Bias Correction and Calibration of Crowdsourced Surface Pressure Observations for Use in High-Resolution Numerical Weather Prediction

This talk will provide an update on a multiyear project to demonstrate the value of massive amounts on crowdsourced surface pressure data on the skill of high-resolution numerical weather prediction over the U.S. Specifically, the UW group is acquiring sea level data from thousands of locations, including smartphone pressures and commercial networks such as Tempest and PurpleAir. Millions of such observations per day are then biased corrected and calibrated using reliable surface network data. It this talk we will talk about some important challenges, including dealing with errors in position in regions of steep terrain. The calibrated pressure observations are then transferred to the NOAA GFL group, which is applying this data in experimental forecasts using the operational RRFS system (3 km). The talk will describe the current state of this effort to improve mesoscale NWP.