The approach of hurricane season underscores the importance of reliable, timely weather forecasting. Unfortunately, the U.S suffers from an inadequate weather model and declining weather satellite capabilities. The European weather forecasting model, known as the European Centre for Medium-Range Weather Forecasts (ECMWF), outperformed the weather model of the US National Weather Service (NWS). ECMWF predicted Winter Storm Nemo, one of the biggest snowstorms in history, three days before the weather model of the US NWS did. ECMWF also predicted that Hurricane Sandy was likely to come ashore while the US NWS model was predicting that the storms were likely to go out to sea. Ultimately the weather models converged, but meteorologists complained of problems with U.S. forecasting ability.
Why are ECMWF’s forecasts generally more accurate than those of the US NWS? If you have read my blog postings in the Saporta Report, you may be able to guess that the answer is. (Supercomputers and big data constitute one of the biggest technological trends. Weather forecasting is but one of their many uses.) The ECMWF uses a much more powerful computer, an IBM system capable of over 600 teraflops, that is one of the most powerful in the world. By contrast, the US NWS uses a much less powerful supercomputer. The supercomputers are used to run simulations of the atmosphere using global data from satellites, weather balloons and other observations. The model simulates the consequences of atmospheric physics into the near future.