Monthly Forecast Improvement with Stratospheric Winds

It is widely accepted that two important influences on monthly forecast accuracy are ENSO and the Arctic Oscillation (AO). Recent investigations in the literature have shown that the stratospheric polar vortex has a significant impact on the AO and hence on the tropospheric weather in winter at northern U.S., northern Europe, and Canadian latitudes. Primary effects include cold air outbreaks and the alteration of storm tracks. Several new papers demonstrate that significant improvements in the mid-range forecast skill would be available if stratospheric winds were monitored globally. These investigations on stratosphere-troposphere coupling include data analyses addressing the effects of the vortex on the AO and increased forecast skill, as well as climate model and forecast model simulations thereof. The physical mechanism behind these strong correlations is now understood. There is also suitable satellite instrumentation available for investigation of this effect. Contributions using data reveal the nature of the coupling of the stratospheric vortex on the arctic oscillation. Model simulations have demonstrated the potential increases in the skill of mid-range forecasts. Currently, stratospheric winds are monitored by ~3000 daily radiosondes. However, the spatial distribution of the current radiosonde network provides poor coverage, being largely confined to major landmasses. Alternatively, a new-generation single satellite wind instrument could provide ~100,000 wind profiles per day extending continuously from ~15 to 200 km, day and night, hence full diurnal, global coverage. A constellation of these instruments on six nanosats would provide ~600,000 profiles daily – and define the mean, planetary wave, and tidal fields at up to 24 local times at each latitude, enabling a complete specification of the large-scale weather and “Space Weather” environment every orbit. The Doppler Wind and Temperature Sounder (DWTS) is a simple, small, light, and cheap IR imager employing gas correlation techniques for monitoring stratospheric and mesospheric winds and temperatures. Initial DWTS development, implementation, data analysis, and demonstration of forecast benefit, for a first instrument, would cost ~$20M. A 6-instrument suite aboard nanosats would cost <$80M to launch and validate. A DWTS 6-instrument suite would cost less than a single previous-generation wind instrument, <10% of the cost of the ESA ADM Aeolus wind lidar, and <1% of the currently planned NOAA or Air Force satellite weather and Space Weather monitoring programs. We argue that NOAA and Environment Canada should take immediate action to fly one DWTS instrument to assess the value of such data for ~3 to 30-day forecasting improvements that could have major societal benefits.