The Need for HPC for Deep Learning with Real-Time Satellite Observations

With the recent launch of new satellites like the Geostationary Operational Environmental Satellite (GOES)-16, the data volume has increased by orders of magnitude and can be difficult to process in a timely manner using traditional methods. Additionally, we have seen a rapid growth of Machine Learning and in particular Deep Learning applications across a variety of fields. Deep learning has shown promising advancements to significantly improve both processing speed and scientific accuracy of results. Our team at the National Oceanic Atmospheric Administration (NOAA) Earth System Research Laboratory (ESRL) is researching the application deep learning techniques for the processing and extracting new value from satellite observation data.

Through our research, we are using a Convolution Neural Network (CNN) to identify regions of interest (ROI) from satellite observations. These areas include cyclones, both tropical and extratropical, cyclogenesis, and eventually convection initiation. Additionally, our team has started preliminary research into the use of CNN’s to generate higher spatial and temporal resolution soil moisture product from satellite radiance observations to improve soil moisture within atmospheric prediction models at the initial time.

This presentation will provide an overview of our research efforts into the application of deep learning, the tradeoffs on computing and accuracy when designing neural networks, along with the challenges of data preparation including creation of “labeled” data, training the model, and where we see these applications heading into the future.