COSP 2 now available
30 views
Skip to first unread message
Robert Pincus
Nov 6, 2017, 12:05:00 PM11/6/17
to cosp...@googlegroups.com
Dear COSP user community -
After several years of hard work by Dustin Swales we are pleased to announce the availability of COSP 2.0. This version substantially revises the infrastructure on which COSP is based, cleanly separating the conversion of model state to optical properties, the pixel-scale pseudo-retrievals performed by each simulator, and the aggregation of pixel-scale retrievals from one or more simulators to grid-scale diagnostics. As one result COSP’s flexibility in representing model clouds is greatly enhanced at the cost of somewhat greater responsibility for users in constructing inputs.
The code is available at https://github.com/CFMIP/COSPv2.0 ; the tag v2.0 is archived at doi:10.5281/zenodo.1040332. A short manuscript describing the code is under review at https://www.geosci-model-dev-discuss.net/gmd-2017-148/ .
COSP 2 has been approved for use in CFMIP3/CMIP6 by the CFMIP Coordination Committee. This use is strictly optional; version 1.4.1 remains a perfectly good choice. COSP2 includes an interface that replicates COSP 1.4.1. Results with this version are scientifically equivalent to version 1.4.1 but not bit-wise reproducible. Note too that this version involves copying relatively large amounts of data. In the Met Office model this copying increases the computational cost of COSP by 20%. More direct implementations can be faster because common inputs are reused much more frequently.
Future developments for COSP, including the development of new simulators and/or diagnostics, should be targeted at COSP 2. We would strongly encourage anyone adding COSP to a new model to begin with COSP 2 as well.
Please direct question and comments to this list.
---
Robert Pincus
University of Colorado/NOAA Earth System Research Lab
http://cires.colorado.edu/researcher/robert-pincus
After several years of hard work by Dustin Swales we are pleased to announce the availability of COSP 2.0. This version substantially revises the infrastructure on which COSP is based, cleanly separating the conversion of model state to optical properties, the pixel-scale pseudo-retrievals performed by each simulator, and the aggregation of pixel-scale retrievals from one or more simulators to grid-scale diagnostics. As one result COSP’s flexibility in representing model clouds is greatly enhanced at the cost of somewhat greater responsibility for users in constructing inputs.
The code is available at https://github.com/CFMIP/COSPv2.0 ; the tag v2.0 is archived at doi:10.5281/zenodo.1040332. A short manuscript describing the code is under review at https://www.geosci-model-dev-discuss.net/gmd-2017-148/ .
COSP 2 has been approved for use in CFMIP3/CMIP6 by the CFMIP Coordination Committee. This use is strictly optional; version 1.4.1 remains a perfectly good choice. COSP2 includes an interface that replicates COSP 1.4.1. Results with this version are scientifically equivalent to version 1.4.1 but not bit-wise reproducible. Note too that this version involves copying relatively large amounts of data. In the Met Office model this copying increases the computational cost of COSP by 20%. More direct implementations can be faster because common inputs are reused much more frequently.
Future developments for COSP, including the development of new simulators and/or diagnostics, should be targeted at COSP 2. We would strongly encourage anyone adding COSP to a new model to begin with COSP 2 as well.
Please direct question and comments to this list.
---
Robert Pincus
University of Colorado/NOAA Earth System Research Lab
http://cires.colorado.edu/researcher/robert-pincus
Bodas-Salcedo, Alejandro
Nov 7, 2017, 8:36:31 AM11/7/17
to cosp...@googlegroups.com
Dear Dustin and Robert,
On behalf of the COSP community, I would like to thank you for all the hard work that you have put into this new version of COSP. This is a great achievement that will facilitate the development of new science through the use of new simulators and multi-instrument diagnostics.
All the best,
Alejandro
> --
> You received this message because you are subscribed to the Google Groups
> "COSP users group" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
> cosp-user+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
On behalf of the COSP community, I would like to thank you for all the hard work that you have put into this new version of COSP. This is a great achievement that will facilitate the development of new science through the use of new simulators and multi-instrument diagnostics.
All the best,
Alejandro
> You received this message because you are subscribed to the Google Groups
> "COSP users group" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
> cosp-user+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
Reply all
Reply to author
Forward
0 new messages