Slide 1

Slide 1 text

Lightning Activity Predictions For Single Buoy Moorings Andrew B. Collier Exegetic Analytics [email protected] http://1.bp.blogspot.com/

Slide 2

Slide 2 text

2 Oceanic lightning... … is difficult to investigate. Brooks, C. E. P. (1925). The Distribution of Thunderstorms over the Globe. Geophysical Memoirs, 3(24), 147–164. “... not much happening over the oceans...”

Slide 3

Slide 3 text

3 Trent, E. M., & Gathman, S. G. (1972). Oceanic thunderstorms. Pure and Applied Geophysics, 100(1), 60–69. doi:10.1007/BF00880227. ocean:land area ratio between 3:1 and 1:1 72% of the Earth covered by ocean sparse lightning over ocean

Slide 4

Slide 4 text

5 Oceanic lightning... … was difficult to investigate.

Slide 5

Slide 5 text

6 Christian, H. J., Blakeslee, R. J., Boccippio, D. J., Boeck, W. L., Buechler, D. E., Driscoll, K. T., Goodman, S. J., et al. (2003). Global frequency and distribution of lightning as observed from space by the Optical Transient Detector. Journal of Geophysical Research, 108(D1). doi:10.1029/2002JD002347.

Slide 6

Slide 6 text

7 Füllekrug, M., Price, C., Yair, Y., & Williams, E. R. (2002). Intense oceanic lightning. Annales Geophysicae, 20, 133–137. doi:10.5194/angeo- 20-133-2002

Slide 7

Slide 7 text

8 World Wide Lightning Location Network (WWLLN)

Slide 8

Slide 8 text

9 http://www.lankhorstropes.com/files/uploads/Offshore/scenarios_TLWPwFPSO_SBM_Atlantia.jpg SBM = Single Buoy Mooring

Slide 9

Slide 9 text

10

Slide 10

Slide 10 text

11

Slide 11

Slide 11 text

12 http://samsa.org.za/

Slide 12

Slide 12 text

13

Slide 13

Slide 13 text

14 WWLLN data partitioned into ● 15 min time bins ● 10, 20, 50, 100 and 200 km ● 4 quadrants Model ● Input: lightning counts ● Output: lightning presence

Slide 14

Slide 14 text

16 A B1 B2 B3 B4 C1 C2 C3 C4 D1 D2 D3 D4 E1 E2 E3 E4 2506 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2507 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2509 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2726 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2729 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2733 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2796 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2880 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 2882 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 2884 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2886 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 2887 FALSE 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 1 2890 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 2891 FALSE 0 0 0 0 0 0 2 0 0 2 3 0 0 1 0 0 2892 FALSE 0 0 0 0 0 0 3 0 0 0 2 0 0 1 0 0 2893 TRUE 0 0 0 0 0 0 1 1 0 5 2 4 0 1 0 1 2894 TRUE 0 0 3 0 0 0 4 5 0 3 1 2 0 3 0 0 2895 TRUE 0 0 3 4 0 2 1 6 1 3 0 1 0 17 0 0 2896 TRUE 2 0 0 2 0 11 1 0 0 1 0 0 0 23 0 0 2897 FALSE 0 0 0 0 2 12 0 0 1 18 0 1 0 6 0 0 2898 FALSE 0 0 0 0 3 4 0 0 17 15 0 3 0 7 0 1 2899 FALSE 1 0 0 0 0 1 0 0 16 29 0 0 0 8 0 0 2900 FALSE 0 0 0 0 1 0 0 0 14 30 0 0 5 14 0 2 2901 FALSE 0 0 0 0 1 0 0 0 16 28 0 0 5 11 0 0 2902 FALSE 0 0 0 0 0 1 0 0 14 8 0 0 12 41 1 0 2903 FALSE 0 0 0 0 0 0 0 0 5 5 0 0 15 64 0 1 2904 FALSE 0 0 0 0 0 0 0 0 3 0 0 0 24 59 0 1 2905 FALSE 0 0 0 0 0 0 0 0 1 0 0 0 45 32 1 0 2906 FALSE 0 0 0 0 0 0 0 0 0 0 0 0 26 17 0 0 2907 FALSE 0 0 0 0 0 0 0 0 0 1 0 0 66 10 0 0 2908 FALSE 0 0 0 0 0 0 0 0 0 1 0 0 31 2 0 0

Slide 15

Slide 15 text

18 Confusion Matrices Neural Network observed predicted FALSE TRUE FALSE 12681 65 TRUE 15 48 Conditional Inference Tree observed predicted FALSE TRUE FALSE 12690 92 TRUE 6 21 Random Forest observed predicted FALSE TRUE FALSE 12696 32 TRUE 0 81 Support Vector Machine observed predicted FALSE TRUE FALSE 12696 22 TRUE 0 91 false positive false negative

Slide 16

Slide 16 text

19 http://www.davey.com/elements/skin/home_tree.png

Slide 17

Slide 17 text

20

Slide 18

Slide 18 text

21

Slide 19

Slide 19 text

22 Random Forest Results

Slide 20

Slide 20 text

24 Alexander Stepanovich Popov 1895

Slide 21

Slide 21 text

25 … susceptible to type I errors.

Slide 22

Slide 22 text

26 Conclusion ● Lightning warnings → disaster prevention ● Few false positives ● Multiple applications