Historical Weather Reconstruction by Cloud Cover Data Assimilation with Gaussian Transformation
Abstract
Due to the limitation of instrumental weather records before the nineteenth century, old descriptive weather records (i.e., old diaries) are significantly important in reconstructing the historical weather on a daily scale. Recently, data assimilation is widely used because it is an effective method to combine model forecasts and various types of meteorological observations. Cloud cover information derived from historical documents demonstrates great potential for reconstructing historical weather. However, cloud cover often shows a non-Gaussian distribution and seriously violates the basic assumption of normal error statistics in most data assimilation schemes. This study aims to assimilate cloud cover more accurately by applying the Gaussian transformation and to achieve effective weather reconstruction. To confirm the feasibility of Gaussian transformation (GT), Observing System Simulation Experiments (OSSEs) are conducted with the Global Spectral Model (GSM) and a local ensemble transform Kalman filter (LETKF). A control experiment without any observation assimilated and three experimental runs with cloud cover assimilated are performed from 1 July 2017 around Japan. Results indicate that at mid-troposphere, 2-month average RMSE of zonal wind, meridional wind, temperature, and specific humidity can be reduced by 8.4%, 5.3%, 4.1%, and 0.8% by GT of cloud cover, respectively. At the surface, a 2-month average RMSE of cloud cover, pressure, rainfall, and shortwave solar radiation can be reduced by 2.2%, 5.2%, 31.0%, and 4.3% by GT of cloud cover, respectively. These results demonstrate that the GT approach has the potential to improve cloud cover assimilation accuracy and historical weather reconstruction using old diaries. Furthermore, historical reconstruction with the GT approach is conducted using actual dairy data during 1831-1833. Since weather information recorded in diaries is less accurate than modern weather, a GT approach with five categories is utilized in historical weather reconstruction. Results based on the category data assimilation will be presented.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMA215.0006W
- Keywords:
-
- 3315 Data assimilation;
- ATMOSPHERIC PROCESSES;
- 0520 Data analysis: algorithms and implementation;
- COMPUTATIONAL GEOPHYSICS