• Abadi, M., and Coauthors, 2016: Tensorflow: A system for large-scale machine learning. Proc. 12th USENIX Symp. on Operating Systems Design and Implementation (OSDI’16), Savannah, GA, USENIX 265–283, https://www.usenix.org/system/files/conference/osdi16/osdi16-abadi.pdf.

  • Adams-Selin, R. D., and C. L. Ziegler, 2016: Forecasting hail using a one-dimensional hail growth model within WRF. Mon. Wea. Rev., 144, 49194939, https://doi.org/10.1175/MWR-D-16-0027.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Adams-Selin, R. D., A. J. Clark, C. J. Melick, S. R. Dembek, I. L. Jirak, and C. L. Ziegler, 2019: Evolution of WRF-HAILCAST during the 2014–16 NOAA/Hazardous Weather Testbed spring forecasting experiments. Wea. Forecasting, 34, 6179, https://doi.org/10.1175/WAF-D-18-0024.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Allen, J. T., and M. K. Tippett, 2015: The characteristics of United States hail reports: 1955–2014. Electron. J. Severe Storms Meteor., 10 (3), http://www.ejssm.org/ojs/index.php/ejssm/article/viewArticle/149.

  • Anderson-Frey, A. K., Y. P. Richardson, R. L. Thompson, and B. T. Smith, 2017: Self-organizing maps for the investigation of tornadic near-storm environments. Wea. Forecasting, 32, 14671475, https://doi.org/10.1175/WAF-D-17-0034.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Breiman, L., 2001: Random forests. Mach. Learn., 45, 532, https://doi.org/10.1023/A:1010933404324.

  • Brimelow, J. C., G. W. Reuter, and E. R. Poolman, 2002: Modeling maximum hail size in Alberta thunderstorms. Wea. Forecasting, 17, 10481062, https://doi.org/10.1175/1520-0434(2002)017<1048:MMHSIA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Brimelow, J. C., G. W. Reuter, R. Goodson, and T. W. Krauss, 2006: Spatial forecasts of maximum hail size using prognostic model soundings and hailcast. Wea. Forecasting, 21, 206219, https://doi.org/10.1175/WAF915.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Buell, C. E., 1975: The topography of empirical orthogonal functions. Preprints Fourth Conf. on Probability and Statistics in Atmospheric Sciences, Tallahassee, FL, Amer. Meteor. Soc., 188–193.

  • Buell, C. E., 1979: On the physical interpretation of empirical orthogonal functions. Preprints, Sixth Conf. on Probability and Statistics in Atmospheric Sciences, Banff, Alberta, Canada, Amer. Meteor. Soc., 112–117.

  • Byers, H. R., and H. R. Rodebush, 1948: Causes of thunderstorms of the Florida peninsula. J. Meteor., 5, 275280, https://doi.org/10.1175/1520-0469(1948)005<0275:COTOTF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569585, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chollet, F., and Coauthors, 2015: Keras: The Python Deep Learning Library. Accessed 31 August 2018, https://keras.io.

  • Cintineo, J. L., T. M. Smith, V. Lakshmanan, H. E. Brooks, and K. L. Ortega, 2012: An objective high-resolution hail climatology of the contiguous United States. Wea. Forecasting, 27, 12351248, https://doi.org/10.1175/WAF-D-11-00151.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Computational and Information Systems Laboratory, 2017: Cheyenne: HPE/SGI ICE XA System (NCAR Community Computing). Tech. Rep., National Center for Atmospheric Research, Boulder, CO, accessed 1 September 2018, https://doi.org/10.5065/D6RX99HX.

    • Crossref
    • Export Citation
  • Dennis, E. J., and M. R. Kumjian, 2017: The impact of vertical wind shear on hail growth in simulated supercells. J. Atmos. Sci., 74, 641663, https://doi.org/10.1175/JAS-D-16-0066.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Edwards, R., and R. L. Thompson, 1998: Nationwide comparisons of hail size with WSR-88D vertically integrated liquid water and derived thermodynamic sounding data. Wea. Forecasting, 13, 277285, https://doi.org/10.1175/1520-0434(1998)013<0277:NCOHSW>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Foote, G. B., 1984: A study of hail growth utilizing observed storm conditions. J. Climate Appl. Meteor., 23, 84101, https://doi.org/10.1175/1520-0450(1984)023<0084:ASOHGU>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gagne, D. J., 2019: Interpretable deep learning for spatial analysis of severe hailstorms: Storm and analysis data. Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory, https://doi.org/10.5065/6CJA-B154.

    • Crossref
    • Export Citation
  • Gagne, D. J., A. McGovern, S. E. Haupt, R. A. Sobash, J. K. Williams, and M. Xue, 2017: Storm-based probabilistic hail forecasting with machine learning applied to convection-allowing ensembles. Wea. Forecasting, 32, 18191840, https://doi.org/10.1175/WAF-D-17-0010.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Goodfellow, I., Y. Bengio, and A. Courville, 2016: Deep Learning. MIT Press, 775 pp.

  • Grant, L. D., and S. C. van den Heever, 2014: Microphysical and dynamical characteristics of low-precipitation and classic supercells. J. Atmos. Sci., 71, 26042624, https://doi.org/10.1175/JAS-D-13-0261.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Greybush, S., S. Haupt, and G. Young, 2008: The regime dependence of optimally weighted ensemble model consensus forecasts of surface temperature. Wea. Forecasting, 23, 11461161, https://doi.org/10.1175/2008WAF2007078.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Haberlie, A. M., and W. S. Ashley, 2018: A method for identifying midlatitude mesoscale convective systems in radar mosaics. Part I: Segmentation and classification. J. Appl. Meteor. Climatol., 57, 15751598, https://doi.org/10.1175/JAMC-D-17-0293.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heymsfield, A. J., A. R. Jameson, and H. W. Frank, 1980: Hail growth mechanisms in a Colorado storm: Part II: Hail formation processes. J. Atmos. Sci., 37, 17791807, https://doi.org/10.1175/1520-0469(1980)037<1779:HGMIAC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ioffe, S., and C. Szegedy, 2015: Batch normalization: Accelerating deep network training by reducing internal covariate shift. Proc. 32nd Int. Conf. on Machine Learning, Vol. 37, Lille, France, PMLR, 448–456, http://proceedings.mlr.press/v37/ioffe15.html.

  • Jewell, R., and J. Brimelow, 2009: Evaluation of Alberta hail growth model using severe hail proximity soundings from the United States. Wea. Forecasting, 24, 15921609, https://doi.org/10.1175/2009WAF2222230.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Johnson, A. W., and K. E. Sugden, 2014: Evaluation of sounding-derived thermodynamic and wind-related parameters associated with large hail events. Electron. J. Severe Storms Meteor., 9 (5), http://www.ejssm.org/ojs/index.php/ejssm/article/viewArticle/137.

  • Kingma, D. P., and J. Ba, 2015: Adam: A method for stochastic optimization. Int. Conf. on Learning Representations, San Diego, CA, https://arxiv.org/abs/1412.6980.

  • Kohonen, T., 1982: Self-organized formation of topologically correct feature maps. Biol. Cybern., 43, 5969, https://doi.org/10.1007/BF00337288.

  • Krizhevsky, A., I. Sutskever, and G. Hinton, 2012: ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems 25 (NIPS 2012), F. Pereira et al. Eds., Neural Information Processing Systems Foundation, Inc., 1097–1105, http://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf.

  • Lagerquist, R., A. McGovern, and T. Smith, 2017: Machine learning for real-time prediction of damaging straight-line convective wind. Wea. Forecasting, 32, 21752193, https://doi.org/10.1175/WAF-D-17-0038.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lakshmanan, V., and T. Smith, 2009: Data mining storm attributes from spatial grids. J. Atmos. Oceanic Technol., 26, 23532365, https://doi.org/10.1175/2009JTECHA1257.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lakshmanan, V., K. Hondl, and R. Rabin, 2009: An efficient, general-purpose technique for identifying storm cells in geospatial images. J. Atmos. Oceanic Technol., 26, 523537, https://doi.org/10.1175/2008JTECHA1153.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • LeCun, Y., B. E. Boser, J. S. Denker, D. Henderson, R. Howard, W. Hubbard, and L. Jackel, 1990: Handwritten digit recognition with a back-propagation network. Advances in Neural Information Processing Systems, D. S. Touretzky, Ed., Morgan Kaufmann Publishers, Inc., 396–404, http://papers.nips.cc/paper/293-handwritten-digit-recognition-with-a-back-propagation-network.pdf.

  • LeCun, Y., Y. Bengio, and G. Hinton, 2015: Deep learning. Nature, 521, 436444, https://doi.org/10.1038/nature14539.

  • Manzato, A., 2012: Hail in northeast Italy: Climatology and bivariate analysis with the sounding-derived indices. J. Appl. Meteor. Climatol., 51, 449467, https://doi.org/10.1175/JAMC-D-10-05012.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mason, I. B., 1982: A model for assessment of weather forecasts. Aust. Meteor. Mag., 30, 291303.

  • McGovern, A., D. J. Gagne, J. K. Williams, R. A. Brown, and J. B. Basara, 2014: Enhancing understanding and improving prediction of severe weather through spatiotemporal relational learning. Mach. Learn., 95, 2750, https://doi.org/10.1007/s10994-013-5343-x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mellor, G. L., and T. Yamada, 1982: Developement of turbulence closure model for geophysical fluid problems. Rev. Geophys. Space Phys., 20, 851875, https://doi.org/10.1029/RG020i004p00851.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 66316 682, https://doi.org/10.1029/97JD00237.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morrison, H., and J. Milbrandt, 2011: Comparison of two-moment bulk microphysics schemes in idealized supercell thunderstorm simulations. Mon. Wea. Rev., 139, 11031130, https://doi.org/10.1175/2010MWR3433.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murphy, A. H., 1973: A new vector partition of the probability score. J. Appl. Meteor., 12, 595600, https://doi.org/10.1175/1520-0450(1973)012<0595:ANVPOT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nelson, S. P., 1983: The influence of storm flow structure on hail growth. J. Atmos. Sci., 40, 19651983, https://doi.org/10.1175/1520-0469(1983)040<1965:TIOSFS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nelson, S. P., 1987: The hybrid multicellular-supercellular storm—An efficient hail producer. Part II: General characteristics and implications for hail growth. J. Atmos. Sci., 44, 20602073, https://doi.org/10.1175/1520-0469(1987)044<2060:THMSEH>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nowotarski, C. J., and A. A. Jensen, 2013: Classifying proximity soundings with self-organizing maps toward improving supercell and tornado forecasting. Wea. Forecasting, 28, 783801, https://doi.org/10.1175/WAF-D-12-00125.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Odena, A., V. Dumoulin, and C. Olah, 2016: Deconvolution and checkerboard artifacts. Distill, accessed 15 July 2018, https://distill.pub/2016/deconv-checkerboard/.

    • Crossref
    • Export Citation
  • Olah, C., A. Mordvintsev, and L. Schubert, 2017: Feature visualization: How neural networks build up their understanding of images. Distill, accessed 15 July 2018, https://distill.pub/2017/feature-visualization/.

    • Crossref
    • Export Citation
  • Pearson, K., 1901: On lines and planes of closest fit to systems of points in space. Philos. Mag., 2, 559572, https://doi.org/10.1080/14786440109462720.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pedregosa, F., and Coauthors, 2011: Scikit-learn: Machine Learning in Python. J. Mach. Learn. Res., 12, 28252830.

  • Púčik, T., P. Groenemeijer, D. Rýva, and M. Kolář, 2015: Proximity soundings of severe and nonsevere thunderstorms in central Europe. Mon. Wea. Rev., 143, 48054821, https://doi.org/10.1175/MWR-D-15-0104.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rasmussen, R. M., and A. J. Heymsfield, 1987: Melting and shedding of graupel and hail. Part I: Model physics. J. Atmos. Sci., 44, 27542763, https://doi.org/10.1175/1520-0469(1987)044<2754:MASOGA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Richman, M. B., 1986: Rotation of principal components. Int. J. Climatol., 6, 293335, https://doi.org/10.1002/joc.3370060305.

  • Richman, M. B., 1993: Comments on “The effect of domain shape on principal components analysis.” Int. J. Climatol., 13, 203218, https://doi.org/10.1002/joc.3370130206.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Richman, M. B., and P. J. Lamb, 1985: Climatic pattern analysis of three- and seven-day summer rainfall in the central United States: Some methodological considerations and a regionalization. J. Climate Appl. Meteor., 24, 13251343, https://doi.org/10.1175/1520-0450(1985)024<1325:CPAOTA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roebber, P. J., 2009: Visualizing multiple measures of forecast quality. Wea. Forecasting, 24, 601608, https://doi.org/10.1175/2008WAF2222159.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schwartz, C. S., G. S. Romine, R. A. Sobash, K. Fossell, and M. L. Weisman, 2015: NCAR’s experimental real-time convection-allowing ensemble prediction system. Wea. Forecasting, 30, 16451654, https://doi.org/10.1175/WAF-D-15-0103.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Simoyan, K., A. Vedaldi, and A. Zisserman, 2014: Deep inside convolutional networks: Visualising image classification models and saliency maps. Int. Conf. on Learning Representations Workshop, Banff, Canada, https://arxiv.org/abs/1312.6034.

  • Skamarock, W. C., and J. B. Klemp, 2008: A time-split nonhydrostatic atmospheric model for weather research and forecasting applications. J. Comput. Phys., 227, 34653485, https://doi.org/10.1016/j.jcp.2007.01.037.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Srivastava, N., G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, 2014: Dropout: A simple way to prevent neural networks from overfitting. J. Mach. Learn. Res., 15, 19291958.

    • Search Google Scholar
    • Export Citation
  • Storm Prediction Center, 2019: Storm Prediction Center WCM Page. NOAA, accessed 1 February 2019, https://www.spc.noaa.gov/wcm/.

  • Thompson, G., and T. Eidhammer, 2014: A study of aerosol impacts on clouds and precipitation development in a large winter cyclone. J. Atmos. Sci., 71, 36363658, https://doi.org/10.1175/JAS-D-13-0305.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Thompson, G., R. M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev., 132, 519542, https://doi.org/10.1175/1520-0493(2004)132<0519:EFOWPU>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Thompson, G., P. R. Field, R. M. Rasmussen, and W. D. Hall, 2008: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: Implementation of a new snow parameterization. Mon. Wea. Rev., 136, 50955115, https://doi.org/10.1175/2008MWR2387.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tibshirani, R., 1996: Regression shrinkage and selection via the lasso. J. Roy. Stat. Soc. B, 58, 267288.

  • Wilks, D. S., 2011: Statistical Methods in the Atmospheric Sciences. 3rd ed. International Geophysics Series, Vol. 100, Academic Press, 704 pp.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 90 90 90
PDF Downloads 76 76 76

Interpretable Deep Learning for Spatial Analysis of Severe Hailstorms

View More View Less
  • 1 National Center for Atmospheric Research, Boulder, Colorado
Restricted access

Abstract

Deep learning models, such as convolutional neural networks, utilize multiple specialized layers to encode spatial patterns at different scales. In this study, deep learning models are compared with standard machine learning approaches on the task of predicting the probability of severe hail based on upper-air dynamic and thermodynamic fields from a convection-allowing numerical weather prediction model. The data for this study come from patches surrounding storms identified in NCAR convection-allowing ensemble runs from 3 May to 3 June 2016. The machine learning models are trained to predict whether the simulated surface hail size from the Thompson hail size diagnostic exceeds 25 mm over the hour following storm detection. A convolutional neural network is compared with logistic regressions using input variables derived from either the spatial means of each field or principal component analysis. The convolutional neural network statistically significantly outperforms all other methods in terms of Brier skill score and area under the receiver operator characteristic curve. Interpretation of the convolutional neural network through feature importance and feature optimization reveals that the network synthesized information about the environment and storm morphology that is consistent with our understanding of hail growth, including large lapse rates and a wind shear profile that favors wide updrafts. Different neurons in the network also record different storm modes, and the magnitude of the output of those neurons is used to analyze the spatiotemporal distributions of different storm modes in the NCAR ensemble.

© 2019 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: David John Gagne II, dgagne@ucar.edu

Abstract

Deep learning models, such as convolutional neural networks, utilize multiple specialized layers to encode spatial patterns at different scales. In this study, deep learning models are compared with standard machine learning approaches on the task of predicting the probability of severe hail based on upper-air dynamic and thermodynamic fields from a convection-allowing numerical weather prediction model. The data for this study come from patches surrounding storms identified in NCAR convection-allowing ensemble runs from 3 May to 3 June 2016. The machine learning models are trained to predict whether the simulated surface hail size from the Thompson hail size diagnostic exceeds 25 mm over the hour following storm detection. A convolutional neural network is compared with logistic regressions using input variables derived from either the spatial means of each field or principal component analysis. The convolutional neural network statistically significantly outperforms all other methods in terms of Brier skill score and area under the receiver operator characteristic curve. Interpretation of the convolutional neural network through feature importance and feature optimization reveals that the network synthesized information about the environment and storm morphology that is consistent with our understanding of hail growth, including large lapse rates and a wind shear profile that favors wide updrafts. Different neurons in the network also record different storm modes, and the magnitude of the output of those neurons is used to analyze the spatiotemporal distributions of different storm modes in the NCAR ensemble.

© 2019 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: David John Gagne II, dgagne@ucar.edu
Save