Be the first to post a text review of Pheno Analysis. Rate and review a project by clicking thumbs up or thumbs down in the right column.

• Project Information Updated

  stockli changed the public information on the Pheno Analysis project

  posted by stockli 158 days ago

• phenoanalysis pheno_analysis-1.1 file released: pheno_analysis-1.1.tar.bz2

  2009/07/17 Reto Stockli (MeteoSwiss / Colorado State University) This file documents changes since the Pheno Analysis Version 1.0 1.10 ---- - re-tested the 4 local and 1 regional experiments (regional experiment is shifted eastwards by 0.25 degrees) - added better VPD downscaling (first downscale temperature, then calculate downscaled VPD with large scale mixing ratio) - corrected bug in ERA-40 reading (ERA Interim was fine) 1.09 ---- - added support for Intel Compiler and Intel MPI - added GEOS4 reanalysis support - time averaging during assimilation for all parameters possible via parameter file (run/data/parameters.dat) higher averaging numbers enable to slow down convergence for parameters that converge too quickly. - reprocessed example datasets with new IDL codes in utils/ (pft's, topography, ECMWF Interim) 1.08 ---- - removed bug in MODIS observation error calculation - removed bad NetCDF output name (double definition of RAINMIN corrected) - new met driver module: cosmo_module.F90 reads COSMO NWP NetCDF surface weather forcing - added driver_module variables: pollon & pollat for treatment of rotated pole grids - pft, topo boundary conditions and met driver grids can be on rotated pole grids --> If both, pft and topo BC are rotated, then the model and the assimilation also runs on the rotated pole grid, and the output is on that grid as well 1.07 ---- - Non-F90 standard system() command removed from all modules. Fully replaced by the glob.c routine - Makefile PGI and XLF define statements removed (obsolete, since code is fully portable now) - fixed bug in pheno_analysis.pl script: did not correctly create runtime directory if it already was present from previous (aborted) run - all photoperiod parameters are now in units hours [h] instead of seconds [s] - optional flag simplifying the use of rainfall instead of vpd; or photoperiod instead of rg as predictors in phenology_module.F90 --> when those are selected then the PHOTOMIN/PHOTOMAX, or RAINMIN,RAINMAX parameters need to be selected for perturbation as well 1.06 ---- - length of sitename and gsitename(:) increased from 50 to 100 characters in driver_module.F90 and init_driver.F90 1.05 ---- - Automatic yearly job resubmission on NCAR bluefire (AIX, PBS) simplified - It is possible to run multiple instances of Pheno Analysis on a single host as long as they have a different experiment name. Before, Pheno Analysis had a single fixed temporary I/O directory, while now for each experiment a temporary output directory is created and driver/boundary/output data linked to this directory. Upon model termination, the temporary output directory is removed again - a new sites file: sites/sites_global.dat with 64 global sites, 4 sites per pft (except water) distributed over all continents that each contain an area with a pft majority for the respective pft - state assimilation disabled temporarily for both local and regional mode. We now only have parameters in the ensemble matrix since the main aim of this framework is to constrain parameters - several modifications in the analysis routines phenology_regional.pro and phenology_local.pro were performed 1.01 ---- - posteriori state and parameter inflation decreased from 1e-3 to 1e-6 times the initial state/parameter variance in assimilation_module.F90 - total time diagnostics overflow for pgf90 fixed (imcomplete yet) in driver.F90

  posted 175 days ago

• phenoanalysis pheno_analysis-1.1 file released: pheno_analysis_datasets-1.1.tar.bz2

  2009/07/17 Reto Stockli (MeteoSwiss / Colorado State University) This file documents changes since the Pheno Analysis Version 1.0 1.10 ---- - re-tested the 4 local and 1 regional experiments (regional experiment is shifted eastwards by 0.25 degrees) - added better VPD downscaling (first downscale temperature, then calculate downscaled VPD with large scale mixing ratio) - corrected bug in ERA-40 reading (ERA Interim was fine) 1.09 ---- - added support for Intel Compiler and Intel MPI - added GEOS4 reanalysis support - time averaging during assimilation for all parameters possible via parameter file (run/data/parameters.dat) higher averaging numbers enable to slow down convergence for parameters that converge too quickly. - reprocessed example datasets with new IDL codes in utils/ (pft's, topography, ECMWF Interim) 1.08 ---- - removed bug in MODIS observation error calculation - removed bad NetCDF output name (double definition of RAINMIN corrected) - new met driver module: cosmo_module.F90 reads COSMO NWP NetCDF surface weather forcing - added driver_module variables: pollon & pollat for treatment of rotated pole grids - pft, topo boundary conditions and met driver grids can be on rotated pole grids --> If both, pft and topo BC are rotated, then the model and the assimilation also runs on the rotated pole grid, and the output is on that grid as well 1.07 ---- - Non-F90 standard system() command removed from all modules. Fully replaced by the glob.c routine - Makefile PGI and XLF define statements removed (obsolete, since code is fully portable now) - fixed bug in pheno_analysis.pl script: did not correctly create runtime directory if it already was present from previous (aborted) run - all photoperiod parameters are now in units hours [h] instead of seconds [s] - optional flag simplifying the use of rainfall instead of vpd; or photoperiod instead of rg as predictors in phenology_module.F90 --> when those are selected then the PHOTOMIN/PHOTOMAX, or RAINMIN,RAINMAX parameters need to be selected for perturbation as well 1.06 ---- - length of sitename and gsitename(:) increased from 50 to 100 characters in driver_module.F90 and init_driver.F90 1.05 ---- - Automatic yearly job resubmission on NCAR bluefire (AIX, PBS) simplified - It is possible to run multiple instances of Pheno Analysis on a single host as long as they have a different experiment name. Before, Pheno Analysis had a single fixed temporary I/O directory, while now for each experiment a temporary output directory is created and driver/boundary/output data linked to this directory. Upon model termination, the temporary output directory is removed again - a new sites file: sites/sites_global.dat with 64 global sites, 4 sites per pft (except water) distributed over all continents that each contain an area with a pft majority for the respective pft - state assimilation disabled temporarily for both local and regional mode. We now only have parameters in the ensemble matrix since the main aim of this framework is to constrain parameters - several modifications in the analysis routines phenology_regional.pro and phenology_local.pro were performed 1.01 ---- - posteriori state and parameter inflation decreased from 1e-3 to 1e-6 times the initial state/parameter variance in assimilation_module.F90 - total time diagnostics overflow for pgf90 fixed (imcomplete yet) in driver.F90

  posted 175 days ago

• File released: /phenoanalysis/pheno_analysis-1.1/pheno_analysis-1.1.tar.bz2

  posted 175 days ago

• File released: /phenoanalysis/pheno_analysis-1.1/pheno_analysis_datasets-1.1.tar.bz2

  posted 177 days ago

• phenoanalysis pheno_analysis-1.0 file released: pheno_analysis.tar.gz

  No Changes since Version 1.0

  posted 299 days ago

• phenoanalysis pheno_analysis-1.0 file released: pheno_analysis_datasets.tar.gz

  No Changes since Version 1.0

  posted 299 days ago

• phenoanalysis pheno_analysis-1.0 file released: stockli_modis_data_assimilation_phenology.2008.pdf

  No Changes since Version 1.0

  posted 299 days ago

• phenoanalysis pheno_analysis-1.0 file released: pheno_analysis_manual.pdf

  No Changes since Version 1.0

  posted 299 days ago

• File released: /phenoanalysis/pheno_analysis-1.0/pheno_analysis_manual.pdf

  posted 301 days ago