The Continued Improvement of NCEP's RTMA/URMA System and Benefits of Field Interaction toward the 2015 operational upgrades

NCEP's 2DVar-based Real Time Mesoscale Analysis (RTMA) and UnRestricted Mesoscale Analysis (URMA) systems have undergone several recent upgrades that result in a much-improved gridded surface analysis. A new background field based on the 3 km HRRR and 4 km CONUS NAM nest (rather than the 13 km RAP) is now being used, new analysis variables continue to be added and quality control continues to be updated to allow for more effective use of observations in the analysis. Several examples of improvements in the analysis will be presented from the scheduled April 2015 RTMA/URMA upgrade package. The examples will show an analysis with a better fit to surface observations and with an improved representation of fine-scale features, especially terrain-driven phenomena such as valley cold pools. Improvements to the precipitation fields will also be shown. Most of these improvements were based on feedback received from forecasters in the field. The methods used to solicit and collect this feedback will also be discussed. Highlights of the planned December 2015 upgrade will also be presented. They include improvements to the analysis scheme itself, to the model downscaling algorithm that produces the background field, and the use of a novel variational observation quality control scheme.