Observation Impact on WRF Model Forecast Accuracy over Southwest Asia. Michael D. McAtee Environmental Satellite Systems Division (ESSD) User Applications and Integration (UA&I) The Aerospace Corporation. ESSD/UA&I May 2014. Approved for Public Release – Distribution Unlimited . Overview. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
• Based on results presented in a JCSDA newsletter, Air Force Weather Agency (AFWA) personnel were encouraged to pursue the development of an observation impact assessment tool based on techniques developed at the Naval Research Laboratory (NRL)
• AFWA tasked the National Center for Atmospheric Research (NCAR) with developing such a tool that could be used with its operational forecast model and data assimilation system (WRF and WRF DA) leveraging previous work that had been done to develop 4DVAR
• Aerospace working with NCAR, AFWA and its contractors installed the system in AFWA’s development environment
Forecast Sensitivity to Observations
Aerospace Corporation
Joint Center for Satellite Data Assimilation (JCSDA)
Aerospace Corporation
Weather Research and Forecasting (WRF) modelWeather Research and Forecasting model Data Assimilation (WRF DA) system
• The tool, which is called Forecast Sensitivity to Observations (FSO), has the ability, using adjoint methods, to quantitatively estimate the impact that assimilating observations has on short-range WRF model forecast accuracy
• An FSO capability was subsequently developed by NCAR for AFWA that works with the Gridpoint Statistical Interpolation (GSI) system
• The FSO system used in this study consists of the GSI with Lanczos minimization, WRF, and the adjoint to WRF
• Meets a need to know not only the impact of observations but also their relative value
• A single run of FSO can provide the relative value of all observations assimilated without the need for multiple and computationally expensive with- and without- model runs
• FSO can provide the critical information needed to intelligently select which channels from space based remote sensors will be assimilated
• FSO can be used to monitor the health of an NWP center’s data assimilation system as well as the health of the observations it uses
• FSO can determine which individual observations improved or degraded the forecast
• The GSI version of the observation impact assessment tool Forecast Sensitivity to Observations was used to determine the relative impact of various types of weather observations on 12-hour WRF model forecast accuracy over South West Asia
• Radiance data accounted for nearly 75% of the total impact achieved through assimilating all observations
• The assimilation of ATMS, CrIS, and SSMIS data was successful and acted to reduce the forecast error
• Apply the FSO tool to more domains which are higher resolution ~15 KM