Enter Note Done
Go to previous page in this tab
Session
  • Presentation | A41F: Numerical Modeling, Data Assimilation (DA), Artificial Intelligence (AI), and Research to Operations (R2O) for Better Analyses and Forecasts of High-Impact Weather Events II Oral
  • Oral
  • Share Icon
    Map Icon
    Bookmark Icon
  • A41F-07: All-sky assimilation of GIIRS infrared hyperspectral radiance in CMA-GFS using 4D-Var
    • Schedule
    Notes
  • Location Icon278-279
    NOLA CC
    Set Timezone
    •  
    View Map

Generic 'disconnected' Message
Author(s):
Senyi Kong, Zhejiang University (First Author, Presenting Author)
Wei Han, Numerical Weather Prediction Center, China Meteorological Administration
Lei Bi, School of Earth Sciences, Zhejiang University


Predicting the weather accurately is really important, and satellites that measure infrared radiation in clear skies help a lot. But what about when there are clouds? It’s much harder to use that satellite data when clouds are present. Our study tackled this problem by improving a weather forecasting system used in China, called CMA-GFS. We developed several new techniques to better use cloudy-sky satellite data, including ways to correct for instrument quirks and account for how clouds affect measurements.


We tested our new methods using 50 specific channels that measure water vapor. The results were exciting: we could use almost 60% of the cloudy-sky satellite observations, compared to only about 3.6% in clear skies before. This led to much better predictions of moisture in the atmosphere, improved wind forecasts, and even more accurate paths for typhoons. Ultimately, our work helps us make better medium-range weather predictions, even when the sky is not perfectly clear.




Scientific Discipline
Neighborhood
Type
Main Session
Discussion