Investigate datasets from SeDOO

[nsoontie]

Decide what information/datasets from the SeDOO gpsc mirror should be added to the navigator.

[nsoontie]

Strategy: I will write my comments/suggested action for each dataset separately

My comments reflect the datasets found here on the gpsc: gpsc:/gpfs/fs2/dfo/hpcmc/gridpt/sdfo501/hubs/netcdf

NOTE: This is all up for discussion and nothing should be implemented until we are all in agreement.

[nsoontie]

GIOPS

giops_analysis_daily

• These are daily instantaneous snapshots from the giops daily analysis cycle at 00 (gu). Data is on the user grids - latlon and polar stereographic
• The analysis cycles generally have some form of data assimilation and are used to provide an initial condition for the forecasts.
• Is there a use case for the analysis or would we prefer to show daily averages calculated from the forecast?
• Suggestion
  • do not include on ON at the moment
  • develop a strategy for if/how to include analysis data

giops_forecast

• These are from the giops 10 day forecasts on the user grids - lat lon and polar stereographic
• 2D files are three hour means
• 3D files are daily means
• There are two forecasts a day: 00 and 12
• Suggestion
  • display only the latlon user grid on the ON. Ignore the polar stereographic data
  • develop a strategy for handling the two forecasts

giops_forecast_native

• There are 3D velocity fields from the 00 forecast on the "native grid" (ie a-grid)
• The data is only available at 24hr into the 00 forecast. It is unclear if this is a snapshot or a daily average.
• There also appears to be a RPN standard file in this directory.
• Suggestion
  • do not include on the ON

giops_forecast_usr

• This is identical to giops_forecast.
• Suggestion
  • do not include on ON

@dwayne-hart @FraserO Please comment.

[nsoontie]

RIOPS

riops_analysis_daily

• These are from the riops daily analysis cycle (ru).
• There are several types of files:
  • daily instantaneous snapshots on 3D U/T/V grid (velocities, temp, salinity)
  • hourly averages of surface T-grid fields (ssh type variables only)
  • daily instantaneous snapshots on polar stereographic grid (2D/3D in separate files)
• Suggestion
  • develop a strategy for including analysis data before adding to the ON

riops_analysis_weekly

• These are from the riops weekly analysis cycle (rr).
• The types of files are the same as those in riops_analysis_daily
• There is a longer history of files than in the riops_analysis_daily (7 days back vs 2 days back)
• Suggestion
  • Figure out the difference between riops_analysis_daily and riops_analysis_weekly
  • Decide if it would be useful to include one or both of the the analysis runs on the ON
  • develop a strategy for including analysis data before adding to the ON

riops_coupled_forecast

• These are from the riops coupled forecast (caps)
• The 2D and 3D files are 3-hour means on the polar stereographic grid
• There are two forecasts a day: 00 and 12. Each forecast is 48 hours
• Do we need to have multiple instances of riops on the ON? It might get confusing....
• Suggestion
  • decide how to organize the presentation of riops data before adding to the ON.

riops_forecast

• These are from the riops forecasts (uncoupled)
• The 2D and 3D files are hourly instantaneous snapshopts.
• The data is on the polar stereographic grid (2D+3D) and latlon grid (2D only)
• There are 4 forecasts a day: 00, 06, 12, 18. Each forecast is 48 hours
• Suggestion
  • This data should be on the ON.
  • Develop a strategy for handling the four forecasts.
  • We need to revisit how the daily averages are calculated.

@dwayne-hart @FraserO Please comment.

[nsoontie]

Other models

ciops_pseudo_analysis

• These are files from the CIOPS-W pseudo-analysis run cycle
• There are 2D hourly instantaneous outputs on T/V/U grid.
• There are 3D daily instantaneous outputs on T/V/U grid.
• There are other files that appear to be subsets/stations at a high frequency.
• The files are zipped in separate 3D/2D folders.
• Suggestion
  • Display on navigator, labelled as CIOPS-W analysis
  • Only use data in the grid_T, grid_V, grid_U files
  • 2D and 3D should be separate datasets
  • If/when we receive forecasts from CIOPS-W we will need to develop a strategy for showing both forecasts and analysis

gsl_forecast

• These are from the Gulf of St. Lawrence forecasts
• There are 3D velocity fields on the "native grid" (a-grid)
• There are instantaneous daily outputs
• Suggestion
  • Don't show on the ON

hdrps_forecast

• These are u_wind/v_wind from the hrdps forecasts
• The winds are at 10 m.
• They are daily, instantaneous values
• It is unclear if the winds are aligned with the model grid or north/south
• Suggestion
  • Don't include on the ON

rdsps_forecast

• These are from the regional storm surge forecast. It appears to be a 10 day forecast launched at 00 and 12
• In the datamart directory, we have instantaneous storm surge every three hours on a lat lon grid. There is one timestamp per file.
• In the main directory, we have instantaneous hourly outputs for each forecast in a single file (241 time outputs) on the same grid as above
• Suggestion
  • Learn more about this product before adding to the ON.
  • For example, what is eta? Would there be interest in seeing just eta or is the total water level (including tides) needed?

@dwayne-hart @FraserO Please comment