Marked Improvements Shown in Global Weather Forecast Model

 
Published: 19 July 2007

Contact: Lynne Roeder, ARM Public Information Officer, 509.372.4331

Example of an ECMWF analysis.

One of the world’s foremost weather forecast models is showing dramatic improvements thanks to the pairing of two recent advancements in the representation of radiative transfer in global weather and climate models. Developed with funding from the Atmospheric Radiation Measurement (ARM) Program, the new components simulate the absorption and scattering of sunlight (“solar radiation”) in the atmosphere and better represent small-scale cloud variability. Their application to the forecast model of the European Centre for Medium-Range Weather Forecasts (ECMWF) solves a long-standing problem in which this model predicted too many convective thunderstorms over the ocean relative to those over near-by land areas.

The significant improvement immediately shown in the ECMWF forecasts was surprising even to the scientists who helped develop the new features.

“I’ve been at the Centre for 20 years and have never been able to achieve this level of impact any other way,” said Jean-Jacques Morcrette of ECMWF in Reading, England, who worked on linking the two new components.

When compared to actual weather observations, the combined codes do a much better job of representing the distribution of cloud convection over ocean and nearby land. Ten-day forecasts using the ECMWF model with the new codes show that improved weather forecasts persist for the entire 10-day cycle of forecasts.

The new solar radiative transfer model (RRTMG_SW) was developed by researchers at Atmospheric and Environmental Research (AER) Inc. It is based on highly detailed observations from the ARM Program and represents the way gases, including carbon dioxide, ozone and water vapor, absorb and scatter sunlight in the Earth’s atmosphere.

The Monte Carlo Independent Column Approximation (McICA) is an efficient, statistical technique for representing small-scale variations in cloud geometry and cloud properties. It was developed by researchers at Environment Canada, the Cooperative Institute for Research in Environmental Studies, and ECMWF.

Since summer 2000, the Centre has been using AER’s longwave radiation model (RRTMG_LW) to calculate the way energy emitted by the Earth is absorbed and reemitted by the atmosphere. After careful testing and evaluation, ECMWF began operational use of McICA and RRTMG_SW in its forecast model on June 5.

“This achievement is a very big deal, and one that I can’t envision happening without the ARM Program,” said Tony Clough, one of the scientists from AER who worked on the radiative transfer code.

The new components also provide the flexibility necessary to incorporate future advances in cloud physics resulting from the ARM Program and other research organizations. In the United States, both the National Center for Atmospheric Research and the Geophysical Fluid Dynamics Laboratory are investigating the use of these cloud and radiation improvements in their global climate models.

See also the press release from AER.

# # #