Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Lightning Activity Predictions For Single Buoy ...

datawookie
September 12, 2013

Lightning Activity Predictions For Single Buoy Moorings

A short talk that I gave at the LIGHTS 2013 Conference (Johannesburg, 12 September 2013). The slides are a little short on text because I like the audience to hear the content rather than read it. The objective with this project was to develop a model which would predict the occurrence of lightning in the vicinity of a Single Buoy Mooring (SBM). I used data from the World Wide Lightning Location Network (WWLLN). I considered four possible models: Neural Network, Conditional Inference Tree, Support Vector Machine and Random Forest. Of the four, Random Forests produced the best performance. The preliminary results of the model are very promising: there is good agreement between predicted and observed lightning occurrence in the vicinity of the SBM.

datawookie

September 12, 2013
Tweet

More Decks by datawookie

Other Decks in Science

Transcript

  1. 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...”
  2. 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
  3. 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.
  4. 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
  5. 10

  6. 11

  7. 13

  8. 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
  9. 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
  10. 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
  11. 20

  12. 21