102 ((D17),) 2140121410.

  • Cummins, K. L., , M. J. Murphy, , E. A. Bardo, , W. L. Hiscox, , R. B. Pyle, , and A. E. Pifer, 1998: A combined TOA/MDF technology upgrade of the U.S. National Lightning Detection Network. J. Geophys. Res., 103 ((D8),) 90359044.

    • Search Google Scholar
    • Export Citation
  • Cummins, K. L., , R. B. Pyle, , and G. Fournier, 1999: An integrated North American Lightning Detection Network. Proc. 11th Int. Conf. on Atmospheric Electricity, Guntersville, AL, NASA/CP-1999-209261, 218–221. [Available from NASA Center for AeroSpace Information, 800 Elkridge Landing Road, Linthicum Heights, MD 21090-2934.].

    • Search Google Scholar
    • Export Citation
  • Fosdick, E. K., , and A. I. Watson, 1995: Cloud-to-ground lightning patterns in New Mexico during the summer. Natl. Wea. Dig., 19 , 1724.

    • Search Google Scholar
    • Export Citation
  • Idone, V. P., , D. A. Davis, , P. K. Moore, , Y. Wang, , R. W. Henderson, , M. Ries, , and P. F. Jamason, 1998a: Performance evaluation of the U.S. National Lightning Detection Network in eastern New York. Part I: Detection efficiency. J. Geophys. Res., 103 ((D8),) 90459055.

    • Search Google Scholar
    • Export Citation
  • Idone, V. P., . 1998b: Performance evaluation of the U.S. National Lightning Detection Network in eastern New York. Part II: Location accuracy. J. Geophys. Res., 103 ((D8),) 90579069.

    • Search Google Scholar
    • Export Citation
  • Krider, E. P., , C. Leteinturier, , and J. C. Willett, 1996: Submicrosecond fields radiated during the onset of first return strokes in cloud-to-ground lightning. J. Geophys. Res., 101 ((D1),) 15891597.

    • Search Google Scholar
    • Export Citation
  • López, R. E., , and R. L. Holle, 1986: Diurnal and spatial variability of lightning activity in northeastern Colorado and central Florida during the summer. Mon. Wea. Rev., 114 , 12881312.

    • Search Google Scholar
    • Export Citation
  • Lyons, W. A., , T. E. Nelson, , E. R. Williams, , J. Cramer, , and T. Turner, 1998a: Enhanced positive cloud-to-ground lightning in thunderstorms ingesting smoke from fires. Science, 282 , 7781.

    • Search Google Scholar
    • Export Citation
  • Lyons, W. A., , M. Uliasz, , and T. E. Nelson, 1998b: Large peak current cloud-to-ground lightning flashes during the summer months in the contiguous United States. Mon. Wea. Rev., 126 , 22172233.

    • Search Google Scholar
    • Export Citation
  • Mackerras, D., , and M. Darveniza, 1994: Latitudinal variation of lightning occurrence characteristics. J. Geophys. Res., 99 , 1081310821.

    • Search Google Scholar
    • Export Citation
  • Murray, N. D., , R. E. Orville, , and G. R. Huffines, 2000: Effect of pollution from Central American fires on cloud-to-ground lightning. Geophys. Res. Lett., 27 , 22492252.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., 1990: Lightning return stroke peak current variation as a function of latitude. Nature, 342 , 149151.

  • Orville, R. E., . 1991: Lightning ground flash density in the contiguous United States–1989. Mon. Wea. Rev., 119 , 573577.

  • Orville, R. E., . 1994: Cloud-to-ground lightning flash characteristics in the contiguous United States: 1989–1991. J. Geophys. Res., 99 ((D5),) 1083310841.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., , and A. C. Silver:, 1997: Lightning ground flash density in the contiguous United States: 1992–95. Mon. Wea. Rev., 125 , 631638.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., , and G. R. Huffines:, 1999: Lightning ground flash measurements over the contiguous United States: 1995–97. Mon. Wea. Rev., 127 , 26932703.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., . 2001: Cloud-to-ground lightning in the USA: NLDN results in the first decade, 1989–98. Mon. Wea. Rev., 129 , 11791193.

    • Search Google Scholar
    • Export Citation
  • Reap, R. M., 1986: Evaluation of cloud-to-ground data from the western United States for the 1983–84 summer seasons. J. Climate Appl. Meteor., 25 , 785799.

    • Search Google Scholar
    • Export Citation
  • Tyahla, L. J., , and R. E. López, 1994: Effect of surface conductivity on the peak magnetic field radiated by first return strokes in cloud-to-ground lightning. J. Geophys. Res., 99 ((D5),) 1051710525.

    • Search Google Scholar
    • Export Citation
  • Wacker, R. S., , and R. E. Orville, 1999a: Changes in measured lightning flash count and return stroke peak current after the 1994 U.S. National Lightning Detection Network upgrade. Part I: Observations. J. Geophys. Res., 104 ((D2),) 21512157.

    • Search Google Scholar
    • Export Citation
  • Wacker, R. S., . 1999b: Changes in measured lightning flash count and return stroke peak current after the 1994 U.S. National Lightning Detection Network upgrade. Part II: Theory. J. Geophys. Res., 104 ((D2),) 21592162.

    • Search Google Scholar
    • Export Citation
  • Zajac, B. A., , and S. A. Rutledge, 2001: Cloud-to-ground lightning activity in the contiguous United States from 1995 to 1999. Mon. Wea. Rev., 129 , 9991019.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    The total area covered by the NALDN is shown in light gray. Blue symbols represent the CLDN composed of 87 sensors, and red represents the NLDN of 106 sensors. Each network is composed of IMPACT/ES and TOA sensors. Triangles mark the TOA sensor locations and circles mark IMPACT/ES sensor locations

  • View in gallery

    Four-panel display shows the annual ground flash density in North America for the individual years (a) 1998, (b) 1999, and (c) 2000, and (d) the mean flash density for the 3 yr combined

  • View in gallery

    The local hours of maximum ground flash density are plotted for two time increments, from 0800 to 2000 LST in red and from 2000 to 0800 LST in blue. Thus, red shows maximum lightning in daytime and blue is at night

  • View in gallery

    Same as in Fig. 2 except the positive ground flash density is plotted by color

  • View in gallery

    Same as in Fig. 2 except the percentage of lightning ground flashes lowering positive charge is color shaded

  • View in gallery

    Same as in Fig. 2 except the negative median peak current is shaded for ground flashes lowering negative charge

  • View in gallery

    Same as in Fig. 2 except for positive median peak currents

  • View in gallery

    Same as in Fig. 2 except for mean negative multiplicity

  • View in gallery

    Same as in Fig. 2 except for mean positive multiplicity.

  • View in gallery

    Median peak current is plotted as a function of the flash multiplicity for each polarity for each of the two networks, the NLDN and the CLDN

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 177 177 12
PDF Downloads 135 135 12

The North American Lightning Detection Network (NALDN)—First Results: 1998–2000

View More View Less
  • 1 Department of Atmospheric Sciences and Cooperative Institute for Applied Meteorological Studies, Texas A&M University, College Station, Texas
  • | 2 Air Force Institute of Technology, Wright-Patterson Air Force Base, Ohio
  • | 3 Meteorological Research Branch, Meteorological Service of Canada, Downsview, Ontario, Canada
  • | 4 Global Atmospherics, Inc., Tucson, Arizona
© Get Permissions
Full access

Abstract

Cloud-to-ground lightning data have been analyzed for the years 1998–2000 for North America (Canada plus the contiguous United States) for all ground flashes, positive flashes, the percentage of positive lightning, peak currents for negative and positive lightning, and for negative and positive multiplicity. The authors examined a total of 88.7 million flashes divided among the three years: 31.1 million (1998), 29.5 million (1999), and 28.2 million (2000). Annual flash densities are derived from 245–424 km2 regions and are uncorrected for flash detection efficiency. The highest flash densities in Canada are along the U.S.–Canadian border (1–3 flashes km−2), and in the United States along the Gulf of Mexico coast and Florida (exceeding 9 flashes km−2). Maximum annual positive flash densities in Canada generally range primarily from 0.1 to 0.3 flashes km−2, and in the United States to over 0.7 flashes km−2 (areas in the Midwest, the Gulf Coast, and Florida). Areas of greater than 20% positive lightning occur throughout British Columbia and the midwest United States extending into Manitoba and Ontario. High percent positive also occurs in Quebec and much of eastern Canada. The median negative peak current is 16.5 kA. The median positive peak current, with the peak currents less than 10 kA removed from the calculation, is 19.8 kA. Median positive peak currents exceed 35 kA in the Midwest from west Texas to Nebraska to the Canadian border. The area of maximum mean negative multiplicity, exceeding 2.6 strokes, occurs in western Canada from just east of the British Columbia–Alberta border to and including Saskatchewan. Mean negative multiplicity also peaks in the southeastern United States. Mean positive multiplicity is observed to have maximum values in Alberta, Saskatchewan, and in a region centered on Tennessee. The authors examined the time of maximum flash rate in North America and find it is over land in the daytime hours with the exception of a region of maximum nighttime lightning in midcontinent that extends from the midwestern United States into Canada. Over the waters surrounding the North American continent, the maximum lightning is principally at night, including the coastal Pacific, the Gulf of California, the Gulf of Mexico, and the coastal waters of the North Atlantic.

Corresponding author address: Dr. Richard E. Orville, Department of Atmospheric Sciences, Texas A&M University, CIAMS, College Station, TX 77843-3150. Email: rorville@tamu.edu

Abstract

Cloud-to-ground lightning data have been analyzed for the years 1998–2000 for North America (Canada plus the contiguous United States) for all ground flashes, positive flashes, the percentage of positive lightning, peak currents for negative and positive lightning, and for negative and positive multiplicity. The authors examined a total of 88.7 million flashes divided among the three years: 31.1 million (1998), 29.5 million (1999), and 28.2 million (2000). Annual flash densities are derived from 245–424 km2 regions and are uncorrected for flash detection efficiency. The highest flash densities in Canada are along the U.S.–Canadian border (1–3 flashes km−2), and in the United States along the Gulf of Mexico coast and Florida (exceeding 9 flashes km−2). Maximum annual positive flash densities in Canada generally range primarily from 0.1 to 0.3 flashes km−2, and in the United States to over 0.7 flashes km−2 (areas in the Midwest, the Gulf Coast, and Florida). Areas of greater than 20% positive lightning occur throughout British Columbia and the midwest United States extending into Manitoba and Ontario. High percent positive also occurs in Quebec and much of eastern Canada. The median negative peak current is 16.5 kA. The median positive peak current, with the peak currents less than 10 kA removed from the calculation, is 19.8 kA. Median positive peak currents exceed 35 kA in the Midwest from west Texas to Nebraska to the Canadian border. The area of maximum mean negative multiplicity, exceeding 2.6 strokes, occurs in western Canada from just east of the British Columbia–Alberta border to and including Saskatchewan. Mean negative multiplicity also peaks in the southeastern United States. Mean positive multiplicity is observed to have maximum values in Alberta, Saskatchewan, and in a region centered on Tennessee. The authors examined the time of maximum flash rate in North America and find it is over land in the daytime hours with the exception of a region of maximum nighttime lightning in midcontinent that extends from the midwestern United States into Canada. Over the waters surrounding the North American continent, the maximum lightning is principally at night, including the coastal Pacific, the Gulf of California, the Gulf of Mexico, and the coastal waters of the North Atlantic.

Corresponding author address: Dr. Richard E. Orville, Department of Atmospheric Sciences, Texas A&M University, CIAMS, College Station, TX 77843-3150. Email: rorville@tamu.edu

1. Introduction

Annual lightning ground flash characteristics for the continental United States have been reported for the past decade beginning with the summary by Orville (1991) for 1989 and continuing for 1990–91 (Orville 1994), 1992–95 (Orville and Silver 1997), 1995–97 (Orville and Huffines 1999), 1995–99 (Zajac and Rutledge 2001) and a recent summary for the first decade, 1989–98 (Orville and Huffines 2001). Regional and seasonal studies have also been reported. These include the latitudinal variation of ground flash return stroke peak current (Orville 1990) and the summer distribution of large peak currents 1991–95 (Lyons et al. 1998b). Regional studies, such as Fosdick and Watson (1995) for New Mexico, are important and contribute to our understanding of thunderstorms and lightning. Zajac and Rutledge (2001) published a comprehensive list of national and regional lightning studies in the United States and the reader should consult this for additional information.

In this paper, we continue to summarize annual lightning characteristics, but with one major enhancement. We add the lightning data from the Canadian Lightning Detection Network (CLDN) (Burrows and King 2000; Burrows et al. 2002), which began its first full year of operation in 1998. Consequently, we report for the first time on the combined annual lightning characteristics for the National Lightning Detection Network (NLDN) and for the CLDN. Cummins et al. (1999) refer to these two networks as the North American Lightning Detection Network (NALDN). We summarize the lightning characteristics for this portion of North America for 1998–2000, including an analysis of the time of maximum lightning activity over North America and surrounding waters. We believe that a number of unique lightning-related observations are now possible by the creation of this single homogeneous network covering nearly 20 million km2, over latitudes ranging from 25° to 67° in the west, and 25° to 55° in the east.

2. Data

The NALDN coverage (Fig. 1) is provided by 187 sensors consisting of the NLDN with 106 sensors (red) and the CLDN with 81 sensors (blue). The Canadian network uses a mix of 26 Improved Accuracy from Combined Technology/Enhanced Sensitivity (IMPACT/ES) sensors and 55 Lightning Position and Tracking System (LPATS-IV) time-of-arrival (TOA) sensors. These are next-generation sensors beyond the capabilities of those currently in the NLDN, and have the capability of detecting and locating inter- and intracloud discharges. However, because these sensors are separated by 300–500 km, only a small percentage of cloud discharges are located by the CLDN. Assuming a nominal detection range of 600 km for each sensor, the area covered by the combined, integrated networks is approximately 20 million km2. Within the NLDN area in Fig. 1, the estimated median accuracy of the lightning locations is 500 m (Idone et al. 1998a) and the flash detection efficiency is 80%–90% (Idone et al. 1998b; Cummins et al. 1998) for peak currents greater than 5 kA. An analysis of the CLDN by Global Atmospherics, Inc. estimates that the flash detection efficiency is 85%–90% out to 200 km from the network periphery, decreasing to 80% at the periphery, and to 10%–30% at a distance of 300 km beyond the periphery. Because of the overall high detection efficiency of the NALDN, no corrections for detection efficiency are applied to the flash density calculations.

It is important to note that in our analyses we have eliminated positive flashes from our analysis with peak currents less than 10 kA. This was originally suggested by Cummins et al. (1998) and explained by Wacker and Orville (1999a,b). To quote Cummins et al., “We recommend that the subset of small positive discharges with peak currents less than 10 kA be regarded as cloud discharges unless they are verified to be cloud-to-ground.” We agree. One can argue that we should take into account the fact that the CLDN sensors are less susceptible to the misclassification of cloud flashes than the NLDN sensors, but we note that the NLDN sensors contribute data up to 500 km into Canada. Therefore, for the sake of consistent statistics, we apply the 10 kA limit to the whole NALDN. It will just have less relevance in Canada.

All geographical plots for this paper were calculated with a spatial resolution of 0.2° corresponding to an approximate resolution of 20 km. Longitudinal lines converge significantly over a north–south geographical area as large as North America. Consequently, the area over which the flash density is calculated varies from 424 km2 at 30°N, to 346 km2 at 45°N, to 245 km2 at 60°N. Our flash density calculations are exact and compensate for the changing area as the longitudinal lines converge.

3. Results

All results for the NALDN are summarized in the categories of flash density, percent positive, first stroke peak currents for both polarities, and multiplicity for both polarities. Table 1 summarizes the data for the unified network, the NALDN. There are two sections in Table 1: all measured flashes in the first set of columns, and all measured flashes without the weak positive flashes (<10 kA) in the second set of columns. We suspect, as discussed above, that the weak positive flashes are mostly cloud flashes (Cummins et al. 1998; Wacker and Orville 1999a,b) and have therefore removed them from the dataset. Our analysis is of the remaining 86 million ground flashes over the period 1998–2000. In the following discussion, we examine the geographical variations of ground flash characteristics in North America and we will be referring back to Table 1.

a. Flash density

The total number of cloud-to-ground flashes recorded by the combined network is listed in Table 1 for 1998–2000. Measured flashes totaled approximately 30 million cloud-to-ground flashes each year, varying from 31.071 million in 1998 to 28.176 million in 2000. When the positive flashes with peak currents less than 10 kA are removed as discussed previously, the annual flash totals decrease to approximately 28 million per year. The geographical variation of flash densities for each of these years is color shaded in Figs. 2a–c along with the average flash density for these 3 years (Fig. 2d). The flash density varies from less than 0.1 flashes km−2 yr−1 over the northern and western extremes of Canada (e.g., British Columbia, Hudson Bay) and western United States to over 9 flashes km−2 yr−1 in Florida.

A particularly noticeable flash density variation in Canada (Fig. 2) is the dramatic increase in the flash density that occurs along the eastern edge of the Canadian Rockies. [López and Holle (1986) discovered a similar boundary on the east side of the Colorado Rockies.] This “boundary” appears as a southeast to northwest line along the British Columbia–Alberta border, separating a western area with 0.1–0.25 flashes km−2 yr−1 from an eastern area with 1–3 flashes km−2 yr−1. At a grid resolution of 20 km, the flash density in Canada does not exceed 3 flashes km−2 yr−1 except for a small region of southwestern Ontario northeast of Lake St. Clair where 3.8 flashes km−2 yr−1 were seen.

Areas of minimum flash density over western Canada occur over western British Columbia and the adjacent Pacific Ocean and over far northern Saskatchewan and Manitoba. These regions are dominated by air masses that are cool and stable in the lower levels, or have relatively low moisture content compared to the continent to the south and east. Across the prairies, stretching eastward from central Alberta is a belt of higher flash occurrence than in surrounding areas. It is likely that storms form along the eastern slope of the Rockies, move eastward with the mean steering flow, and are able to persist long enough to cause the belt of high flash occurrence extending east into Saskatchewan. South of this belt is a relative minimum. Southeast from there is a large region of higher flash occurrence straddling the international border along southern Saskatchewan, Manitoba, and Ontario interrupted by a relative minimum in the vicinity of the Manitoba–North Dakota border.

Lightning flash density is sometimes greater over the slopes leading to higher terrain and is minimal over large areas of rough terrain (Reap 1986; Fosdick and Watson 1995). This is seen in Fig. 2d in the Appalachian Mountains in western Virginia and eastern West Virginia, and over the Ozark plateau in Arkansas. In contrast, we sometimes see a strong relationship between flash density and significant elevated terrain features, where flash density increases with elevation along the sides of terrain features and then decreases at the top as was found in the CLDN data by Burrows et al. (2002). A good example is in southern Alberta in Fig. 2d, where a sharp maximum lies over the Rocky Mountain foothills just east of a sharp minimum over the Continental Divide, marked by the Alberta–British Columbia border. Another flash density contour change is noted by Burrows et al. (2002) over the Adirondack Mountains of New York, and by Orville and Huffines (2001) over the Appalachian Mountains. The latter is clearly visible in Fig. 2.

Many of the flash density features appearing in the NLDN decade lightning summary published by Orville and Huffines (2001) appear in the annual panels (Figs. 2a–c) and in the composite (Fig. 2d). High flash densities persist in Florida, along the Gulf Coast, in the upper Midwest and over the Gulf Stream. A region of high flash density is apparent in northwestern Mexico. With the addition of the Canadian lightning network data, we note the broad regions of lower flash density that extend from the Canadian northwest to the northeastern United States, a feature that was not apparent with the limited coverage of just the NLDN for the United States. A steady decline of flash density occurs from the lower Great Lakes to the upper Great Lakes.

b. Time of maximum flash rate

In this section we examine the time of the maximum flash rate in North America for the broad categories of daytime and nighttime lightning and plot the annual results for each year (Figs. 3a–c) and for the 3 years (Fig. 3d). The daytime maxima are plotted in red and cover the local time from 0800 to 2000 local standard time (LST) whereas the nighttime maxima are plotted in blue and cover the local time from 2000 to 0800 LST. Note that the time of the maximum flash rate over land is in the daytime hours with the exception of a region of maximum nighttime lightning in the midcontinent, which extends from the midwestern United States into Canada. Over the waters surrounding the North American continent, the time of maximum lightning is principally at night, including the coastal Pacific, the Gulf of California, the Gulf of Mexico, and the coastal waters of the North Atlantic.

c. Positive flash density

Positive flashes are separated from the total ground flash count and plotted in Fig. 4. A similar distribution was published by Orville and Huffines (2001) for the NLDN, but here we have the extension of the positive flash density into Canada. There are no surprises. The positive flash densities in the upper Midwest in 1998 (Fig. 4a) appear slightly higher than in the two subsequent years (Figs. 4b,c). This maximum may reflect the enhanced positive lightning flash density in 1998, first detected and reported by Lyons et al. (1998a) in a study of fires in central Mexico in April–June 1998. [See also Murray et al. (2000) for further analyses of the relation between these central Mexico fires and the enhanced positive lightning activity in the United States.]

d. Percent positive polarity

The percentage of ground flashes lowering positive charge—ignoring peak currents less than 10 kA—is 8.9% for all 3 years (Table 1), but is seen to have considerable geographic variation in the NALDN (Fig. 5). Values less than 10% occur throughout the western and eastern United States. Higher values of 10%–15% occur in the lower Midwest, and then increase in the upper Midwest to over 20%. These high values, first reported by Orville (1994) for the years 1989–91, had an annual persistence that was verified to occur in subsequent years (Orville and Silver 1997; Orville and Huffines 1999). A 10-yr NLDN summary was published by Orville and Huffines (2001) and now we know that the high percentage of positive flashes is seen to extend into and throughout most of Canada. In fact, a large area of more than 20% positive lightning is seen over British Columbia and the Yukon Territory in the west, and much of Quebec, Labrador, and Newfoundland in the east in all 3 years.

There appears to be considerable interannual variation in Figs. 5a–c. For example, Alberta shows values less than 10% in 1998, but then in subsequent years, increases to over 10% with many areas exceeding 20%. In the 3-yr-average map, Fig. 5d, variation in the percentage of positive lightning parallels the British Columbia–Alberta border. It exceeds 20% to the west of the border and is about 10% to the east. We see that the high percentage of positive lightining in the upper Midwest (Orville and Huffines 2001), as we have noted, does extend into Canada. We also note the high interannual variability in the region of north Texas, particularly when comparing 1998 to subsequent years. This may be the influence of the high percentage of positive lightning associated with the smoke from the Mexican fires, first reported by Lyons et al. (1998a) as well as interannual variability in the number of mesoscale convective systems in the region.

e. Median peak currents

Median peak currents for the 3 years are summarized in Table 1. Peak currents for all positive flashes have a median of 15.3 kA. If we eliminate the positive flashes with peak currents less than 10 kA, the median is 19.8 kA. Negative peak currents for all negative flashes have a median of 16.5 kA. The geographical distribution of the negative median peak currents for the NALDN is shown in Fig. 6. It is immediately apparent that, in general, over the North American landmass the peak negative current in first strokes decreases from lower to higher latitudes, albeit in a discontinuous manner. Over the North American landmass the peak negative current in first strokes decreases northward from 20–24 kA in lower latitudes to a minimum of 12–18 kA in central Canada, then increases north of that to 18–20 kA or more. Note carefully that the relatively high peak currents in northern Mexico are on the fringe of the NALDN, where a natural filtering out of the weaker flashes is caused by range, or distance from the nearest sensors. A similar filtering effect is seen along the northern fringe of the NALDN.

This does not explain, however, the relatively sharp transition from low to high median negative peak currents along the coasts of the United States that are shown in Fig. 6. The sharp transition of the negative peak currents between the land–saltwater interfaces was first noted by Lyons et al (1998b) and confirmed by Orville and Huffines (2001), but we note that it does not occur around freshwater sources such as the Great Lakes. We see in Fig. 6 an interesting freshwater effect along the coast where the Mississippi River flows into the Gulf of Mexico. In this region where the freshwater enters the Gulf ground flashes occur with median peak currents in the range 20–24 kA (yellow). Beyond the mouth of the river, approximately 100 km from the coast, we see that the Gulf is dominated by negative flashes with median peak currents exceeding 24 kA (red). The lower median negative peak currents observed near the mouth of the Mississippi River and extending into the Gulf is a curious effect and contrasts with the sharp “yellow-to-red” transition to higher peak currents that we see along most of the Gulf Coast land–water interface.

Along the Maine coast and further north there is an increase in the peak current, but the magnitude is less compared to other coastal regions. The reason for this apparent change in the recorded peak currents is unknown to us; it is clearly not solely an instrumentation effect. It should be noted that the NALDN infers peak current from measurement of electric and magnetic field strength. It is possible that the relationship between peak current and peak field strength is different for lightning striking saltwater than freshwater (Rakov 2001, personal communication). We do know from the work by Krider et al. (1996; their Fig. 4), for example, that the range-normalized change of the electric field with time is different for signals received from lightning over land (lower) to those received from lightning occurring over ocean (higher). It should be recalled, in addition, that the effect of surface conductivity on the peak magnetic field radiated by first return strokes was investigated by Tyahla and López (1994). They concluded that the conductivity of the underlying surface does not significantly affect the magnitude of the peak magnetic field, and hence, the peak current, in the first return stroke of a cloud-to-ground lightning flash.

In contrast to negative peak currents, the geographical distribution of the median positive peak currents (Fig. 7) shows significant differences. Note that there is not as sharp a transition at the coasts; positive peak current values over land show continuity at the coast over the ocean. As distance from the coasts and from the sensors increases, natural range filtering occurs and higher median peak current values are recorded. Over land the highest median positive peak currents occur in the upper Midwest and Great Plains of the United States and continue into southern Manitoba in Canada. Median positive peak currents less than 15 kA occur in western Canada, primarily in Saskatchewan and Alberta. Other zones of relatively high positive peak currents are seen over the Appalachian Mountains (Fig. 7d). Apparently, the high positive peak currents are uniquely associated with the mesoscale convective systems that occur often in this area and were first reported by Lyons et al. (1998b). Lyons noted that “… concentrations of large positive current CGs in this region are likely associated with at least two classes of thunderstorms known to produce copious number of positive CG's: supercells and nocturnal MCSs.” This is consistent with our observation of a high percentage of positive lightning in the Midwest (Fig. 5) and our observation of maximum ground flash activity at night in the Midwest (Fig. 3). Zajac and Rutledge (2001) investigated ground flash activity over the Midwest and found that “… positive lightning was produced primarily during summer in the hours around sunset by isolated storms and convective lines in various stages of mesoscale convective system development. These convective events usually contained one or more storms that were characterized by predominantly positive lightning, high positive flash rate, and large positive peak currents.” These results indicate that the maximum over the Midwest is caused by intense convective elements—either isolated or organized into convective lines—but not exclusively mesoscale convective systems.

f. Multiplicity

The geographical distributions of mean negative and mean positive multiplicity are shown in Figs. 8 and 9. The highest values of mean negative multiplicity exceed 2.6 strokes per flash in the United States over the Midwest, southeastern states, and northern sections of Montana and North Dakota, and in Canada over the three prairie provinces, northern Ontario, and the Gulf of St. Lawrence. A transition from low (1.5–1.8) to high (greater than 2.6) mean multiplicity occurs at the British Columbia–Alberta border, where we have noted other significant lightning characteristic changes, such as flash density (Fig. 2), the percentage of positive flashes (Fig. 5), and negative median peak currents (Fig. 6). Lower values of mean negative multiplicity are observed to occur throughout the western United States.

Highest values of mean positive multiplicity in the NALDN in Canada are in the same regions in which high mean negative multiplicity occurs. They are a fraction more than one and consistent with the observation that positive flashes most often have one stroke (Beasley 1985). Highest values in the United States are in the southeastern states of Kentucky, Tennessee, the Carolinas, and northern portions of Georgia, Mississippi, and Alabama. This latter region is also characterized by a close spacing of the lightning sensors [see Orville and Huffines (2001), their Figs. 1 and 2], which enhances the sensitivity of the network to subsequent strokes. This latter geographical region of relatively high mean multiplicity is therefore an artifact of the network's enhanced sensitivity due to more strokes per flash being detected. This effect is somewhat reflected in the low median negative peak current (Fig. 6) and high mean multiplicity for negative flashes (Fig. 8).

We have no reasons or hypotheses for the observed high mean negative and positive multiplicity over Alberta and Saskatchewan and other northern regions of the NALDN. Clearly further studies are needed.

4. Discussion and conclusions

The mapping of cloud-to-ground lightning for the first 3 yr of the NALDN provides considerable insight into the large-scale geographical variations of lightning characteristics over the largest region with continuous lightning data studied to date. Previously unknown facts about lightning occurrence in northern latitudes of North America were discovered. A complex pattern of lightning occurrence has been revealed, showing strong geographical, diurnal, and interannual variations. There are substantial influences by elevated terrain features and major land–water boundaries. The association of lightning ground flash density with terrain is a compelling relationship when examining the results for the first 3 yr of operation of the NALDN.

The large north–south extent of the NALDN provides us with the opportunity to add to our knowledge of the characteristics of lightning as a function of latitude following the work of Orville (1990) and Mackerras and Darveniza (1994). A comparison of the negative multiplicity (Fig. 8) with the negative peak current (Fig. 6) shows that in general the multiplicity increases with higher negative peak currents. This relationship does not hold up in Alberta and Saskatchewan. To examine this further, we plotted the median peak current (both positive and negative) versus the multiplicity for the NLDN and for the CLDN. The result is Fig. 10 and it shows clearly that the negative median peak current increases with multiplicity for both networks. On the other hand, the positive median peak current decreases with multiplicity for both networks. Similar results have been reported at the individual flash level by Cummins et al. (1999) and in Austria, where cloud-to-ground flash and stroke detection efficiency are quite high (G. Diendorfer and W. Schulz 2001, personal communication). These findings might suggest that the key factors that determine the charge in the lower portion of the lightning channel are also related to the total charge available for producing a flash. We also note in Fig. 6 that the median negative peak current is generally higher at lower latitudes; a relationship first suggested when the lightning network only existed along the east coast of the United States (Orville 1990).

In the future we will have more data from the continuing operation of the NALDN and we will be in a better position to resolve some of the network problems that can, for the moment, only be identified in this paper. One of the most important questions we need to resolve is the problem of contamination of the ground flash densities with intracloud lightning. At the moment, we exclude the positive lightning flashes that have a peak current less than 10 kA, based primarily on the recommendation of Cummins et al. (1998), who stated, “We recommend that the subset of small positive discharges with peak currents less than 10 kA be regarded as cloud discharges unless they are verified to be cloud-to-ground.” In a paper published in the subsequent year, Wacker and Orville (1999a) provided experimental data and analysis supporting this recommendation. On the other hand, Cooray (1997) suggests that a cloud flash dissipates more energy than a ground flash, posing the question of whether a simple 10-kA filter is adequate to discriminate against cloud flashes in the NALDN.

One of the more curious observations that has so far defied our understanding is the change in estimated negative peak current at the coastal interface, particularly at the NLDN interface. Note that the negative median peak current in Fig. 6 undergoes an abrupt change from land to sea, from less than 24 kA to more than 24 kA. This is underscored by the geographical area where the Mississippi River freshwater flows into the saltwater of the Gulf. In contrast, median peak current changes much more steadily at the coast for positive first strokes (Fig. 7). In fact, in Fig. 7 there is no abrupt change in the positive median peak current. We know of no reason why the peak current should change at the land–sea interface, where this change is occurring for negative but not for positive flashes.

Acknowledgments

The lightning data were obtained from Global Atmospherics, Inc., Tucson, Arizona. We thank Barry Greer of the Meteorological Service of Canada for releasing the Canadian data for this research. We thank Barbara Orville for her editorial assistance. Data handling at Texas A&M University is under the direction of Jerry Guynes and we thank him for his assistance. This research is part of a lightning program supported by the National Science Foundation (ATM-9806189 and ATM-0119476) and the National Oceanic and Atmospheric Administration (Cooperative Agreement NA17WA1011).

REFERENCES

  • Beasley, W., 1985: Positive cloud-to-ground lightning observations. J. Geophys. Res., 90 ((D4),) 61316138.

  • Burrows, W. R., , and P. King, 2000: Neuro-statistical models for predicting lightning occurrence in Canada: Climatology and potential predictors. Proc. 2000 Int. Lightning Detection Conf., Tucson, AZ, Global Atmospherics, Inc., Paper 15. [Available from Global Atmospherics, Inc., 2705 E. Medina, Tucson, AZ 85706.].

    • Search Google Scholar
    • Export Citation
  • Burrows, W. R., , P. J. Lewis, , B. Kochtubajda, , B. Snyder, , and V. Turcotte, 2002: Lightning occurrence patterns over Canada and adjacent United States from lightning detection network observations. Atmos.–Ocean,40, 59–80.

    • Search Google Scholar
    • Export Citation
  • Cooray, V., 1997: Energy dissipation in lightning flashes. J. Geophys. Res., 102 ((D17),) 2140121410.

  • Cummins, K. L., , M. J. Murphy, , E. A. Bardo, , W. L. Hiscox, , R. B. Pyle, , and A. E. Pifer, 1998: A combined TOA/MDF technology upgrade of the U.S. National Lightning Detection Network. J. Geophys. Res., 103 ((D8),) 90359044.

    • Search Google Scholar
    • Export Citation
  • Cummins, K. L., , R. B. Pyle, , and G. Fournier, 1999: An integrated North American Lightning Detection Network. Proc. 11th Int. Conf. on Atmospheric Electricity, Guntersville, AL, NASA/CP-1999-209261, 218–221. [Available from NASA Center for AeroSpace Information, 800 Elkridge Landing Road, Linthicum Heights, MD 21090-2934.].

    • Search Google Scholar
    • Export Citation
  • Fosdick, E. K., , and A. I. Watson, 1995: Cloud-to-ground lightning patterns in New Mexico during the summer. Natl. Wea. Dig., 19 , 1724.

    • Search Google Scholar
    • Export Citation
  • Idone, V. P., , D. A. Davis, , P. K. Moore, , Y. Wang, , R. W. Henderson, , M. Ries, , and P. F. Jamason, 1998a: Performance evaluation of the U.S. National Lightning Detection Network in eastern New York. Part I: Detection efficiency. J. Geophys. Res., 103 ((D8),) 90459055.

    • Search Google Scholar
    • Export Citation
  • Idone, V. P., . 1998b: Performance evaluation of the U.S. National Lightning Detection Network in eastern New York. Part II: Location accuracy. J. Geophys. Res., 103 ((D8),) 90579069.

    • Search Google Scholar
    • Export Citation
  • Krider, E. P., , C. Leteinturier, , and J. C. Willett, 1996: Submicrosecond fields radiated during the onset of first return strokes in cloud-to-ground lightning. J. Geophys. Res., 101 ((D1),) 15891597.

    • Search Google Scholar
    • Export Citation
  • López, R. E., , and R. L. Holle, 1986: Diurnal and spatial variability of lightning activity in northeastern Colorado and central Florida during the summer. Mon. Wea. Rev., 114 , 12881312.

    • Search Google Scholar
    • Export Citation
  • Lyons, W. A., , T. E. Nelson, , E. R. Williams, , J. Cramer, , and T. Turner, 1998a: Enhanced positive cloud-to-ground lightning in thunderstorms ingesting smoke from fires. Science, 282 , 7781.

    • Search Google Scholar
    • Export Citation
  • Lyons, W. A., , M. Uliasz, , and T. E. Nelson, 1998b: Large peak current cloud-to-ground lightning flashes during the summer months in the contiguous United States. Mon. Wea. Rev., 126 , 22172233.

    • Search Google Scholar
    • Export Citation
  • Mackerras, D., , and M. Darveniza, 1994: Latitudinal variation of lightning occurrence characteristics. J. Geophys. Res., 99 , 1081310821.

    • Search Google Scholar
    • Export Citation
  • Murray, N. D., , R. E. Orville, , and G. R. Huffines, 2000: Effect of pollution from Central American fires on cloud-to-ground lightning. Geophys. Res. Lett., 27 , 22492252.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., 1990: Lightning return stroke peak current variation as a function of latitude. Nature, 342 , 149151.

  • Orville, R. E., . 1991: Lightning ground flash density in the contiguous United States–1989. Mon. Wea. Rev., 119 , 573577.

  • Orville, R. E., . 1994: Cloud-to-ground lightning flash characteristics in the contiguous United States: 1989–1991. J. Geophys. Res., 99 ((D5),) 1083310841.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., , and A. C. Silver:, 1997: Lightning ground flash density in the contiguous United States: 1992–95. Mon. Wea. Rev., 125 , 631638.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., , and G. R. Huffines:, 1999: Lightning ground flash measurements over the contiguous United States: 1995–97. Mon. Wea. Rev., 127 , 26932703.

    • Search Google Scholar
    • Export Citation
  • Orville, R. E., . 2001: Cloud-to-ground lightning in the USA: NLDN results in the first decade, 1989–98. Mon. Wea. Rev., 129 , 11791193.

    • Search Google Scholar
    • Export Citation
  • Reap, R. M., 1986: Evaluation of cloud-to-ground data from the western United States for the 1983–84 summer seasons. J. Climate Appl. Meteor., 25 , 785799.

    • Search Google Scholar
    • Export Citation
  • Tyahla, L. J., , and R. E. López, 1994: Effect of surface conductivity on the peak magnetic field radiated by first return strokes in cloud-to-ground lightning. J. Geophys. Res., 99 ((D5),) 1051710525.

    • Search Google Scholar
    • Export Citation
  • Wacker, R. S., , and R. E. Orville, 1999a: Changes in measured lightning flash count and return stroke peak current after the 1994 U.S. National Lightning Detection Network upgrade. Part I: Observations. J. Geophys. Res., 104 ((D2),) 21512157.

    • Search Google Scholar
    • Export Citation
  • Wacker, R. S., . 1999b: Changes in measured lightning flash count and return stroke peak current after the 1994 U.S. National Lightning Detection Network upgrade. Part II: Theory. J. Geophys. Res., 104 ((D2),) 21592162.

    • Search Google Scholar
    • Export Citation
  • Zajac, B. A., , and S. A. Rutledge, 2001: Cloud-to-ground lightning activity in the contiguous United States from 1995 to 1999. Mon. Wea. Rev., 129 , 9991019.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

The total area covered by the NALDN is shown in light gray. Blue symbols represent the CLDN composed of 87 sensors, and red represents the NLDN of 106 sensors. Each network is composed of IMPACT/ES and TOA sensors. Triangles mark the TOA sensor locations and circles mark IMPACT/ES sensor locations

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 2.
Fig. 2.

Four-panel display shows the annual ground flash density in North America for the individual years (a) 1998, (b) 1999, and (c) 2000, and (d) the mean flash density for the 3 yr combined

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 3.
Fig. 3.

The local hours of maximum ground flash density are plotted for two time increments, from 0800 to 2000 LST in red and from 2000 to 0800 LST in blue. Thus, red shows maximum lightning in daytime and blue is at night

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 4.
Fig. 4.

Same as in Fig. 2 except the positive ground flash density is plotted by color

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 5.
Fig. 5.

Same as in Fig. 2 except the percentage of lightning ground flashes lowering positive charge is color shaded

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 6.
Fig. 6.

Same as in Fig. 2 except the negative median peak current is shaded for ground flashes lowering negative charge

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 7.
Fig. 7.

Same as in Fig. 2 except for positive median peak currents

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 8.
Fig. 8.

Same as in Fig. 2 except for mean negative multiplicity

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 9.
Fig. 9.

Same as in Fig. 2 except for mean positive multiplicity.

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Fig. 10.
Fig. 10.

Median peak current is plotted as a function of the flash multiplicity for each polarity for each of the two networks, the NLDN and the CLDN

Citation: Monthly Weather Review 130, 8; 10.1175/1520-0493(2002)130<2098:TNALDN>2.0.CO;2

Table 1. 

Lightning Summary for the North American Lightning Detection Network (NALDN), 1998–2000

Table 1. 
Save