How big data will bring weather forecasting into the 21st century — reliable, predictable, meaningful

To simply say that the accuracy of weather forecasts is important is an understatement. Weather impacts everything and every person on a global scale: from international supply chains to managing municipal road salt usage to deciding what to wear each morning. The unfortunate reality is that while weather modeling and simulations have continued to advance (the output), the amount of raw data available for analysis has consistently dwindled (the input). Imagine a powerful machine with no fuel to keep it running, let alone enough to experience its true potential. The need for increased weather data collection is at an all-time high; a significant majority of raw data that we rely upon today for weather analysis across the globe comes from roughly 20 satellites, and most of those same satellites are past their intended decommission date (running in the red), the likelihood of a catastrophic failure and diminishing reliability increasing with each passing day. The U.S. economy already sees a $500 billion fluctuation in annual GDP due to weather uncertainty. Globally the impact is even larger, particularly in territories where weather prediction and data collection are not as well established. Now imagine the impact as the world's forecast simulations and models become increasingly starved for data.