ISSN 1006-9895

CN 11-1768/O4

Error diagnosis and Assessment of Sub-seasonal forecast using GRAPES-GFS
Author:
Affiliation:

1.NOAA/NWS/NCEP/EMC, College Park, Maryland, USA;2.Numerical Weather Prediction Center of CMA

Fund Project:

  • Article
  • |
  • Figures
  • |
  • Metrics
  • |
  • Reference
  • |
  • Related
  • |
  • Cited by
  • |
  • Materials
    Abstract:

    Using the analysis and predictions of leading 35 days related to Global and Regional Assimilation Prediction System (GRAPES)-Global Forecast System (GFS) during the period from September 2018 to August 2019, we diagnosed the prediction errors and evaluated the extended forecast capability to improve a numerical weather guidance for the sub-seasonal timescale. Result show that, GRAPES-GFS could capture the spatial distribution characteristics of 2m temperature and 500hPa geopotential height during winter in 2018 and summer in 2019, however there exists large system bias related to 2m temperature analysis in the desert plateau areas which have the thermal forcing effect significantly, especially in arid areas of Africa. Related to 2m temperature, the Root-Mean-Square Errors (RMSE) of leading 1 to 3 weeks predictions approximate to the linear growth. GRAPES-GFS posses the high prediction skill in the East Asia and Austria but have the lower prediction skills in the ocean areas compares with the land areas. Related to 500hPa geopotential height, when leading 1 to 3 weeks predictions, there exist higher prediction skills in low latitude than in high latitudes of East Asia. Also, the prediction skills of the tropics is much lower than other regions and the northern hemisphere is higher than the southern hemisphere. Related to the Madden-Julian Oscillation (MJO), it is found that, GRAPES-GFS can reproduce the propagation characteristic of spatial-temporal variations related to the upper and lower zonal wind and can capture the location of strong convective activity signals. However, the Outgoing Long Wave Radiation (OLR) positive anomaly is much weaker and the negative anomaly is much stronger. GRAPES-GFS have a skillful MJO forecast for 11 lead days when it is useful for ACC and for the selected two strong MJO cases, GRAPES-GFS could describe the MJO propagation process exactly but have a stronger signal during MJO developing and decaying period.

    Reference
    Related
    Cited by
Get Citation
Share
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
  • Cited by:
History
  • Received:May 21,2020
  • Revised:August 04,2020
  • Adopted:December 23,2020
  • Online:
  • Published: