Evaluating Observation Innovation of Surface Temperature and Humidity for the Assimilation of the Observations in GFDL SHiELD Model

Host: Mingjing Tong

Assimilation of surface observations can potentially improve the prediction of surface variables. More accurate analysis of the atmospheric surface layer or even the entire planetary boundary layer (PBL) can also help improve severe weather prediction. Evaluating observation innovation (the difference between the observations and their model counterpart) is the first step toward assimilation of any new observations. This project is part of the project of assimilating air temperature and humidity from surface stations into GFDL System for High-resolution prediction on Earth-to-Local Domains (SHiELD), The evaluation and statistical analysis of the observation innovation will provide key information for quality control and observation error estimation.

The intern will:

  • Provide statistical evaluation and understanding of the surface temperature and moisture observation innovation.
  • Evaluate the innovations with different model or data assimilation configurations.
  • Test the effectiveness of quality control and observation error estimation
  • If time permits, learn to run a single cycle analysis to understand the impact of the surface observations.

Some knowledge of atmosphere science, programming skills (e.g., Unix/Linux, Python, Fortran), and statistical analysis and visualization skills, will be helpful. The student will gain knowledge of numerical weather prediction (NWP), learn concepts of data assimilation, have hands-on experience working with the observational data and numerical model, and understand the impact of the observations in numerical weather prediction. The student will also have the opportunity to interact with the FV3 modeling team at GFDL to gain understanding of the model and get help with model issues.