Short-range Forecasts of Satellite-derived Moisture and Stability Products using an Ensemble of Satellite Moisture Retrievals
Ralph Petersen (University of Wisconsin) talks about forecasting satellite-derived moisture using all available observation data on moisture and Lagrangian methods to give forecasters more information on the possibility of storm formation.
The CIMSS Lagrangian NearCasts system 1) expands the utility of clear-air sounding and products related to the pre-convective environment (from MTG-IRS Sounder and MTG-Imager) into the 1-9 hours period before storm formation and 2) now combines Geostationary Infrared Products with less frequent microwave products from multiple Polar Orbiting systems to fill information gaps in cloudy areas. For a heavy precipitation event, quantitative measures of both retrieval and short-range forecast accuracy are provided, including a new, non-uniform bias correction approach, and explorations of predictive clear-sky RGBs.
Including “all-weather”, real-time MiRS retrievals not only provide a more visually pleasing product (improving coverage by 30-40%), but also exposes forecasters to here-to-fore underutilized POES observations over land.
Validation against hourly surface-based GPS-TPW observations testify to the ability of the Satellite-based products to capture observed small-scale moisture features properly. Results show error growth rates < 1%/hour (without initial shocks) and support applying non-uniform bias corrections derived over 5mm bands to assure realistic TPW distributions.
Because RGB presentations are more popular than quantitative retrieval product displays the parcel-based Lagrangian NearCast approach was also used to “predict” clear-air RGBs based on the projection of radiance, with quantitative values overlaid. Examples of initial tests will be shown.