WRF model moisture adjustment method; a case study with winter-time cloudy biases in Xinjiang, China.

View More View Less
  • 1 Envision Digital International Pte Ltd, Singapore.
  • 2 Shenzhen Institute of Artificial Intelligence and Robotics for Society, and Institute of Robotics and Intelligent Manufacturing, The Chinese University of Hong Kong, Shenzhen, China
© Get Permissions
Restricted access

Abstract

For most locations on Earth the ability of a Numerical Weather Prediction (NWP) model to accurately simulate surface irradiance relies heavily on the NWP model being able to resolve cloud coverage and thickness. At horizontal resolutions at or below a few kilometres NWP models begin to explicitly resolve convection and the clouds that arise from convective processes. However, even at high resolutions, biases may remain in the model and result in under- or over-prediction of surface irradiance. In this study we explore the correction of such systematic biases using a moisture adjustment method in tandem with the Weather Research and Forecasting model (WRF) for a location in Xinjiang, China. After extensive optimisation of the configuration of the WRF model we show that systematic biases still exist—in particular for wintertime in Xinjiang. We then demonstrate the moisture adjustment method with cloudy days for January 2019. Adjusting the relative humidity by 12% through the vertical led to a Root Mean Square Error (RMSE) improvement of 57.8% and a 90.5% reduction in bias for surface irradiance.

Abstract

For most locations on Earth the ability of a Numerical Weather Prediction (NWP) model to accurately simulate surface irradiance relies heavily on the NWP model being able to resolve cloud coverage and thickness. At horizontal resolutions at or below a few kilometres NWP models begin to explicitly resolve convection and the clouds that arise from convective processes. However, even at high resolutions, biases may remain in the model and result in under- or over-prediction of surface irradiance. In this study we explore the correction of such systematic biases using a moisture adjustment method in tandem with the Weather Research and Forecasting model (WRF) for a location in Xinjiang, China. After extensive optimisation of the configuration of the WRF model we show that systematic biases still exist—in particular for wintertime in Xinjiang. We then demonstrate the moisture adjustment method with cloudy days for January 2019. Adjusting the relative humidity by 12% through the vertical led to a Root Mean Square Error (RMSE) improvement of 57.8% and a 90.5% reduction in bias for surface irradiance.

Save