The Cooperative Institute for Meteorological Satellite Studies (CIMSS) at the University of Wisconsin (UWisc) provides near real-time imagery from different satellite platforms for analysis of tropical storms. This data is used to make weather predictions for tropical cyclones. Recently, CIMSS posted about a machine learning approach (developed at UWisc) for predicting the intensity of tropical cyclones. The model, called “DeepMicroNet”, is a deep learning convolutional neural network (CNN) trained on various microwave images of tropical cyclones. The model utilizes images in the 37- and 85-92 GHz microwave bands to make predictions about storm intensity. Training data comes from past forecaster intensity predictions made from other satellite data and aircraft observations. The resulting model was able to predict intensities within 11.5 miles per hour. This is near current state of the art predictions which can be off by as much as 10 miles per hour.
When predicting a hurricane’s maximum sustained winds, DeepMicroNet’s results differed from the historical record of forecaster-estimated values by about 16 miles per hour. DeepMicroNet’s results improved, however, when the data sets were limited to data measured directly by aircraft. Then, DeepMicroNet was off by less than 11.5 miles per hour. By comparison, estimates using state-of-the-art methods are typically off by around 10 miles per hour.
Checkout the abstract below:
A deep learning convolutional neural network model is used to explore the possibilities of estimating tropical cyclone (TC) intensity from satellite images in the 37- and 85–92-GHz bands. The model, called “DeepMicroNet,” has unique properties such as a probabilistic output, the ability to operate from partial scans, and resiliency to imprecise TC center fixes. The 85–92-GHz band is the more influential data source in the model, with 37 GHz adding a marginal benefit. Training the model on global best track intensities produces model estimates precise enough to replicate known best track intensity biases when compared to aircraft reconnaissance observations. Model root-mean-square error (RMSE) is 14.3 kt (1 kt ≈ 0.5144 m s−1) compared to two years of independent best track records, but this improves to an RMSE of 10.6 kt when compared to the higher-standard aircraft reconnaissance-aided best track dataset, and to 9.6 kt compared to the reconnaissance-aided best track when using the higher-resolution TRMM TMI and Aqua AMSR-E microwave observations only. A shortage of training and independent testing data for category 5 TCs leaves the results at this intensity range inconclusive. Based on this initial study, the application of deep learning to TC intensity analysis holds tremendous promise for further development with more advanced methodologies and expanded training datasets.
If you would like to learn more about this work check out the publication titled, “Using Deep Learning to Estimate Tropical Cyclone Intensity from Satellite Passive Microwave Imagery”. If you would like to learn more about models for predicting tropical storms, checkout this presentation by NASA titled, “Tropical Cyclone Intensity Estimation Using Deep Convolutional Neural Networks”.