Mesonet wind quality monitoring allowing assimilation in the RUC and other NCEP models

Mesonet observations (not including winds) began to be assimilated into the operational Rapid Update Cycle (RUC) running at NCEP in 2005. Mesonet winds were withheld since widespread poor siting of these observations frequently led to degraded surface wind analyses in preliminary testing.

An improved mesonet wind monitoring has been developed to allow assimilation of some mesonet winds into the RUC. Summaries of observation-background (O-B) differences are produced hourly in the RUC analysis output to monitor observation network behavior for dozens of different observation types. This capability has found to be very effective, allowing detection of past problems with some observation networks (e.g., certain types of aircraft or profilers).

A version of the RUC O-B statistics run hourly at NOAA/ESRL/GSD has now been extended to distinguish observation quality over 40 different mesonet wind providers, including the Oklahoma Mesonet, Automated Weather Source (AWS), RAWS, the Citizens Weather Observers Program, and many state-based networks. Differences between observations and 1-h RUC forecasts for temperature, wind speed, RH, and surface pressure are accumulated for each mesonet provider for each hour. From these accumulations, mean absolute differences and biases are calculated.

Based on this new hourly monitoring of mesonet wind biases, an objectively based mesonet wind provider uselist has been established and is now being used to safely allow assimilation of mesonet wind obsevations from specified providers in an experimental version of the RUC run at NOAA/ESRL/GSD. Assimilation of wind observations from the mesonet uselist providers is planned to begin with the operational RUC at NCEP starting in summer 2007 as part of a larger RUC change package.