Wiki » History » Version 13

« Previous - Version 13/15 (diff) - Next » - Current version
Malcolm Roberts, 02/09/2017 01:33 PM


Recent simulations with high-resolution climate models have demonstrated the added value of enhanced resolution. They showed significant improvement in the simulation of aspects of the large scale circulation such as storm tracks and blockings, compared to the CMIP3 and CMIP5 archive. In addition, the increased resolution enables more realistic simulation of small scale phenomena with potentially severe impacts such as tropical cyclones and polar lows.

Until now these high-res simulations have been performed in a few research centers without an overall coordination. Due to the large computers resources needed for these simulations synergy will be gained if these runs are done in a coordinated way, which enables the construction of a multi-model ensemble with common integration periods, forcing and boundary conditions. The CMIP3 and CMIP5 data bases provide outstanding examples of the success of this approach. The multi-model mean has proven often to be superior to individual models in seasonal and decadal forecasting. Moreover, significant scientific understanding has been gained from analyzing the inter-model spread and attempting to attribute to model formulation.

The project will coordinate the efforts in the high-res modeling community. As a consequence of the large computer resources, the number and length of the high-res simulations will be limited. This emphasizes the need to perform these simulations with a common well defined protocol that enables inter-model comparison and the construction of a multi-model ensemble. Joint analysis, based on process-based assessment and seeking to attribute model performance to emerging processes, will further highlight the impact of enhanced resolution on the simulated climate.

  • The latest draft of the protocol and the interested institutes can be find at Documents
  • HadISST2.2.0.0 SST and sea-ice forcing on 1/4 degree daily grid * The ability to produce 0.25 degree fields of actual SST comes from our use of a satellite-era climatology. The analysis of residuals from that climatology (aka anomalies) which we do on a 5-day timescale is performed on a 1 degree grid. On that spatial and temporal scale, that analysis is perfectly achievable with the data available and is done both for pre-satellite data and satellite-era data alike. Information on covariances between anomalies in different locations gained from the satellite (and in situ) data enables the analysis to be done at the same resolution throughout. The ensemble captures the uncertainty in this process and its spread is wider prior to the satellite era.

Rein Haarsma
Malcolm Roberts