Bolt out of the blue

May 10th, 2022 |

Florida is one of the lightning capitals of the world, so residents need to be constantly aware of lightning safety. NOAA/CIMSS LightningCast might be able to help with that.

A tree and home in Sebring, Florida were suddenly struck by lightning on the morning of Saturday, May 7th. A line of storms was edging its way eastward. A neighbor who was outside at the time said, “It wasn’t raining. It was nice and warm. It was cloudy, but that was it. And then boom!” This underscores how easy it is to be caught unaware of potential lightning danger.

LightningCast can help with users’ situational awareness. LightningCast is an experimental deep-learning model trained on thousands of GOES-R ABI and GLM images to predict the probability of next-hour lightning occurrence. In the animation below, the red dot is Sebring, Florida.

LightningCast probability contours, GOES-16 ABI imagery (grayscale background), and GOES-16 GLM flash-extent density (shaded color). The red dot is the approximate location of the “bolt out of the blue”.
Florida homeowner stunned by nearby lightning strike. Credit: FOX13 Tampa Bay

Below is a time series of LightningCast probability and GLM observations around the home in Sebring. Lightning struck the tree and home at 8:21 EDT, marked by the vertical black line below. You can see a rapid increase in probability of lightning from 11:26 to 11:36 UTC (7:26 to 7:36 EDT), reaching 70%. This was about 25 minutes before the first nearby lightning strike and 45 minutes before the Sebring home was struck.

The animation below from the National Weather Service lightning safety page shows that most lightning casualties occur before a thunderstorm is fully overhead, or before it fully departs the area, when people might not realize their vulnerability to lightning and don’t seek shelter soon enough or leave shelter too soon.

Animation depicting the threat of lightning casualties as a function of a hypothetical storm moving into the area.

Using GOES ABI and deep learning to nowcast lightning

September 2nd, 2020 |

NOAA and CIMSS are developing a product that uses a deep-learning model to recognize complex patterns in weather satellite imagery to predict the probability of lightning in the short term. Deep learning is a branch of machine learning based on artificial neural networks, which have the ability to automatically learn targeted features in the data by approximating how humans learn.

A convolutional neural network (CNN) was trained on over 23,000 images of GOES-16 Advanced Baseline Imager (ABI) data to predict the probability of lightning, within any given ABI pixel, in the next 60 minutes. The CNN was trained using 118 days of data collected between May and August of 2018. Images of GOES Geostationary Lightning Mapper (GLM) flash-extent density (created with glmtools) were used as the source of lightning observations. Note that GLM is an optical sensor that observes both in-cloud and cloud-to-ground lightning.

The CNN currently uses four ABI channels: band 2 (0.64-µm), band 5 (1.6-µm), band 13 (10.3-µm), and band 15 (12.3-µm). Bands 2 and 5 are only utilized under sunlit conditions. Utilization of additional channels and time sequences of images is under investigation. The model uses a semantic image segmentation architecture to assign the probability of lighting in the next 60 minutes to each pixel in the image. The model is very computationally efficient, only needing 30 seconds to process the ABI CONUS domain and 3 seconds to process an ABI mesoscale domain using multithreading on a 40-CPU linux server.

Currently, the model only utilizes satellite radiances. Thus, it can be applied to nearly any spatial domain covered by the ABI or an ABI-like sensor (e.g. AHI). Based on near-real-time testing, the model routinely nowcasts lightning initiation with 10-30 minutes of lead-time. We expect the skill and lead-time will increase as new predictors (e.g. more ABI fields, NWP, radar where available) are added to the model.

Below are a sampling of recent examples. The base images are 0.64-µm reflectance, with GLM-derived flash-extent density overlaid as filled semi-transparent polygons. The flash-extent density is the accumulated number of flashes within the previous 5 minutes. The CNN-derived probabilities are displayed as contours at select levels (near-real-time output is available through RealEarth).

The overall objective is to improve lightning nowcasts in support of aviation, mariners, and outdoor events/activities. Beyond improving the CNN, our work will focus on packaging the output into actionable information for forecasters and other decision makers.

A cold front in Iowa


Thunderstorm development on sea-breeze boundaries in Florida and the Bahamas


Diurnally and orographically forced storms in the Southwest U.S. and Rocky Mountains


Storms in central Oklahoma, on the edge of Hurricane Laura’s cloud shield


A couple of examples over the Northeast U.S.


A boundary of convection on the southern bank of Lake Ontario


Storms in a warm sector in IL/IN/OH, perhaps along an outflow boundary


Southeast U.S. offshore region

The background image in this example transitions from 10.3-µm brightness temperature to 0.64-µm reflectance, while the flash-extent density enhancement also changes, in an attempt to enhance contrast.